MORAL MOTIVATION
Ethics appears to be a normative discipline.  When discussing the normative adequacy of conceptions of practical reason, we talked about two dimensions of normativity -- motivational pull and normative authority.  In thinking about the normativity of ethics, we might focus on moral motivation.  But moral motivation involves a puzzle. Cognitivists interpret moral judgments as expressing cognitive attitudes, such as belief, rather than noncognitive attitudes, such a desire.  Internalists believe that moral judgments necessarily engage the will and motivate.  It is a common view that motivation involves pro-attitudes, such as desires, and that no belief entails any particular desire.  But these assumptions are in tension, as can be seen in this puzzle about moral motivation.

1. Moral judgments express beliefs.
2. Moral judgments entail motivation.
3. Motivation involves a desire or pro-attitude.
4. There is no necessary connection between any belief and any desire or pro-attitude.

THE AUTHORITY OF MORALITY
Another way to think of the normativity of ethics is, not in terms of whether morality does motivate, but in terms of whether it should.  One way to understand this aspect of normativity is in terms of the authority of morality.  But the authority of morality raises its own puzzles.  It is common to think of morality as both objective and impartial, in particular, as containing various other-regarding duties of cooperation, forbearance, and aid that apply to agents independently of their own aims and interests.  Most of us also regard moral obligations as authoritative practical considerations.  But heeding these obligations appears sometimes to constrain the agent's pursuit of his own interest or aims.  If we associate rationality with the agent's own point of view, we may wonder whether moral conduct is always rationally justifiable.  We can capture this tension in terms of a puzzle about the authority of morality.

1. Moral requirements include impartial other-regarding obligations that do not apply to agents in virtue of their aims or interests.
2. Moral requirements necessarily provide agents with reasons for action.
3. Rational action is action that advances the agent's aims or interests.
4. Fulfilling other-regarding obligations need not advance the agent's aims or interests.

RELATIVISM AND MINIMALISM
We might appeal to (2)-(4) to reject (1).  One view that is relativist is Harman's in "Moral Relativism Defended".  He combines rationalism about moral requirements (expressed by "inner" judgments) with an instrumental conception of practical reason to yield a "soberly logical" relativist conclusion that moral duties must be relativized to the desires of the agent.

1. Moral requirements generate reasons for action.
2. Reasons for action are desire-dependent.
3. Hence, moral requirements are desire-dependent.

Another view that is both skeptical and minimalist is Callicles' position in Plato's Gorgias.  Callicles resolves the apparent conflict between the demands of virtue and the agent's own interest by arguing that "real" or natural justice does not require the agent to help others or forbear from harming them, as conventional morality supposes (482de, 483ab, 488b-490a).  The naturally just person satisfies her own unrestrained desires (488b).

1. Justice is a virtue.
2. Virtues must benefit their possessor.
3. Conventional justice benefits others, not the agent.
4. Hence, conventional justice is not real justice.
5. Hence, real justice consists in benefiting oneself, not others.

One attempt to meet this challenge is embodied in that strand of social contract theory -- including Epicurus, Hobbes, and Gauthier -- that understands the scope, content, and authority of morality in terms of rational agreement.  In its classical form, this view appeals to a prudential or egoist conception of rationality and argues that it is in the long-term interest of agents to develop, maintain, and act on other-regarding attitudes.  The general story is familiar enough and can be illustrated with an example from Hume.

Your corn is ripe today; mine will be so tomorrow. 'Tis profitable for us both, that I shou'd labour with you today, and that you shou'd aide me tomorrow.  I have no kindness for you, and know you have as little for me.  I will not, therefore, take any pains upon your account; and shou'd I labor upon my own account, in expectation of a return, I know I shou'd be disappointed, and that I shou'd in vain depend upon your gratitude.  Here then I leave you to labour alone; You treat me in the same manner.  The seasons change; and both of us lose our harvests for want of mutual confidence and security [Treatise III.ii.5/520-21].

For every rational individual, noncooperation strictly dominates cooperation; whatever the other does, each is better-off not cooperating.  Every rational individual would reason in this way, with the result that rational individuals cannot, in principle, secure mutual benefits.  (For the classical PD situation, substitute "stay silent" for "contribute" and "rat" for "don't contribute" and put the following values in boxes : NW box = A and B each serves 2 years in prison; SW box = A goes free and B serves 12 years; NE box = B goes free and A serves 12 years; and SE box = A and B each serves 10 years.  It's essential, of course, that prisoners consider the value only of jail terms -- for instance, they are not to consider the negative reputational effects of being a "stool pigeon" -- in other words, the pay-offs should really be in terms of utilities.)

But typically A and B have opportunities for ongoing mutually advantageous interaction with each other and others.  Here we have a so-called iterated PD situation, which changes the pay-off structure of the original game (so that we no longer have a genuine PD).  If A contributes, it is better for B to contribute, because the benefits of contribution and the costs of noncontribution are higher than in the single-shot PD; if B fails to contribute, A will cease to deal with him and/or retaliate and B will suffer bad reputational effects with others.

Insofar as moral interaction can be modeled as an iterated PD situation, the rational egoist can perhaps justify other-regarding moral behavior.  Much of impartial other-regarding morality involves norms of cooperation (e.g. fidelity and fair play), forbearance, and aid.  Each individual has an interest in the fruits of interaction conducted according to these norms.  Though it might be desirable to reap the benefits of others' compliance with norms of forbearance and cooperation without incurring the burdens of one's own, the opportunities to do this are infrequent.  Noncompliance is generally detectable, and others won't be forbearing and cooperative toward those who are known to be noncompliant.  For this reason, compliance is typically necessary to enjoy the benefits of others' continued compliance.  Moreover, because each has an interest in others' cooperation and restraint, communities will tend to reinforce compliant behavior and discourage noncompliant behavior.  If so, compliance is often necessary to avoid such social sanctions.  Whereas noncompliance secures short-term benefits that compliance does not, compliance typically secures greater long-term benefits than noncompliance.  The structure of this justification is something like this.

1. It is in an agent's interest to receive the benefits of norms of cooperation and restraint.
2. Others won't be cooperative and restrained toward those who are known to be noncompliant with these norms; the benefits of systems of cooperation and restraint are available only to those who appear to be cooperative and restrained.
3. Given the costs of detected noncompliance and the psychic costs (e.g. anxiety) that noncompliance will be detected, the least costly way to maintain the appearance of compliance is to be compliant.
4. Hence, it is in an agent's interest to be cooperative and restrained.
5. Hence, there are other-regarding duties that have prudential foundations.

• First, (1) is not always true.  The price of detected noncompliance is exclusion from the benefits of the system of cooperation and restraint.  But the penalty may not always outweigh the rewards of noncompliance.
• Second, (2) is not always true.  In the case of public goods that are nonexcludable, the appearance of compliance is not necessary to receive the benefits of compliance by others, though fairness typically seems to require compliance.
• Third, (3) is sometimes false.  Selective noncompliance may go undetected where it is difficult to monitor compliance and detect deception.  Where detection is unlikely, it would be unreasonable to be terribly anxious about detection.  In such cases, compliance would not be the least costly means of maintaining the appearance of compliance.
• Fourth, (4) and (5) may be too weak.  For even if the defense of other-regarding moral behavior were extensionally adequate, it would still be counterfactually fragile.  On this justification of compliance with other-regarding norms, compliance involves costs, as well as benefits; it must remain a second-best option, behind undetected noncompliance, in which one enjoys the benefits of others' compliance without the costs of one's own.  But then if one had some way of ensuring that one's own noncompliance would go undetected -- for instance, one had sole access to the ring of Gyges -- one could enjoy the benefits of others' compliance without the burdens of one's own, and one would have no reason to be compliant.  Whether this is a limitation in morality itself or in the rational authority of morality depends on how we understand the proposed link between morality and self-interest.  If we understand morality in act-egoist terms, then morality extends just as far as self-interest.  In those circumstances, whether actual or counterfactual, in which one can get the benefits of others' compliance without the costs of one's own, compliance is not morally required.  This is a revisionary moral claim; commonsense morality condemns the noncompliance whose rationality the ring of Gyges makes possible.  Alternatively, we might understand morality to consist in compliance with mutually advantageous norms of interaction, such as norms of cooperation and restraint.  On this view, morality demands compliance even in those circumstances, actual or counterfactual, in which one can get the benefits of others' compliance without the burdens of one's own.  In these circumstances, moral demands cease to be rationally authoritative.  This is a kind of anti-rationalist position.  What is true on both views is that the strategic justification of compliance with the sort of other-regarding norms contained in commonsense morality is counterfactually fragile.  Whereas the strategic justification of compliance must be unstable in this way, we, like Glaucon and Adeimantus, might demand that any justification of other-regarding morality be counterfactually stable.
• Fifth, this view must assume that the scope of other-regarding moral demands is limited to strategic contexts.  But such a view about the scope of morality is morally revisionary.  For instance, it's a common view that morality has wide scope; it imposes obligations of beneficence or aid where the agent stands to gain nothing from the cooperation or restraint of the beneficiary.  So, for instance, on this view (a) there can apparently be no obligations to future generations, and (b) if the wealthy and talented have sufficient strength and resources so as to stand to gain nothing by participating with the weak and handicapped in a system of mutual cooperation and forbearance, the former can have no duties of aid, however modest, toward the latter.  These are morally repugnant conclusions.

Foot illustrates one version of anti-rationalism in "Morality as a System of Hypothetical Imperatives."  She motivates this form of anti-rationalism by noting an ambiguity within Kantian rationalism.  Kant, of course, distinguishes between hypothetical and categorical imperatives (Groundwork 414).  Hypothetical imperatives are conditional on whether the conduct enjoined promotes the agent's antecedent interests or desires, whereas categorical imperatives are not.  Kant claims that moral requirements express categorical, rather than hypothetical, imperatives (416).  Following Foot, we might identify two distinguishable claims here.  In one sense, imperatives are categorical just in case they apply to people independently of their aims or interests.  If so, they express categorical norms.  The inescapability thesis claims that moral requirements express categorical norms.  In another sense, imperatives are categorical just in case they provide pro tanto reason for action, independently of the agent's contingent empirical aims and interests.  If so, they enjoy rational authority.  The authority thesis claims that moral requirements possess rational authority.  Though Kant clearly accepts both inescapability and authority theses, Foot argues -- by analogy with etiquette -- that only the inescapability thesis is plausible.

Foot is right that inescapability and authority are distinct claims.  But her denial of authority rests on a brute appeal to instrumental or prudential assumptions about  practical reason.

... [I]t is supposed [by Kant and others] that moral considerations necessarily give reasons for acting to any man.  The difficulty is, of course, to defend this proposition which is more often repeated than explained.  ...  The fact is that the man who rejects morality because he sees no reason to obey its rules can be convicted of villainy but not of inconsistency.  Nor will his action necessarily be irrational.  Irrational actions are those in which a man in some way defeats his own purposes, doing what is calculated to be disadvantageous or to frustrate his ends.  Immorality does not necessarily involve any such thing [Virtues and Vices, pp. 161-2].

But it would be nice to be able to show that those who don't have promoral sentiments have reason to acquire them and that those that do have reason to retain them.  This would apparently require rationalism.  The rationalist might question Foot's assumptions about practical reason, as the friends of impartial practical reason (e.g. Nagel) do.  Foot could appeal to Williams's defense of internal reasons.  But Nagel thinks that motivation is consequential on judgments of practical reason, so that practical reason need not be constrained by what the agent antecedently desires.