It has often been observed that pure hedonism is self-defeating because a person whose ultimate desires are desires for pleasure will not, in any usual case, get as much out of life as someone with a variety of interests who genuinely cares for other people. The present point is that a variety of interests and a concern for others can be derived from an initial desire for pleasure—not because these concerns and interests have pleasure as their ultimate object (in the sense that they are ultimately concerns for and interests in one's own pleasure, which they are not), but because having those interests and that concern can be a means to one's own pleasure. This is the sort of thing that the child learns as it grows from a selfish hedonistic infant to a loving adult with a concern for society and interests in the arts. Psychological hedonism may be right in seeing the ultimate source of adult desires in the primitive desires of the infant but is wrong in taking this to show that adult intrinsic desires are ultimately desires for the same sort of things that the infant intrinsically desires.
One can have a selfish reason to develop a genuinely unselfish concern for others. For, if one has an unselfish concern for others, it is likely that the others will reciprocate. Furthermore, and more importantly, one's life will be richer and more interesting. But, given such a concern, one may also acquire, for example, an unselfish desire about the distribution of one's property at one's death, which can lead one to write a will in order to bring about that distribution, even though one will not be around at that point to be affected by that distribution.
Similarly, it is often useful, for reasons of efficiency, to develop an intrinsic interest in something in which one already has an instrumental interest, for example, money or physical fitness. When such intrinsic interests have developed, they may lead to actions that go beyond what would have been needed to satisfy one's original purposes. One may come to want to amass a large fortune, more than anyone would "need," or one may wish to retain the sort of physical fitness only an athlete ordinarily has. An analogous point helps to explain why one may hope to be remembered after death. It is often in one's interest that others should think of one and not forget one; fame has its uses. It is, therefore, useful to acquire an intrinsic interest in being remembered and thought well of. Having acquired that interest, however, one may well want to be remembered after one has died, even though being remembered at that point will not be useful in terms of one's original purposes.
(Gilbert Harman, "Practical Reasoning," Review of Metaphysics 29 [March 1976]: 431-63, at 460)
Note from KBJ: Ethical egoism, like utilitarianism, is a normative ethical theory. It says that "An act is morally right if and only if no alternative to that act has higher agent utility than it has" (Fred Feldman, Introductory Ethics [Upper Saddle River, NJ: Prentice-Hall, 1978], 82). Philosophers take pride in their charitableness, but I have yet to see a philosopher who is charitable toward ethical egoism. Some philosophers, e.g., James Rachels and Fred Feldman, say that it implies that one should walk all over others. Since when is it in one's interest to walk all over others? If you get a reputation as a liar, a cheat, a thief, a brute, or a scoundrel, you'll be ostracized, and being ostracized is hardly in one's interest. As Harman shows in the passage quoted, a selfish (better: self-interested) person has reason to take the interests of others into consideration prior to acting. I can't make the case here, but I believe, with Hobbes, that ethical egoism provides strong support for the rules of conventional morality.