Rarely does a long and densely argued philosophical work come to the attention of the general reader. Reviews of such books are generally restricted to academic journals; it is assumed that they will be of interest only to specialists. Occasionally, however, wider claims are made for a work of philosophy, and such has been the case with Derek Parfit’s Reasons and Persons. Writing in The New York Review of Books, the distinguished British philosopher P. F. Strawson praises Parfit’s book not only for its “intellectual illumination and delight” but also for its potential practical effect; Reasons and Persons, he suggests, “may point the way to the emergence of a satisfactory theory of rational beneficence, and this might, in the long run be capable of influencing political behavior.” Samuel Scheffler, in the Times Literary Supplement, writes that “Reasons and Persons may be the greatest work of substantive moral philosophy in the utilitarian tradition” since Henry Sidgwick’s classic study The Methods of Ethics (1874). These are large claims indeed, and they demand a close reading of Parfit’s book.
Parfit describes himself as one who by temperament is a revisionist: “Philosophers should not only interpret our beliefs; when they are false, they should change them.” The fundamental ideas that Parfit thinks we ought to abandon are “beliefs about our own nature, and our identity over time.” In short, Parfit contends that we have no good reasons to believe that we exist throughout our lives as the same person.
Unlike many revisionists, however, Parfit does not write merely to shock. Instead, he writes as a champion of that venerable moral principle—impartial benevolence. His argument, although somewhat complex, is ultimately straightforward: Impartial benevolence is the most rational principle to guide one’s life choices and decisions, because the only plausible alternative—the principle of enlightened self-interest—falsely assumes that an individual remains the same person throughout his life. Without this assumption, self-interest theorists would have to recommend a life of unrestrained sensual gratification. All but the crassest of Philistines, however, realize that while an evening of wine, women, and song may be enjoyable, the costs paid on the day after almost certainly cancel out the benefits of the previous night’s pleasures, and even if there were some cases of pleasure without payment, there can be no doubt that an entire life so spent would not be the most conducive to an individual’s own happiness.
Suppose, however, that there is only a tenuous or nonexistent connection between the intemperate adolescent and the prematurely old man dying of kidney failure. If this is so, the principle of enlightened self-interest loses its appeal. Part 3 of Reasons and Persons argues that there are only these tenuous or nonexistent connections between individuals over time.
Does this mean that if we are presently in the full flower of youth, we simply ought to abandon any concern we might have for that prematurely old man? Parfit is sure that we should not. “Some outcomes are good or bad, in a sense that has moral relevance: it is bad for example if people become paralyzed, and we ought, if we can, to prevent this.” Thus, it is wrong for anyone to act in a way which leads to an early death, even if it is his own. What Parfit believes his revisionist theory of personal identity shows is that we have no better reason to be concerned with our own future well-being than we have to be concerned with the future or present well-being of someone else. Thus, the long-range partiality to self that is recommended by enlightened self-interest is impossible to implement because there is no enduring self to which a person can be partial.
This conclusion is one which almost all moralists can welcome. Moreover, Parfit carefully makes room for the entitlements and individual rights upon which many contemporary critics of utilitarianism insist. Nevertheless, many philosophers would say that this is not the book that solves what Sidgwick regarded “as the profoundest problem of Ethics”—the conflict between self-interest and morality. These philosophers would argue that Parfit’s denial of an enduring self pays too great a price, especially because there are less costly ways of solving the problem.
Aristotle, Bishop Joseph Butler, and those in the twentieth century who refer to themselves as “descriptivists” are among those who think that there is a less costly way. Instead of appealing to the “metaphysical fact” of a nonexistent self, these philosophers appeal to less esoteric facts which can be discovered through shrewd observation and inculcated by good sermons. As Butler wrote and preached, “Surely that character we call selfish is not the most promising for happiness,” and again, “Surely the man of benevolence hath as great enjoyment as the man of ambition.” Butler and others argue that impartial benevolence is reasonable because in the long run it coincides with a person’s own happiness. Partial benevolence is impossible because it is self-defeating. If a person asks, Why should I not devote myself to a life of enlightened self-interest? Peter Geach observes that an “obviously relevant sort of replyis an appeal to something the questioner wants, and cannot get if he does so-and-so.” Furthermore, Geach adds, “only such a reply is relevant and rational.”
Parfit would disagree, and his objections are found in part 1 of his book. His argument focuses on what has come to be known as “the Prisoner’s Dilemma.” Two thieves are caught, and while the prosecutor has sufficient evidence to win a two-year conviction, it is impossible to win the stiffer ten-year sentence which they both deserve without the testimony of one of the thieves. The prosecutor therefore questions them separately and makes the following offer to both: If you confess and give evidence for the state while your partner remains silent, you will go free and he will receive a twelve-year sentence. If you both confess, you will receive a ten-year sentence.
One might at first think that the prosecutor’s offer of freedom contingent upon the other’s silence too generous an offer. The prosecutor, however, is convinced that all thieves operate under the principle of enlightened self-interest and is therefore sure that neither will go free. Suppose either thief is a gambler and opts for freedom. If so, he will confess and hope his partner remains silent. On the other hand, if one opts to play it safe and minimize his loss, no matter what the other partner does, he will confess. The dilemma arises when we remember that if there were honor among thieves, the prosecutor’s case would virtually collapse, and both would only receive a two-year sentence.
In such cases, Parfit argues, we see the fundamental weakness of self-interest theories and the need for impartial benevolence. Furthermore, such cases make it clear that enlightened self-interest and impartial benevolence do not coincide. If one thief is an egoist and the other honorable, the egoist wins big and the honorable man gets the fool’s reward. Thus, Parfit writes, “Prisoner’s Dilemmas need to be explained. So do their moral solutions. Both have been too little understood.”
One wonders, however, how many times individuals actually find themselves in a prisoner’s dilemma where there can be no communication and hence no reciprocity. The possibility of reciprocity is crucial for those who maintain that self-interest and benevolence coincide; certainly, that calculating secularist Benjamin Franklin would not have recommended honesty as the best policy in a world where reciprocity was impossible. “Outside prisons, or the offices of game-theorists,” Parfit admits, this crucial condition is rarely met. Parfit admits that even the much-discussed arms race between the United States and the Soviet Union is not a true Prisoner’s Dilemma because “the choice made by each may affect the later choices made by the other.”
Why, then, even discuss such artificial problems? Parfit writes, “Though we can seldom know that we face a Two-Person Prisoner’s Dilemma, we can very often know that we face Many-Person Versions. And these have great practical importance. The rare Two-Person Case is important only as a model for the Many-Person Versions.” For example, in cases where “wages depend on profits, and work is unpleasant or a burden,” says Parfit, “it can be better for each if others work harder, worse for each if he himself does.” Another example involves soldiers: “Each will be safer if he turns and runs, but if all do more will be killed than if none do.” A third example involves fishermen: “When the sea is overfished, it can be better for each if he tries to catch more, worse for each if all do.” In all of these cases, we “need moral solutions. We must be directly disposed to make the altruistic choice.” As Parfit asserts, these solutions will often involve self-denial where each person does for moral reasons what he knows will be worse for him. Thus, self-interest and morality do not always coincide.
Why is this? Cannot union workers, soldiers, and fishermen communicate with one another? Is there no possibility of reciprocity in such situations? Is there no camaraderie in these professions? It may be true that in these professions, if one does what is morally right, while others do not, he is at a distinct disadvantage. If, however, we suppose that a person acting selfishly may influence others to act similarly, it may also be the case that someone behaving unselfishly may have a beneficial effect on the actions of others. What we have is something similar to William James’s “faith which creates the fact.”
Suppose a person meets someone with whom he would like to be friends. The reasonable thing to do is to make some sort of friendly gesture. Of course, one may be spurned, and one’s feelings hurt, but if one succeeds in making friends, one has demonstrated how kindness and self-interest can coincide. An exceptionally cautious person might wait for the other person to make the first move. This would indeed eliminate some risk of being hurt, but such a “safe” course of action may prevent...
(The entire section is 4260 words.)