This thesis proposes a reading of David Gauthiers moral contractarianism (hereinafter Mb(CM)A) that demonstrates how cooperation can be rational in situations where expected utilities (EU) are stacked too high against cooperation. The dissertation critically examines Mb(CM)A and contends that it breaks down in the test of application, i.e. the problem of secession because of the conception of rationality it appeals to. Mb(CM)A identifies rationality with utility-maximization, where utility is the measure of considered coherent preferences about outcomes. Mb(CM)A links morality to reason, and reason to practical rationality, and practical rationality to interest, which it identifies with individual utility. On this view, an action (or a disposition) is rational if that action (or disposition) maximizes an agents EU. This conception of rationality the essay claims is both nave and misleading because it does not take into account an agents considered preference for the acts that are available, in addition to the EU of those acts. Therefore, the thesis argues that Mb(CM)As account of rationality be abandoned in favor of a decision-value/symbolic utilitys or morals by decision-value agreements conception of practical rationality. Morals by decision-value agreement (henceforth Mb(DV)A), the dissertation claims, handles serious problems, like the problem of secession in ways that Mb(CM)A cannot. Mb(CM)A breaks down in the test of application because when applied to the problem of secession, it suggests a single-tracked silver bullet solution. Specifically, it tracks only EU-reasons and claims that insofar as cooperation does not maximize the EU of better-off agents, it is not rational for them to cooperate with or support those that are less well-off. By contrast, Mb(DV)A offers a multi-tracked framework for solutions to the problem, namely: it factors in an agents considered preference for the acts that are available, in addition to EU of those acts. It is the argument of the thesis that when EU is stacked too high against cooperation, it may or may not be rational for an agent to cooperate, depending on which way symbolic utility (SU) for that agent points toward. If SU points in the direction of secession, then it is DV-rational for an agent not to cooperate, but if SU points toward non-secession, then it is DV-rational for that agent to cooperate.

本論文題旨為：「論代間正義：一個羅爾斯式的觀點」。代間正義是晚近三十年來新興且益愈受到重視的倫理學議題之一，其主要關切的核心問題，乃在於追問「當代之於後代所應擔負的責任」。本文透過當代政治哲學與倫理學家羅爾斯（John Rawls）有關社會正義理論的設計與主張，分兩部分處理此問題。 / 第一部份所處理的問題為由帕菲特（Derek Parfit）所深化之「後代人格不同一問題」（The Non-Identity Problem），旨在探究「代間正義是否可能」。帕菲特指出，前代不同的行為選擇，將造成不同後代的存在，而這使得現存既有之各種權利與責任相對應的理論，無法合理地適用於代間關係。帕菲特主張，可以採取「忽略特定人格的比較（品質）原則」以解決此後代人格不同一問題。然而，帕菲特此舉卻陷入「不特定人格的後代無法追究前代之行為責任」的理論困境。本文主張，透過羅爾斯原初位置（original position）的理論設計啟發，即便在代間存在著「前代不同的行為選擇，將造成不同後代的存在」的後代人格不同一疑慮，當代仍可採納原初位置的理論啟示，區分人的屬性（properties）與獨特性（particular）的差異，在後代存有人格不同一問題（獨特性）的情形下，針對「屬性」而確立追問當代之於後代所應擔負責任之正當性。 / 第二部分主要處理羅爾斯有關代間正義觀點的內部論證問題，並進一步藉此說明「代間正義如何可能」。羅爾斯以「正義的儲蓄原則」（just savings principle）說明代間的分配正義問題，並《正義論》（A Theory of Justice）中將它納入正義二原則之中，成為建構社會基本體制的基本原則之一。然而，羅爾斯早期解釋與證成儲蓄原則的相關理據如動機假定（motivation assumption）與家族模式等，卻可能與其他理論假定如締約者的理性等相衝突、衍生解釋融貫上的困難，而遭到眾多學者們的質疑。羅爾斯在一九九三年的《政治自由主義》（Political Liberalism）中對此做出回應，並將關切下一代的動機假定，修正為「要求前代也承諾遵守他們所遵守的儲蓄原則，無論向前或向後追溯多遠」。除此之外，羅爾斯於《正義論》以外的其他著作，在提及正義二原則時，皆不再表述「正義的儲蓄原則」。本文認為羅爾斯後期所提出的證立主張，不僅整合了代內分配正義（差異原則）與代間分配正義（儲蓄原則）的論證理據（小中取大規則的應用），更與其主張「社會作為一世代相繼之公平的合作體系」時所標舉之「相互性理念」（the idea of reciprocity）的核心概念相符應。本文認為，在論證理據得到順利整合的前提下，羅爾斯仍可在後期表述正義二原則時，將「正義的儲蓄原則」放回其中，並可據此呼應當代永續發展理念「既滿足當代人的需求，又不對後代人滿足其需求的能力造成危害」的核心主張。 / The topic “Intergenerational Justice” is one of the newest but getting more important ethics problems to contemporaries. One of the key points of this issue is how to make sense of our obligations to the posterity (include future people) if possible. In this dissertation, I intend to clarify this problem by Rawls’s theory of justice into two parts. / The first part is “The Non-Identity Problem” held by Derek Parfit. This problem shows that “in the different outcomes, different people would be born”, and it seems inactive the traditional theories of rights. Parfit suggests that we can through it by the priinciple Q: “if in either of two outcomes the same number of people would ever live, it would be bad if those who live are worse off, or have a lower quality of life, than those who would have lived.” But this principle makes new difficulties about this problem. According Reiman, I argue that we can adjust this non-identity problem by the theory hypothesis “Original Position” of Rawls’s theory of justice, and that there are obligations from contemporaries to the future people. / The second part is about the arguments of Rawls’s theory of justice between generations. According to early Rawls in 1971, the theory of justice between generations represented by the “just savings principle” and was one part of the Two Principles of Justice in A Theory of Justice. But there are some argument troubles about the assumptions that makes the theory of justice between generations difficultly, especially on the “motivation assumption” and the family mode and so on. In 1993, Rawls changed his arguments about the theory of justice between generations, but he also take off the just savings principle from the Two Principles of Justice in other books or articles besides A Theory of Justice. I argue that latter arguments seem more reasonable to the theory of justice between generations, and they also makes the whole theory of social justice comprehensive. Then I argue that Rawls can still presents the just savings principle when he says about the Two Principles of Justice.