Derek Parfit is a philosopher who specializes in problems of personal identity, rationality and ethics, and the relations between them. His 1984 book, Reasons and Persons (described by Alan Ryan of The Sunday Times as "something close to a work of genius") has been very influential in the field.
Ethics and Rationality
Reasons and Persons is a four part work each subtly building on the last. Parfit believes that non-Religious ethics is a young and fertile field of inquiry. In many ways he is somewhat of a foil to Wittgenstein turning his study inwards towards the everyday mechanics of moral problems, actions that are right or wrong, and away from meta-ethics which focuses more on logic and language. In Part I of Reasons and Persons Parfit discusses "self-defeating theories" -- namely the self interest theory on rationality (S)and two ethical frameworks: common sense morality (CSM) and consequentialism (C). He posited that the self-interest theory has been dominant in Western culture for over two millennia, often making bedfellows with religious doctrine, which united self interest and morality. The self interest theory demands that we always make self interest our supreme rational concern and consequently we pursue it over whole lives, that is in a temporally neutral manner. It would be irrational to act in ways that we know we would prefer later to undo. It is irrational for my 17 year old self to listen to punk rock and get arrested for protesting an oil spill if I know that I have aspirations to be a chemical engineer and my actions would assuredly detract significantly from my future well-being. Most notably, the self interest theory holds that it is irrational to make any acts of self denial or act on desires that negatively affect our well-being. Consider an aspiring author who's strongest desire is to write an award-winning novel yet in doing so she suffers greatly due to lack of sleep and depression. Parfit holds that it is plausible that we have desires other than our own well-being and that it is not irrational to act to fufill these desires.
Aside from the initial appeal to plausibility Parfit contrives situatitions where S is indirectly self-defeating. That is, it makes demands that it initially posits and irrational. It does not fail on its own terms, but it does recommend adoption of an alternative framework of rationality. Such a feature is clearly undesirable thus we need to search for a new theory of rationality. Parfit offers the Critical Present Aim Theory (CP), a broad catch-all that can be formulated to accommodate and competing theory. Parfit contructs CP to exclude self interest as our over riding rational concern and to allow the time of action to become critically important. He leaves the question open however, if it should include as our highest concern, "to avoid acting wrongly." Such an inclusion would pave the way for ethics.
S is not the only self-defeating theory however. Where S puts too much emphasis on the seperateness of persons, C fails to recognize the importance of bonds and emotional responses that comes from allowing some people privaledged positions in one's life. If we were all pure do-gooders, perhaps following Sidgwick, then it would not constitute the outcome that would maximize happiness. It would be better if some of us, a small percentage of the population were pure do-gooders, but that others acted out of love, etc. Thus C too makes demands of agents that it initially deemed immoral; it fails not on its own terms, for it still demands the outcome that maximizes total happiness, but does demand that each agent not always act as impartial happiness promoters. C thus needs to be revised as well.
S and C fail indirectly, while CSM is directly collectively self-defeating. Parfit shows, using interesting examples and borrowing from Nashian games, that it would often be better for us all if we did not put the welfare of our loved ones before all else. For example, we should care not only about our kids, but everyone's kids. Parfit often poses more questions than he answers. In ethics, he points to a need for a dynamic framework that combines CSM and C but does not offer any specific solution. Such an attitude tracks his stance that non-Religious ethics is a young, fertile field.
Parfit uses many science fiction inspired examples to suss out our emotions and feelings concerning self identity. He is a Reductionist; he believes that self identity can be reduced to a set of criterion that need to suppose that people do exist. Identity can be fully descibed impersonally. To the question "Am I about to die, will the person that remains be me" there need to be a determinate answer. He concludes that we are mistaken in assuming personal identity is what matters, what matters, he posits is Relation R, psychological connectedness (of namely memory and character) and continuity (overlapping chains of strong connectedness).
We are nothing more than our brains and our bodies but identity cannot be reduced to either, for identity, in the classical sense is not what matters. Relation R is what matters. Parfit concedes that his and rival Reductionist theories rarely conflict in everyday life, and are only brought to blows by the introduction of out of this world examples, but defends the usage in that they seem to arrouse genuine and strong feelings in many of us. Identity is not as determinate as often suppose it is, such determinancy owes itself mainly to the way we talk. People exist in the same way that nations or clubs exist.
For more on this issue, search "Parfit Identity" on Google for a number of informative articles.
Parfit's most famous postulations come in Part IV of Reasons and Persons ' where he discusses possible futures for the world. He shows that, in the discussion of possible futures, both average and total utilitarian standards lead to unwelcome conclusions. Applying total utitilitarian standards (absolute total happiness) to possible growth paths of population and welfare leads one to the Repugnant Conclusion. Parfit illustrates this with a simple thought experiment. Imagine a choice between possible futures, in A 10 billion people would live during the next generation all having extremely happy lives, lives far happier than anyone lives today. In B, there are 20 billion people all living lives, while slightly less happy than those in A, are still very happy. Under total utility maximization we would prefer B to A, and through a regressive process of population increases and happiness decreases (in each the happiness decrease is more than outweighed by the population increase) we are forced to prefer Z, a world of 100's of billion people all living lives barely worth living, over A. If we do not assume that causing to exist can benefit someone then we must at least admit that Z is no worse than A.
The absurd conclusion takes a similar form. If all we care about is average happiness, we would be forced to conclude that an extremely small population, say 10 people, over the course of human history is the best outcome if we assume that these first 10 people (Adam and Eve et al) had lives happier than we could ever imagine. Consider the case of American immigration. Presumably alien welfare is less than American, but the would-be alien benefits tremendously from moving from his homeland. Assume also that Americans benefit from immigration (at least in small doses) because they get cheap labor, etc. Under immigration both groups are better off, but the average welfare is lower. Thus although, everyone is better off, this is not the preferred outcome. Parfit asserts that this is simply absurd.
Parfit explorerd the question about identity in relation to the time of conception. I would not be me if my parents waited 2 more years to have a kids. While they would still have a child it would certainly be someonen else, no me. Because of this fact, identity is dependent on the time of conception and the accompanying details. The initial conditions of a system can have drastic effects on the resulting people that exist after a seemingly short ammouunt of time. A drastic environmental policy shift would shift the initial conditions of the conception process so much that after 60 years none of the same people are born. Different couples meet each other and concieve at different times.
We could therefore craft disasterous policies that would be worse for nobody, becase none of the same people exist under the different policies. If we consider the moral ramifications of potential policies in person affecting terms, then we will have no reason to prefer a sound policy over an unsound one. This is the non-identity crisis: the identity of future generations is causually dependent, in a very sensitive way, on the actions the present generations. So much so that they effectively have no identity if one looks as little as a half century into the future.