Nagel both (a) rejects the need for pro-attitudes in moral motivation (including the Humean view about motivation) and defends a purely cognitive account of motivation, and (b) rejects agent-centered assumptions about practical reason (that it is instrumental or prudential) (including the Humean view about rationality) and defends impartial practical reason. Nagel may see (a) and (b) connected, if he assumes that practical reason must motivate. If so, he may see (a) as necessary to defend his main claim (b).
ALTRUISM/IMPARTIALITY
Altruism implies nonderivative reason to be concerned about and help
others (PA: 15-16). Is altruism the same as agent-neutrality?
Reasons are agent-relative iff their general form involves essential reference to the agent who has them; otherwise, they are agent-neutral (The View from Nowhere, pp. 152-53).Agent-neutrality is usually thought to imply a consequentialist normative doctrine in which if one person has a reason to do or care about something (fear or relieve his own pain) then anyone else has a reason do or care about that thing (e.g. to care about or relieve his pain). By contrast, there are many different possible forms of agent-relativity. Instrumentalism and prudence or egoism are both agent-relative. But so are deontological normative doctrines that recognize side-constraints on promoting the good and C.D. Broad's self-referential altruism, which claims that one has non-derivative reason to benefit anyone it is in one's power to benefit but that the weight of one's reasons to benefit are a function of the relationship between benefactor and beneficiary ("Self and Others"). Nagel himself appears to reject an agent-neutral conception of altruism insofar as he expresses doubts about whether consequentialist normative conceptions can recognize the separateness of persons and the prohibition on demanding uncompenstaed sacrifices (138-42). While some agent-relative conceptions, such as instrumentalism and prudence, deny the existence of nonderivative reason to be concerned about or help others, others, such as self-referential altruism, embrace the existence of nonderivative reason to be concerned about and help others.
Would Nagel regard self-referential altruism as an acceptable conception of altruism? While self-referential altruism does recognize nonderivative reason to be concerned about others, it may not recognize the sort of equal normative concern that Nagel appears to endorse (138-42). His challenge is to find an interpretation of equal concern that does not require the agent-neutrality, which he rejects.
In fact, Nagel is concerned with the necessity, and not just the possibility, of altruism (96; cf. 88, 90, 98, 145-46).
INTERTEMPORAL AND INTERPERSONAL DISTRIBUTION
In contrast with instrumentalism, prudence claims that reasons extend
across time; in contrast with prudence, altruism claims that reasons extend
across persons. Just as the agent's future self provide him with
reasons now, so too the interests of others provide him with reasons for
actions. Failure to recognize prudence involves temporal dissociation
(failure to recognize the present as one time among others); failure to
recognize altruism involves personal dissociation (failure to recognize
oneself as one among others).
One concern about this defense of altruism is that it seems to appeal to the parity of intertemporal and interpersonal distribution -- interpersonal distribution should privilege the agent's own interests no more than intertemporal distribution should privilege the agent's current desires or interests. But Nagel appears to defend temporal neutrality. Parity would require interpersonal neutrality or agent neutrality, which Nagel resists as an interpretation of altruism.
In any case, intertemporal and interpersonal distribution are importantly disanalogous if, as Nagel appears to believe, there must be compensation for sacrifice (142). Intrapersonal balancing is automatically compensated, because benefactor and beneficiary are the same. But interpersonal balancing is not automatically compensated, because benefactor and beneficiary are distinct.
CONSTANCY OF MOTIVATIONAL CONTENT ACROSS PERSPECTIVES
Nagel believes that motivational content (= authority) must be constant
across both temporal and personal perspectives.
Nagel claims that interpersonal univocity of motivational content involves universalizability (PA: 107). But altruism's rival, egoism also respects universality. Does interpersonal univocity add anything (109)?
RESENTMENT
Nagel rests his intuitive case for altruism on the existence of emotional
reactions such as resentment. If you fail to tell me that there is
a wasp on my hamburger or refuse to walk around my gouty toes, I am likely
to respond to you with resentment (16n, 85). Or consider another
case. I've been building an elaborate and fragile model in my backyard.
(a) If a sudden gust of wind destroys my model, I am likely to be disappointed,
angry, or frustrated, but I won't be resentful. There is no one to
resent. (b) But I won't be resentful if you inadvertently destroy
my model, provided that I don't think that your behavior was negligent.
If your destructive behavior is the result of Tourette's Syndrome, it would
again be appropriate me to feel disappointment, anger, or frustration,
but not resentment. (c) But if you willfully destroy my model, because
you are unwilling to suffer any delay whatsoever on your way to the mailbox,
then it seems appropriate for me to feel resentment. Presumably,
what distinguishes resentment from disappointment, anger, and frustration
is that it implies, as they do not, that one has been wronged and that
someone else has acted as he ought not to have. Moreover, resentment
is not predicated on the belief that the offending agent had a self-interested
reason not to engage in the destructive behavior. But then it looks
like feelings of resentment are evidence against egoism -- the idea that
all reasons for action are prudential -- and for altruism.
This argument seems to present a problem of normative accommodation for the egoist that is independent of Nagel's official line of argument that exploits parallels between the justification of prudence and of altruism and that appeals to considerations of universalizability or interpersonal univocity. But Nagel might deny this appearance of independence. He might claim that the egoist who experiences resentment is making judgments that are not interpersonally stable. My feeling of resentment implies the judgment that you (or others) have reason to refrain from causing me unnecessary pain, independently of your own interests. But this requires me to judge that anyone similarly situated (including me) has reason to refrain from causing unecessary pain to others, independently of his own (my own) interests.
What this shows is that the egoist cannot consistently harbor feelings of resentment toward the inconsiderate. Of course, he has self-interested reason to object to such behavior both prospectively and retrospectively, and he may have self-interested reason to act as if they had necessarily flouted reasons to behave otherwise. But he cannot actually resent them inasmuch as this would imply that they did have and flout reasons for behaving otherwise. But we can avoid inconsistency by abandoning either egoism or feelings of resentment toward the inconsiderate. Abandoning reactive attitudes such as resentment may seem like a steep price to pay for a philosophical commitment, revealing the unsustainable cost of the egoist living his egoism. Perhaps so. But Nagel also purports to show that egoism is an unstable normative position and that the egoist must recognize the demands of altruism. But the egoist can resist the demands of altruism if he is willing to abandon his attitudes of resentment. The egoist need not be inconsistent, but are willing to pay the price of being egoists?