Obligation

Obligation is normative, concerned with how things should be (not necessarily with how they are). In particular, it is a deontically normative—concerned with duty, or what is permissible, forbidden, and the like—as opposed to axiologically normative, concerned with what is good, bad, better, or worse. A given deontic normative perspective classifies actions, laws, institutions, or whatever as permissible (acceptable) or impermissible (unacceptable, wrong, forbidden). The classifications are interdefinable. For instance, an option is permissible when it is not impermissible, obligatory when permissible and every alternative option is impermissible.

Obligation is a relation with as many as three places: the person or entity who has an obligation (call this the agent), what that agent is obligated to do (the performance), and the person or entity to whom the agent owes the performance (the object). Because the object of an obligation is owed the performance, that person or entity has a right to it, and when an obligation has an object, obligations and rights are reciprocal: if someone has a right to your ing, then you have an obligation to; and if you have an obligation to someone to, then that someone has a right to your ing. Philosophers differ over whether all or just some obligations have objects, and so whether all or just some obligations come with reciprocal rights. For instance, some hold that a person is morally obligated to help others, but this obligation does not give anyone a right to be helped. Often, obligations that do not generate reciprocal rights for their objects are called imperfect obligations, while those that do generate reciprocal rights are called perfect obligations.

Some obligations are agent-relative (or agent-centered) and others are agent-neutral. Agent-relative obligations are those whose performance contains within its scope an essential reference back to the agent and to the agent carrying out that performance (McNaughton and Rawling, 1991). Hence, such obligations state that the agent is obligated himself to perform some action. An agent-neutral obligation, by contrast, states only that the agent ought to ensure the performance of the action, whether by himself or someone else. For instance, my obligation to help a stranger is really just the obligation that I make sure the stranger is helped—whether by me or by someone nearby does not matter. By contrast, my obligation to go to my son's birthday party is an obligation that I go to his party. I do not discharge that obligation by sending someone else. Likewise, if I promise that I will baby-sit your children, the obligation I have undertaken is that I baby-sit, not that a baby-sitter of some sort is provided. To the degree that obligations and rights are reciprocal, this distinction in agent-relative and agent-neutral obligations brings with it a distinction in agent-relative and agent-neutral rights. For instance, you have an agent-relative right that I baby-sit if I have promised as much.

What is obligatory is what one must do. A citizen has an obligation to obey the law, and so must obey the law; it is obligatory for a Boy Scout to keep himself physically strong, thus he must do so; parents are under an obligation and so must care for their children. Because of this, deontic classifications have been treated as modal notions—as what it is normatively necessary, possible, or impossible to do. Moreover, these modalities are often characterized much like causation and other kinds of necessity, as relations grounded in the governance of laws. Heat must boil water because there is a natural law governing these events; likewise, a Boy Scout must keep himself physically strong because there is a Scouting law governing his actions. Thus, an action is obligatory when a normative law or rule makes it normatively necessary, permissible when a normative law makes it normatively possible, and so on. The differences between moral, legal, etiquette, and club obligations can thus be made out as differences in the sets of normative rules or laws grounding the modal categorization. For instance, a club obligation is an action made necessary by the rules of the club.

An obligation is thus typically held to be an action that is normatively necessary according to some normative rule or law. A normative law, however, does not necessitate one's action, in the way that the law of gravitation makes a person stay on or near the surface of the earth. The laws of a state, club, or morality necessitate one's action not as nature or logic makes things necessary, because it is logically and naturally possible for a person not to act as these laws direct. What is normatively necessary is both logically and naturally not necessary, even if both logically and naturally possible.

For any deontic realm, there will be obligations within or defined by that realm, and there may also be obligations not defined by that realm, but whose performance is regarding the obligations within that realm—separate obligations to fulfill the obligations laid down by that deontic realm (Green, 1988). For instance, legal systems define certain legal obligations, such as the obligation not to exceed a certain speed on a given road. These are obligations within the law. Yet a club, such as the Boy Scouts, may define certain obligations regarding those obligations created within the law. Hence, one obligation of the Boy Scouts is never to fail in one's legal or moral obligations. This is an obligation within one deontic realm, the Boy Scouts, regarding the obligations within a distinct deontic realm, the law. In general, if there is an obligation regarding a given deontic realm (i.e., an obligation to fulfill the obligations laid down by a club, a state, or etiquette), then there must be some second deontic realm independent of it within which the obligation regarding that realm exists. One of the most important kinds of obligation regarding a deontic realm is the moral obligation to conform to legal obligations, often termed political obligation.

Normative laws can make one's actions necessary in a variety of ways. One way in which they might do this is if the laws are created as means to ends that one has. In order to vote, one must register; in order to get badges in the Boy Scouts, one must keep oneself physically strong; and so on. This can be called instrumental necessity. Many have thought that the necessity of any obligation must be instrumental. The laws of a state, for instance, would generate obligations for a given person when and only when those laws necessary are necessary for her security and the avoidance of sanctions for disobedience.

Most philosophers, however, have thought instrumental necessity is neither necessary nor sufficient for deontic necessity. It is not sufficient because a law backed up by sanctions alone seems not to be an obligation but rather extortion. The law must at least be in some sense a legitimate law of that deontic realm. Following the legal philosopher John Austin, a mechanism to establish the legitimacy of a normative law or rule can be called a pedigree. Systems of normative rules or laws require a pedigree that establishes the legitimacy of those laws or rules in creating a genuine obligation.

Instrumental necessity is not necessary for deontic necessity either. For instance, according to the rules of chess, it is necessary that the bishop stay on his color; this is a normative necessity arising from a rule of chess. But that rule is not backed up by some clause such as "in order to avoid such and such penalty." The rules of chess are in this sense "categorically" necessary. What is required for deontic necessity are laws that are nonoptional rules of behavior, and that at a minimum have a pedigree that makes them legitimate rules of the system.

Different deontic realms can have different pedigrees, and philosophers have offered different pedigrees for some deontic realms. For instance, philosophers such as St. Thomas Aquinas tied the legitimacy of a state's laws to divine pedigree. Genuine obligations, both within and to the law, arise only from demands conforming to laws God laid down for our nature when he created us. Hence, genuine legal obligations arise only from laws that comport with this natural law. And the moral obligation to obey the law is grounded in our obligation to obey God. The view that the legitimacy of law arises from acts of will, or "voluntarism," underwent a transformation as the vision of the universe grounded in God's will receded (for recent accounts of the history of modern thought about obligation, see Darwell, 1995, and Schneewind, 1998). Human acts of will—acts of choice, consent, agreement, promise—came to take the place of God's will as a source of authority. Thomas Hobbes and John Locke, for instance, rejected Thomas's view that each person's good naturally comports with the good of others. Although we are naturally entitled to seek our own good, each person's good is at odds with that of others. This conflict inevitably leads to a devastating loss for all as everyone tries to secure their own good at the cost of others. It is only by leaving this natural state that all can secure their good. Each person's natural authority over himself is transferred to the sovereign or the community, which then comes to possess an authority to enact and enforce laws obligating them. And our obligation to obey those laws is based on something like a promise we have made.

Some version of the law-based source of obligation and some version of voluntarism as a source of the legitimacy of those laws and the obligation to obey has held a dominant position among philosophers since the modern era began. Nevertheless, many have defended nonvoluntarist positions and/or have rejected the model of obligations as actions made necessary by a law. Some, for instance, hold that obligations are a special kind of reason that make actions necessary. Hence, the obligation to obey the law, for instance, is a reason that necessitates obeying it. Among voluntarism's important detractors was David Hume, who had misgivings both about the law-based analysis of obligation and about voluntarism as the required pedigree. Hume argued that consent "has very seldom had place in any degree, and never almost in its full extent" (II.XII. 20). We are born into or find ourselves with obligations. Obligations that we have because of our consent, such as promises, are few and far between. Hume's own view of obligation was complex and hard to deem entirely consistent. Nevertheless, the gist of his view was that deontic necessity should be explained in subjective terms: When I judge that someone has an obligation to perform some action, I judge that there is a admirable motive impelling her to do it. The motive might be unreflective and natural, as is the love of children. Or it may arise from reflection on the advantages of various institutions, such as property or promising. Such a motive gains legitimacy through my approval of it from an impartial point of view.

John Stuart Mill followed the spirit of this pattern of analysis, but focused on external actions rather than internal motives. His still influential account holds that an action is obligatory just in case failure to perform it justifies internal or external sanctions, "if not by law, by the opinion of his fellow creatures; if not by opinion, by the reproaches of his own conscience" (pp. 48–49). The pedigree granting authority to such internal or external punishments in the case of moral and legal obligations is general happiness. As David Lyons (1977) has argued, however, Mill's analysis in terms of justifiably sanctionable behavior is itself quite separate from his view of what justifies such sanctions, namely, the general happiness. Indeed, anarchists argue for a given analysis of political obligation but then deny that any such thing exists on the grounds that no legitimacy can be given for it (see, e.g., Simmons, 1979; Wolff, 1998).

Most philosophical positions separate the legitimacy or authority of a law laying down obligations (moral or legal) from the reason to comply. Whether a rule or law in some deontic realm is valid is one thing; whether a person has reason to comply with its obligations is quite another. The most famous attempt to unify the legitimacy and reason to comply with moral obligations was that of Immanuel Kant. Kant's insight was to view moral obligations as stemming from laws that each person imposes on himself rather than from laws whose origin is external to the agent. In particular, moral laws are requirements stemming from deep inside—from a person's very capacity to act and choose on the basis of reason itself. If the source of a law was the moral agent himself, that is, if it were created and enacted by the agent's own rational will, then its authority would come from the agent himself, namely, the authority of his own rational will over his actions. Kant referred to this as the "autonomy" of a rational will. Moral obligations, in his view, stem from the demands of our own reason, and it is in virtue of this source that we have reason to comply.