Rawls’s Triple Critique of Utilitarianism

More than fifty years have passed, and Rawls’s critique of utilitarianism remains an important theoretical backdrop for those engaged in the study of utilitarianism. One of the primary goals of Rawls’s construction of “justice as fairness” was to replace the prevailing utilitarianism. As Rawls himself states in the preface to *A Theory of Justice*: “What I have attempted to do is to generalize and carry to a higher order of abstraction the traditional theory of the social contract as represented by Locke, Rousseau, and Kant. In this way I hope to develop a theory of justice that can withstand those objections which have often seemed fatal to it. Moreover, this theory seems to offer a systematic account of justice that I believe is superior to, or at least I shall argue is superior to, the dominant utilitarian tradition.” Freeman also points out that one of Rawls’s “primary aims is to articulate the most appropriate moral conception of justice for a democratic society, a conception that, in contrast to the prevailing utilitarian tradition, would better account for the democratic values of freedom and equality.” In *A Theory of Justice*, Rawls creatively divides utilitarianism into three distinct forms—classical utilitarianism, average utilitarianism, and limited utilitarianism—and contrasts them with the two principles of justice, thereby completing a comprehensive critique of utilitarianism. However, the theoretical force of Rawls’s critique weakens progressively across these three forms, which conversely indicates that utilitarianism exerts a persistent and difficult-to-resolve theoretical pressure on “justice as fairness.”

50多年过去了，罗尔斯对功利主义的批评仍是人们从事功利主义研究的重要理论背景。罗尔斯建构的“公平的正义”的首要目标之一就是替代盛行的功利主义。正如罗尔斯在《正义论》序言中所说：“我试图做的就是要进一步概括洛克、卢梭和康德所代表的传统的社会契约理论，使之上升到一种更高的抽象水平。借此，我希望能把这种理论发展得能经受住那些常常被认为对它是致命的明显攻击。而且，这一理论看来提供了一种对正义的系统解释，这种解释在我看来不仅可以替换，而且还优于（或至少我将如此论证）占支配地位的传统的功利主义解释。”弗雷曼也指出，罗尔斯的“首要目标之一在于，为民主社会阐明最恰当的正义道德观，与盛行的功利主义传统相比，那种道德观将更好地解释自由平等的民主价值”。在《正义论》中，罗尔斯创造性地将功利主义分为古典功利主义、平均功利主义和有限功利主义三种不同的表现形式，并与两个正义原则相对照，从而完成对功利主义的全面批判。然而，罗尔斯对功利主义三种不同的表现形式给予的理论批判力道越来越弱，这反向说明功利主义对“公平的正义”形成了难以消解的理论压力。

I. Critique of Classical Utilitarianism

一、古典功利主义批判

There are many theories of justice; Rawls calls his own “justice as fairness.” From his first important essay on justice in 1958, “Justice as Fairness,” to his final work, *Justice as Fairness: A Restatement*, Rawls’s use of the concept “justice as fairness” never changed. For Rawls, fairness is a “marker term” within his theory of justice. Although Rawls does not provide an explicit definition of “what fairness is,” he offers clear discussions in various chapters on how to guarantee the initial conditions of fairness and how to embody the idea of fairness. Among these, the theoretical contrast between “justice as fairness” and classical utilitarianism is an important way to reveal the meaning of fairness. It is precisely through this theoretical comparison of the two conceptions of justice that “justice as fairness” comes to replace utilitarianism.

正义理论存在很多种，罗尔斯将自己的正义理论称之为“公平的正义”。从1958年关于正义问题的第一篇重要文章“作为公平的正义”起，一直到生前最后的著作《作为公平的正义：正义新论》止，罗尔斯对“公平的正义”概念的使用就没有变过。就正义理论来说，公平对于罗尔斯是一个“标识词”。罗尔斯虽然没有对“什么是公平”给予明确的界定，但是，就如何保证公平的初始条件和体现公平的理念，罗尔斯却在不同的章节里有明确的论述。其中，“公平的正义”同古典功利主义的理论对照是公平含义展现的重要方式。正是通过对两种正义观的理论比较，“公平的正义”才会取代功利主义。

The classical utilitarianism Rawls refers to is primarily represented by Hume, Smith, Bentham, Mill, and Sidgwick. The basic thesis of classical utilitarianism is: “if the main institutions of a society are arranged so as to achieve the greatest net balance of satisfaction summed over all the individuals belonging to it, then the society is properly organized and therefore just.” Utilitarianism is simple, clear, and intuitively appealing when explaining individual behavior. When individuals choose actions or calculate their own interests, they typically measure losses against gains. In daily life, people also choose varying degrees of self-sacrifice in exchange for larger future benefits. As Nozick criticizes utilitarianism: “Individually, we each sometimes choose to undergo some pain or sacrifice for a greater benefit or to avoid a greater harm: we go to the dentist to avoid worse suffering later; we do unpleasant work for its results; some people diet to improve their health or appearance; some save money for their old age. In these cases, some costs are borne for a greater overall good.” For most people, individual behavior based on utilitarian calculation seems unobjectionable. The key question is whether principles applicable to individual conduct can be applied to society. It is at this point that utilitarianism and “justice as fairness” diverge.

罗尔斯所说的古典功利主义主要以休谟、斯密、边沁、密尔和西季维克等为代表。古典功利主义的基本观点是：“如果一个社会的主要制度被安排得能够达到所有社会成员满足总量的最大净余额，那么这个社会就是被正确地组织的，因而也是正义的。”功利主义在解释个人行为的时候简单、清晰并且符合人们的直觉。当个人在抉择行动或计算自己利益的时候，通常都会根据自己的所得来衡量所失。在日常生活中，人们也会为换取将来较大的利益而选择不同程度的自我牺牲。就像诺奇克在批评功利主义时所说：“就个人而言，我们每一个人有时愿意为了更大的利益或避免更大的伤害而经受某些痛苦或牺牲：我们去看牙医以避免以后遭更大的罪；我们为了其结果而做某种令人不快的工作；某些人节食以改善他们的健康或外貌；某些人储蓄金钱以供他们到老年时用。在这些场合，为了更大的整体利益，一些代价被承受了。”对于大多数人来说，以功利计算为基础的个人行为没有不妥之处。问题关键是，适用于个人行为的原则能不能运用到社会之中，功利主义和“公平的正义”在此分道扬镳。

Classical utilitarianism holds that the principle for an individual can be extended to society. An individual can measure future gains and losses against present ones, and since society is composed of individuals, society too can trade off among individuals for a greater overall good. If the individual’s principle is to advance his own welfare and satisfy his own system of desires as much as possible, then society’s principle is likewise to advance the welfare of the group as much as possible and to realize the total system of desires of all its members. Through this analogy between the individual and society, utilitarianism naturally arrives at the conviction that the principle of choice for a human society is simply an extension of the principle of choice for an individual.

古典功利主义认为可以将个人原则运用到社会原则中。一个人可以根据现在的所得所失来衡量未来的所得所失，而社会是由个人构成的，为了更大的整体利益，社会也可以在个人之间进行权衡。既然个人的原则是尽可能地推进自己的福利，满足自己的欲望体系，那么，社会的原则也是要尽可能地推进群体的福利，最大程度地实现包括它的所有成员在内的总的欲望体系。通过个人与社会的类比，功利主义自然就达到了这样的信念：一个人类社会的选择原则就被解释为个人选择原则的扩大。

The secret of utilitarian reasoning lies in its natural incorporation of individual desires into the total system of social desires. Only by understanding utilitarian reasoning can one discern its fallacy. Utilitarianism sees no difference between the principle for an individual and the principle for society. In Rawls’s view, this follows from the theoretical aims of utilitarianism. Generally speaking, utilitarianism is not concerned with how the total sum of desires is distributed, but with how that total sum can be increased. To determine how society can maximize the net balance of desires, utilitarianism can only reason from the total sum of individual desires. That is, the aggregative principle adopted for society is the same as the rational choice principle for an individual. To achieve this goal, utilitarianism posits an impartial spectator who sympathetically experiences and identifies with each person’s desires and appraises them appropriately. In this process, the impartial spectator is conceived as necessarily composing, integrating, and forming a coherent system of desires out of the desires of all individuals. In constructing this system of desires, many persons are fused into one. Rawls launches a fierce theoretical critique of classical utilitarianism through three theoretical contrasts.

功利主义推理的奥秘在于它很自然地将个人的欲望纳入社会的总体欲望体系中。只有弄清功利主义的推理，人们才能洞察功利主义推理的谬误。功利主义认为个人原则同社会原则之间不存在任何区别。在罗尔斯看来，这是由功利主义的理论旨趣决定的。一般来说，功利主义不关心欲望的总额如何分配，而是关心欲望的总额如何增加。社会如何最大限度地增加欲望的净余额，功利主义只能从个人欲望的总额进行推理。也就是说，对社会采取的整体性原则适用于个人的合理选择原则。为了达到这个目标，功利主义假定了不偏不倚的观察者，以同情的方式来体验和认同每个人的欲望并恰当地估价它。在这个过程中，不偏不倚的观察者就被设想为对所有人的欲望进行必要的组成、整合并形成一个融贯的欲望体系。在建构欲望体系时，许多人就融合成了一个人。罗尔斯通过三个理论对照对古典功利主义展开了激烈的理论批判。

The first is the contrast between the principle of total welfare and the principle of rights. Utilitarianism advocates the principle of increasing total welfare, while “justice as fairness” insists on the priority of liberty and rights. Because it takes the increase of total welfare as its fundamental principle, utilitarianism naturally treats all values that promote this increase as effective means, including the liberties and rights cherished by liberalism. Utilitarianism does not disregard liberty and rights; within the liberal tradition, utilitarianism is also a firm defender of civil liberty and freedom of thought. However, the overall stance of utilitarianism is not individualistic; liberty and rights are merely effective means for promoting total welfare. Utilitarianism permits sacrificing the interests of some for a greater total sum of utility, and may even deprive a minority of their liberty and rights for a greater good. “Justice as fairness,” by contrast, treats liberty and rights as prior values. Justice denies that it is right to deprive some persons of their liberty in order to secure a greater benefit for others. In a just society, basic liberties are taken for granted. The rights secured by justice are not subject to political bargaining or the calculus of social interests. Liberty and rights possess an indisputable priority over welfare.

第一个是福利总量原则与权利原则的对照。功利主义主张福利总量增加的原则，而“公平的正义”坚持自由和权利优先的原则。由于把福利总量的增加视为基本原则，功利主义自然会把促进福利总量增加的一切价值都视为有效的手段，即使是自由主义珍视的自由和权利也不例外。功利主义并不是不重视自由和权利，在自由主义思想谱系上，功利主义也是公民自由和思想自由的坚定捍卫者。但是，功利主义的总体立场不是个人主义，自由和权利只不过是促进福利总量的有效手段。功利主义允许为了得到更大的功利总额而牺牲某些人的利益，甚至会为了更大的利益而剥夺少数人的自由和权利。“公平的正义”则把自由和权利视为优先价值，正义否认为使一些人享受较大利益而剥夺另一些人的自由是正当的。在一个正义的社会里，基本的自由被看作理所当然的。由正义保障的权利不受制于政治的交易或社会利益的权衡。自由和权利相对于福利来说具有无可争议的优先性。

The second is a contrast of the conceptions of the individual and society. Utilitarianism naturally extends the principle of individual choice to society, while “justice as fairness” holds that the principle of social choice is the object of the original contract. Regarding how to treat individuals within society, the positions of utilitarianism and “justice as fairness” are vastly different. Utilitarianism regards the overall system of social desires as a collection of individual desires; hence, the principle of individual choice and the principle of society are the same principle. In this respect, utilitarianism does not take the separateness of persons seriously. “Justice as fairness,” by contrast, emphasizes that the separateness of persons is a fundamental feature of human society: “There exist only individual persons, distinct individual persons each with their own separate life.” The individual is the entity; society is composed of many individuals. Each person in society has a unique value that cannot be measured; each person has entirely different goals, and these goals cannot be aggregated into a larger goal. Based on the uniqueness of individual value, “justice as fairness” holds that the principle of the individual cannot be applied to society. Since this is the case, where does the principle of society come from? “Justice as fairness” holds that it is the result of the construction of human practical reason and the unanimous choice of all under the original contract. The contrast between the conceptions of the individual and society fully reveals the essential divergence between utilitarianism and “justice as fairness” in their treatment of the individual and society. The good society envisioned by utilitarianism is a society of efficient resource management, while the good society hoped for by “justice as fairness” is a cooperative system of mutual benefit. In their fundamental attitude toward society, utilitarianism focuses on interests, while “justice as fairness” focuses on justice.

第二个是关于个人和社会的观念对照。功利主义把个人的选择原则自然地扩展到社会，而“公平的正义”则认为社会选择的原则是原初契约下的目标。如何对待社会中的个人，功利主义与“公平的正义”的立场大相径庭。功利主义将社会总体欲望体系视为个人欲望的集合，因而，个人选择原则和社会原则是同一个原则。就此而言，功利主义没有认真对待个人的分离性。“公平的正义”则强调个人的分离性是人类社会的一个基本特征：“存在的只是个体的人，具有他们自己个别生命的不同的个体的人。”个人是实体，社会是由众多个人构成的。社会中的每个人都有其独特的价值且不可衡量；每个人都有着自己完全不同的目标，这些目标不能被集合成一个更大的目标。基于个人价值的独特性，“公平的正义”认为不能将个人原则运用到社会。既然如此，社会原则从何而来？“公平的正义”认为它是人类实践理性建构的结果并在原初契约下所有人一致的选择。个人和社会观念的对照将功利主义与“公平的正义”对待个人与社会的本质分歧完全展现出来。功利主义设想的好社会是资源有效管理的社会，而“公平的正义”期冀的好社会则是互利互惠的合作体系。在对待社会的基本态度上，功利主义着眼利益，“公平的正义”聚焦正义。

The third is the contrast between teleology and non-teleology. Broadly speaking, utilitarianism belongs to teleology, while “justice as fairness” follows deontology. It must be emphasized that “justice as fairness” is not deontology in the general sense, but a deontology that accepts or considers consequences. Rawls holds: “All ethical doctrines worth our attention must take consequences into account in judging rightness. An ethical theory that does not do so is irrational and crazy.” Strict deontology does not consider consequences; “justice as fairness,” while insisting on the priority of the right over the good, also considers consequences. This deontology differs from strict deontology, and Rawls calls it “non-teleology.” The opposition between teleology and non-teleology is merely a difference in standpoint. So why should the teleology of utilitarianism be criticized? Rawls provides two lines of argument: one is the predicament of the teleological theoretical system, and the other is the constraint of justice on desires.

第三个是目的论与非目的论的对照。大体来说，功利主义属于目的论，而“公平的正义”遵从义务论。需要强调的是，“公平的正义”不是一般意义上的义务论，而是接受或考虑某种结果的义务论。罗尔斯认为：“所有值得我们注意的伦理学都必须在判断正当时考虑结果，不这样做的伦理学理论是不合理的和疯狂的。”严格的义务论不考虑结果，“公平的正义”在考虑正当优先于善的同时还要考虑结果，这种义务论同严格的义务论存在差别，罗尔斯称之为“非目的论”。目的论与非目的论只是立场上的分歧，那么功利主义的目的论为什么就应该被批判呢？罗尔斯给出了两个方面的论证：一个是目的论理论体系的困境，另一个是正义对欲望的约束。

The predicament of the teleological theoretical system lies in the fact that teleology cannot provide a highest end that encompasses all ends, and therefore cannot form a unified theoretical system. Teleology holds that the good is independent of the right. Because the understanding of the good differs, teleology takes various forms. If the good is defined as the realization of human excellence, there is the perfectionism of Aristotle and Nietzsche; if the good is defined as pleasure, there is hedonism; if the good is defined as happiness, there is eudaimonism. Utilitarianism does not specify what the good means, but based on a general form, defines the good as the satisfaction of rational desires. However, the problem lies in whether there is a good that encompasses all desires. The difference between utilitarianism and “justice as fairness” regarding the conception of the individual reappears here. Because concrete individuals exist, there are differences in the degree to which individuals satisfy their desires, and the good, as the goal people pursue, also varies from person to person. Even when facing the same good, each person’s attitude toward it and the weight they assign to it differ, and may even conflict. From the perspective of society, utilitarianism needs a highest good to unify the various and individually differing goods, and needs a dominant end to command the diverse special ends; but utilitarianism cannot find that highest end that encompasses all ends. If there is no dominant end, then teleology itself is mistaken.

目的论理论体系的困境在于目的论无法提供统摄所有目的的最高目的，因而不能形成统一的理论体系。目的论主张善独立于正当。由于对善的理解不一样，目的论有各种形式。如果善被定义为个人卓越性的实现，就有亚里士多德和尼采式的完善论；如果善被定义为快乐，就有了快乐主义；如果善被定义为幸福，就有了幸福论。功利主义并没有说明善是何种含义，而是基于一般的形式，将善定义为理性欲望的满足。然而，问题就在于有没有统摄所有欲望的善。功利主义与“公平的正义”关于个人的观念差异再次出现。由于存在具体的个人，因而，个人对欲望的满足存在程度差异，并且，善作为人们追求的目标也因人而异。即使是面对同一种善，每个人对待它的态度和赋予的权重也不相同甚至互相冲突。从社会的角度看，功利主义需要一种最高的善来统一各种各样、因人而异的善，需要一种支配性目的来统领五花八门的特殊目的；而功利主义找不到那种统摄所有目的的最高目的。如果没有支配性目的，那么目的论本身就是错误的。

The constraint of justice on desires is reflected in the fact that the satisfaction of rational desires is meaningful only under the premise of justice. Utilitarianism holds that the satisfaction of any desire has some value in itself; therefore, when calculating the maximum balance of satisfaction, utilitarianism considers only the satisfaction of the desire itself, without regard to the source or nature of the desire. In contrast, “justice as fairness” imposes strict restrictions on people’s desires and conceptions of the good. Which satisfactions of desires have value, and which pursuits of the good are reasonable—these must all depend on the principles of right and justice. Without the constraint of justice, the satisfaction of desires leads to intuitive theoretical difficulties. If people derive pleasure from mutually discriminating against others or harming others’ freedom to enhance their own dignity, utilitarianism sees nothing wrong, while “justice as fairness” resolutely opposes it. Sacrificing the freedom and rights of others to satisfy one’s own desires is not permitted by justice: “Those interests that require violating justice to obtain are themselves worthless.” “Justice as fairness” also does not regard desires and preferences as given, and then seek the best way to satisfy them; rather, it holds that desires and preferences are constrained by justice, and that desires and preferences that conform to the requirements of right can be developed from just institutions.

正义对欲望的约束体现在理性欲望的满足只有在正义的前提下才有意义。功利主义认为任何欲望的满足本身都具有某种价值，因而，在计算满足的最大余额时，功利主义只考虑欲望本身的满足，不考虑欲望的来源和性质。与之相反，“公平的正义”则对人们的欲望和善观念施加了严格的限制。哪些欲望的满足有价值，哪些善的追求具有合理性，这些都必须依赖正当原则和正义原则。离开正义的约束，欲望的满足会带来直觉上的理论困难。如果人们在相互歧视或者损害别人的自由以提高自己尊严的行为中得到快乐，功利主义认为没有什么不妥，而“公平的正义”则坚决反对。牺牲别人的自由和权利来满足自己的欲望为正义所不许，“那些需要违反正义才能获得的利益本身毫无价值”。“公平的正义”也并不认为欲望和偏好是既定的，然后再寻找满足他们的最好方式，而是认为欲望和偏好受到正义的约束，能够从正义的制度中发展出符合正当要求的欲望和偏好。

II. Critique of Average Utilitarianism

二、平均功利主义批判

Rawls’s critique of classical utilitarianism focuses mainly on basic concepts and principles. When faced with a slightly more complex utilitarianism, merely criticizing classical utilitarianism is insufficient to support “justice as fairness.” When the utilitarian principle is applied to the basic structure of society, people will reject the classical utilitarian principle and instead accept the principle of average utility. The classical utilitarian principle requires institutions to maximize the absolute weighted sum of the expectations of each relevant representative person. This sum is obtained by multiplying the expectation for each position by the number of people in that position, and then adding the results. In contrast, the principle of average utility instructs society not to maximize the total sum of utility, but to maximize average utility. Institutions are to be set so as to maximize the percentage-weighted sum of the expectations of each representative person. To calculate this sum, we multiply the expectation by the social share of the people in the corresponding position. The difference between classical utilitarianism and average utilitarianism lies in whether the focus is on maximizing total utility or average utility; the point of intersection of the theories is population size. When the population size is constant, classical utilitarianism and average utilitarianism are indistinguishable; when the population size changes, the divergence immediately appears.

罗尔斯对古典功利主义的批判主要集中在基本的观念和原则上。面对稍微复杂的功利主义，仅靠批判古典功利主义的方式不足以支持“公平的正义”。当把功利原则运用到社会基本结构上时，人们会拒绝古典功利原则，转而接受平均功利原则。古典功利主义原则要求制度能够最大化各相关代表人的期望的绝对加权和。这一总和是通过用处于相应地位的人数来乘以每一期望，然后再把结果相加得到的。与之相对，平均功利原则指示社会不是要最大化功利总数，而是要最大化平均功利。制度的设定要能最大化各代表人的期望的百分比加权和。为计算这一总额，我们用期望乘以相应地位的人所占的社会份额来得到这个总和。古典功利主义和平均功利主义的分歧在于是关注功利总额最大化还是平均功利最大化，理论的交集点在于人口数量。当人口数量不变时，古典功利主义和平均功利主义没有区别；当人口数量出现变化时，两者分歧立现。

Since total utility is obtained by multiplying the expectation for a given position by the population in that position and then adding, classical utilitarianism would encourage unlimited population growth regardless of a decline in average utility. As a result, classical utilitarianism would accept a very low average welfare. For average utilitarianism, when the population doubles, utility does not necessarily double. As long as the percentages of people in different positions remain unchanged, the total utility remains unchanged. In average utilitarianism, a mere change in population size does not affect total utility; what truly affects total utility is the percentage of expectations of these people. If the percentages do not change and only the population size increases, this will only lead to an infinite reduction in average welfare. Under the condition of constant total utility, average utilitarianism holds that average welfare cannot be reduced infinitely; rather, there should be a minimum level of average welfare. The idea of average utilitarianism was already reflected in Sidgwick: if we can foresee that an increase in population will be accompanied by a decrease in average happiness, or conversely, that an increase in average happiness will be accompanied by a decrease in population, this raises a problem that has not only never been formally pointed out but has actually been overlooked by many utilitarians.

由于总功利是占某一地位的人口数量乘期望再进行相加，因此古典功利主义会无视平均功利的下降而鼓励人口的无限增长。在结果上，古典功利主义会接受很低的平均福利。对于平均功利主义来说，当人口翻番时，功利并不一定翻番。只要不同地位的人的百分比没有改变，功利总额就仍保持不变。在平均功利主义那里，人口数量的单纯变化不会影响功利总量，真正影响功利总量的是这些人口期望所占的百分比。如果百分比没有变化而只是单纯的人口数量增加，只会导致平均福利额无限降低。在总功利不变的情况下，平均功利主义认为平均福利额不能无限降低，而是应该存在一个平均福利的最低额度。平均功利主义思想在西季威克那里已经有所体现：如果我们能预见人数的增加将伴随着平均幸福的降低，或者反过来，平均幸福的提高伴随着人口的减少，就产生了一个不仅从未被正式指出过而且实际上被许多功利主义者忽略了的问题。

If people were to choose between classical utilitarianism and average utilitarianism, average utilitarianism would undoubtedly be the first choice. For the individuals or their representatives making the choice, the purpose of selecting a principle of justice is “to further their own interests, and they have no desire to maximize the total sum of satisfaction under any circumstances.” [1]1 John Rawls, *A Theory of Justice*, rev. ed. (Cambridge, MA: Harvard University Press, 1999), p. 160. To illustrate this, Rawls lists two scenarios of choice under different information conditions: The first is choice under very full information. In this situation, people know that several different societies exist along with their adopted redistribution strategies, also know their own relevant circumstances, and can foresee the welfare they would enjoy in each society. At this point, people would not rely on probabilities but would directly choose the society that favors maximizing their own welfare. The second is choice under insufficient information. When all specific particular information is screened off by the veil of ignorance, people have the same chance of entering different societies or attaining any social position. In this case, a person’s life prospects are highly consistent with the society’s average utility. To obtain better life prospects, people would choose the society that maximizes average utility rather than the society that maximizes total social utility. The above reasoning and argument can prove that average utilitarianism is superior to classical utilitarianism. When the two principles of justice face the challenge of the average utility principle, Rawls must provide more theoretical argumentation and more solid reasons.

如果让人们在古典功利主义和平均功利主义之间进行抉择，平均功利主义无疑将是首选。对于选择的个人或代表人来说，他们选择某种正义原则的目的是“旨在推进他们自己的利益，他们在任何情况下都没有最大限度地增加满足总额的欲望”。为了说明这点，罗尔斯列举了两种信息场景下的选择：第一种是信息非常充分下的选择。在此情境下，人们知道存在几种不同的社会及其采用的再分配策略，也知道自己的相关情况，同时能够预见自己在各个社会中能够享受到的福利。此时，人们就不会依赖概率而直接选择有利于实现自己福利最大化的社会。第二种是信息不充分下的选择。当各种特殊的具体信息被无知之幕遮蔽后，人们进入不同的社会或获得任何社会地位的机会相同。此时，这个人的生活前景就和社会的平均功利高度一致。为了获得更好的生活前景，人们将选择平均功利最大化而不是社会总功利最大化的社会。上述推理和论证能够证明平均功利主义优于古典功利主义。当两个正义原则面对平均功利原则的挑战时，罗尔斯必须给出更多的理论论证和更加坚实的理由。

On the one hand, Rawls believes that average utilitarianism has three major theoretical difficulties that are hard to overcome. The first difficulty is probability calculation. When average utilitarianism engages in reasoning and choice, it must calculate different social positions and their expected utility magnitudes. Generally speaking, the more sufficient the information, the closer the probability calculation is to correctness. When people do not possess all the relevant knowledge, reasoning can only rely on the principle of insufficient reason. The so-called principle of insufficient reason means that people “determine the probabilities of outcomes in the absence of any information whatsoever.” [2]2 Rawls, *A Theory of Justice*, p. 164. When complete evidence is lacking, all possible situations have the same probability. Obviously, whether in a real society or in the original position, individuals do not possess a complete information base, and thus cannot possibly carry out accurate probability calculations. The difficulty of probability calculation will lead people to choose an adventurous approach. People’s choice of the average utility principle actually depends on a presupposed assumption, namely, that they have an equal opportunity to enter the society that maximizes average utility. The real situation might be entering an extremely bad society, which would entail unbearable consequences for the choosing individual and their descendants. Rational people should avoid taking risks.

一方面，罗尔斯认为平均功利主义有着难以克服的三大理论困难。第一个困难是概率计算。平均功利主义在进行推理选择时，要计算不同社会地位及其预期功利大小。通常而言，信息越充分，概率计算就越接近正确。当人们并不具备全部的相关知识，推理就只能依赖不充足理由原则。所谓不充足理由原则是指人们“在不具备任何信息的情况下来确定结果的概率”。当缺乏完全的证据时，各种可能的情况就具有相同的概率。显然，无论是在现实社会中还是在原初状态下，个人都不具备全部的信息基础，因而根本无法进行准确的概率计算。概率计算的困难将导致人们选择采取冒险的方式。人们选择平均功利原则实际上依赖一个前提假定，即具有同等的机会进入到平均功利最大化的社会中。真实的情况则可能是进入一个极其糟糕的社会，这对于选择的个人及其后代来说，都会面临不可忍受的后果。理性的人应该避免冒险。

The second difficulty is that average utilitarianism cannot determine a unified expectation. Although the average utility principle does not rely entirely on utility calculation and interpersonal comparison like the classical utility principle, as a new form of utilitarianism, average utilitarianism also faces the problems inherent to utilitarianism itself. Regarding utility, average utilitarianism treats the objects of individual preferences as individual values. According to Rawls’s view, each person has their own specific preferences, and the objects of choice also differ enormously. There are as many different utilities as there are distinct individuals. However, the reasoning of the average utility principle relies on a unified expectation or utility. A unified expectation requires utility calculation, and the prerequisite for utility calculation is still interpersonal comparison. The problem has already been revealed in classical utilitarianism: because different people have different utilities, both utility calculation and interpersonal comparison are impossible. Based on this, average utilitarianism naturally cannot determine a unified expectation either.

第二个困难是平均功利主义无法确定一个统一的期望。平均功利原则虽然不像古典功利原则那样完全依赖功利的计算及其人际比较，但是，作为功利主义的新形式，平均功利主义也面临功利主义本身具有的问题。对于功利，平均功利主义把个人偏好的对象作为个人的价值看待。按照罗尔斯的看法，每个人都有自己的具体偏好，选择对象也存在巨大的差异。有多少独特的人，就有多少不同的功利。然而，平均功利原则的推理却要依赖统一的期望或功利。统一的期望需要功利计算，而功利计算的前提还是人际比较。问题已经在古典功利主义那里得到揭示，由于不同的人有不同的功利，功利计算和人际比较都不可能。基于此，平均功利主义自然也无法确定统一的期望。

The third theoretical difficulty is that average utilitarianism cannot avoid the serious charge of sacrificing individuals. Rawls does not directly address this issue, but we can deduce it from his critique. Average utilitarianism considers maximizing average utility, or it hopes to maintain a certain level of average welfare. If total utility remains unchanged, average utilitarianism might require optimizing the population size. The optimal population will achieve the highest average utility. Under the demand for the highest average utility of the optimal population, although society would not allow extreme cases of deliberately sacrificing individual lives, it cannot theoretically escape this kind of doubt. In a negative sense, average utilitarianism, like classical utilitarianism, would endorse this policy or behavior: it can turn a blind eye to the dire circumstances of those who reduce social utility or have negative utility indices. The average utility principle ostensibly cares about everyone’s interests, but in reality it can also sacrifice everyone’s interests.

第三个理论困难是平均功利主义无法避免牺牲个人的严重指控。罗尔斯没有直接论及这个问题，但是，我们能够从他的批评中推导出来。平均功利主义考虑平均功利最大化，或者说它希望保持一定水平的平均福利额，在总功利不变的情况下，平均功利主义可能会要求人口数量最优化。最优人口将实现最高的平均功利。在最优人口的最高平均功利的要求下，社会虽然不会允许出现故意牺牲个别人生命的极端情况，但是，它无法从理论上摆脱这种质疑。从消极意义上讲，平均功利主义同古典功利主义一样都会认可这种政策或行为：可以对那些减少社会功利或功利指数消极的人的糟糕处境熟视无睹。平均功利原则表面上关注每个人的利益，实际上也会牺牲每个人的利益。

On the other hand, compared with the average utility principle, the two principles of justice have greater theoretical advantages. The first theoretical advantage is that the two principles of justice do not require probability calculation. The two principles of justice do not consider probability issues at all, so naturally there is no difficulty of probability calculation. The second theoretical advantage is that the two principles of justice will not bring unbearable consequences to the choosing individuals. People in the original position will not adopt an adventurous attitude to choose the average utility principle, but will choose the two principles of justice, because the difference principle maximizes the protection of people’s welfare level. The third theoretical advantage concerns the index of justice. Average utilitarianism cannot calculate utility, while the two principles of justice use primary goods to identify the least advantaged members of society, thereby establishing an effective reference point for justice. More importantly, the two principles of justice place basic liberties and rights in a priority position; any practice that sacrifices liberties and rights to obtain benefits violates the requirements of justice. The population optimization problem faced by average utilitarianism will not exist in “justice as fairness.”

另一方面，同平均功利原则相比，两个正义原则具有较大的理论优势。第一个理论优势是两个正义原则不需要进行概率计算。两个正义原则根本不考虑概率问题，自然也就不存在概率计算的困难。第二个理论优势是两个正义原则不会给选择的人带来不可忍受的后果。原初状态中的人不会采取冒险的态度去选择平均功利原则，而是选择两个正义原则，因为差别原则最大程度地保障了人们的福利水平。第三个理论优势是关于正义的指标问题。平均功利主义无法计算功利，而两个正义原则采用基本善来识别社会最不利者，从而为正义确定了有效的参照点。更为重要的是，两个正义原则将基本自由和权利置于优先地位，任何牺牲自由和权利获取利益的做法都违背正义的要求。平均功利主义面临的人口最优化问题在“公平的正义”中不会存在。

Not only do the two principles of justice have these theoretical advantages, they can also obtain absolute theoretical support. The first theoretical support is the strain of commitment. In Rawls’s view, the contract in the original position is not a contract in the ordinary sense, but is meant to settle once and for all the distribution of social benefits; its impact on people’s life prospects is absolute. Because the original contract is final and permanent, there will be no second chance. The choosing parties must consider not only their own interests but also take responsibility for their descendants; therefore, any form of unbearable consequences will be rejected. The average utility principle entails a huge risk of placing people in unbearable consequences. Given this huge risk, the strain of commitment becomes particularly acute, and the parties in the original position will choose the two principles of justice. The second theoretical support is stability. Regarding stability, the utility principle requires an identification of interests among people, while the two principles of justice require that everyone have a sense of justice. From the perspective of moral psychology and human motivation, identification of interests among people is very difficult, while the two principles of justice can generate a sense of justice that supports themselves, and the latter is more stable. The third theoretical support is the requirement that “man is an end.” Regarding man as an end, the average utility principle, because it maximizes average utility, treats people as means that can be sacrificed, whereas the basic structure of society regulated by the two principles of justice “shows a desire on the part of everyone to treat one another not merely as means but as ends in themselves.” [3]3 Rawls, *A Theory of Justice*, p. 157.

两个正义原则不但具有这些理论优势，而且能够获得绝对性的理论支持。第一个理论支持是承诺的压力。在罗尔斯看来，原初状态中的契约不是一般意义上的契约，而是要一劳永逸地解决社会利益的分配问题，它对人们生活前景的影响是具有绝对性的。由于原初的契约是最终的和永久性的，就不会有第二次机会。选择的当事人既要考虑自己的利益，还要为自己的子孙后代负责，因而，任何形式的不可忍受的后果都将被拒绝。平均功利原则存在将人们置于不可忍受的后果之巨大风险，鉴于这种巨大风险，承诺的压力就显得特别尖锐，原初状态中的当事人会选择两个正义原则。第二个理论支持是稳定性。对于稳定性，功利原则要求一种人与人之间的利益认同，而两个正义原则要求人们都有正义感。从道德心理学和人类动机来看，人与人之间的利益认同非常困难，而两个正义原则能够产生出支持自己的正义感，后者更具稳定性。第三个理论支持是“人是目的”的要求。对于人是目的，平均功利原则会因为平均功利最大化而把人视为可以牺牲的手段，而两个正义原则规范下的社会基本结构“表明了人们希望相互不把对方仅仅作为手段、而也作为自在的目的来对待的意愿”。

III. Critique of Limited Utilitarianism

三、有限功利主义批判

Although Rawls has consistently regarded utilitarianism as his main theoretical opponent, what truly poses a challenge to “justice as fairness” is neither classical utilitarianism nor average utilitarianism, but limited utilitarianism. Limited utilitarianism comes from what Rawls calls “mixed conceptions,” sometimes also referred to as “the limited utility principle.” “Mixed conceptions” are mixtures of the average utility principle and the two principles of justice to varying degrees. In the list of conceptions of justice for choice in the original contract, Rawls specifically enumerates different forms of mixed conceptions:

罗尔斯虽然一直把功利主义视为主要的理论对手，真正对“公平的正义”构成挑战的不是古典功利主义，也不是平均功利主义，而是有限功利主义。有限功利主义来自罗尔斯所说的“混合的观念”，有时候也称为“有限功利原则”。“混合的观念”是平均功利原则同两个正义原则在不同程度上的混合。罗尔斯在原初契约的正义观选择清单中专门列举了混合观念的不同形式：

B. Mixed conceptions. One of the following is substituted for A2 above:

B.混合的观念。以下面的一个代替上面的A2

1. the principle of average utility, or

1.平均功利的原则，或

2. the principle of average utility subject to either of the following constraints:

2.受到下面任一限制的平均功利原则：

(a) that a certain social minimum be maintained,

(a)应当维持某种社会的最低受惠值，

(b) that the total distribution not be too wide, or

(b)总分配不应太广泛，或者

3. the principle of average utility subject to either (a) or (b) of B2 and combined with the principle of fair equality of opportunity.

3.受到B2中(a)或(b)的限制并与机会公平平等原则相结合的平均功利原则。

The A2 that Rawls refers to denotes the principle of fair equality of opportunity and the difference principle. Depending on the differing elements of the mixture, there are three distinct forms of mixing: the first form replaces the entire second principle of justice with the principle of average utility, i.e., the principle of equal liberty plus the principle of average utility; the second form replaces the second principle of justice with a restricted principle of average utility, i.e., the principle of equal liberty plus the restricted principle of average utility; the third form replaces the difference principle with a restricted principle of average utility, i.e., the principle of equal liberty plus the principle of fair equality of opportunity plus the restricted principle of average utility. The restrictive conditions include a certain social minimum and that the overall distribution should not be too wide (i.e., a small distributional deviation). Because the principle of fair equality of opportunity holds an extremely important status and value and must be retained, Rawls focuses his critique most heavily on the third form of mixing rather than the first two.

罗尔斯所说的A2是指公平的机会平等原则和差别原则。根据混合的要素不同，混合有三种不同的形式：第一种混合，用平均功利原则代替整个第二个正义原则，即平等的自由原则+平均功利原则；第二种混合，是以受到限制的平均功利原则代替第二个正义原则，即平等的自由原则+受到限制的平均功利原则；第三种混合，是以受到限制的平均功利原则代替差别原则，即平等的自由原则+公平的机会平等原则+受到限制的平均功利原则。限制条件包括某种社会最低受惠值和总分配不应太广泛（即分配偏差较小）。由于公平的机会平等原则具有极为重要的地位和价值，公平的机会平等原则要予以保留，因此，罗尔斯关注和批判得最多的是第三种形式的混合而不是前两种。

In *A Theory of Justice*, Rawls does not analyze and critique the restricted utility principle within the original position, but only discusses and compares it within the institutional arrangements for distributive justice. On the one hand, the argument against the mixed conception is evidently more difficult than the argument against the principle of average utility. As Rawls himself states: “Many writers seem to accept some form of utilitarian view (though it is vaguely expressed as a balance and harmony of various social interests), and they clearly presuppose a mixed constitutional system that guarantees basic liberties at a minimum level. Thus, they actually hold some mixed doctrine, and the strong argument from liberty can no longer be used as before.” On the other hand, the social minimum that restricts the principle of average utility is an institutional arrangement, and how an institution concerning distribution functions can only be elaborated after some principle of justice has been determined.

在《正义论》中，罗尔斯并没有在原初状态下分析和批判有限功利原则，而是在分配正义的制度安排中才进行了讨论和比较。一方面，反对混合观念的论证显然比反对平均功利原则的论证更为困难。就像罗尔斯所说：“许多作家看来承认某种变形的功利主义观点（尽管这种观点被含糊地表达为各种社会利益的平衡与和谐），他们明显地预先假设了一种在最低程度上保证基本自由的混合宪法体系。这样，他们实际上便持有某种混合学说，因而来自自由的强有力论据就不可能像以前那样使用了。”另一方面，限制平均功利原则的社会最低受惠值是一种制度安排，而一种有关分配的制度如何发挥作用也只有在某种正义原则确定后才能展开。

Regarding the grounds for critique, an important reason Rawls criticizes classical utilitarianism is that individual basic liberties and rights are inviolable, and this reason can also be applied to average utilitarianism. The priority of liberty no longer serves as a powerful argument, which places theoretical pressure on “justice as fairness.” If other grounds are to be invoked, these grounds must possess strong theoretical persuasiveness. In fact, apart from the argument from liberty, there is no better moral reason that can deliver a decisive critique of restricted utilitarianism. Rawls can only focus his theoretical efforts on the “utilitarian standard” and the “reasons for the utilitarian standard.” The former aims to emphasize that the utilitarian standard cannot be determined, while the latter aims to emphasize that no utilitarian standard can be established on appropriate moral grounds. Rawls also acknowledges that these reasons are far less powerful than the argument from liberty. Even so, “we still need to examine the reasons against the utilitarian standard, although these reasons are evidently not as decisive as those against the typical and average utilitarian theories.” For the sake of clarity in argumentation, Rawls further divides the third form of mixing into two types: one type replaces the difference principle with the principle of average utility restricted by a certain social minimum; the other type replaces the difference principle with the principle of average utility restricted by a certain distributive requirement (either according to this requirement itself or according to its connection with a suitably chosen social minimum).

就批判理由来说，罗尔斯批判古典功利主义的重要理由是个人基本自由和权利不容侵犯，这一理由在平均功利主义那里也能得到运用。自由的优先性不再成为一个有力的论据，这给“公平的正义”带来了理论论证上的压力。如果诉诸其他理由，那么，这些理由就必须具有强大的理论说服力。实际上，除了自由的论据，没有更好的道德理由能够给予有限功利主义以决定性的批判。罗尔斯只能将理论的着力点集中于“功利标准”及其“功利标准的理由”上面。前者意在强调功利标准无法确定，后者意在强调任何一种功利标准的确立都找不到合适的道德理由。罗尔斯也承认这些理由相比于自由来说力量都小得多，即使是这样，“我们仍需要考察反对功利标准的理由，尽管这些理由显然不像反对典型的和平均的功利理论的理由那样具有决定性的作用”。为了论证的清晰化，罗尔斯又将第三种混合形式分为两种形式：一种是由某种社会最低受惠值所限制的平均功利原则代替差别原则；另一种是用某种分配要求（或者根据这种要求自身或者根据它同某一适当选择的最低受惠值的联系）来限制的平均功利原则代替差别原则。

The first type, namely the restricted utility principle conditioned by a social minimum, possesses an intuitionist character. From a factual standpoint, those who use the two principles of justice might also think this way: society should maintain a balance between pursuing the maximum of average utility and maintaining an appropriate social minimum. How the social minimum is determined and how it adjusts to changes in circumstances, Rawls argues, utilitarianism cannot explain clearly; it can only rely on intuition. Conversely, from a moral standpoint, the social minimum in fact depends on the difference principle. Because the difference principle takes the least advantaged as the standpoint of justice, and the social minimum is determined with reference to the least advantaged. The determination and variation of the social minimum lie in the conception of justice, and this conception of justice can only be “justice as fairness” with the difference principle at its core. More crucially, the difference principle demands more than what the social minimum stipulates; it not only sets a higher standard for the social minimum but also aims to achieve greater equality above that minimum.

第一种形式即以社会最低受惠值为限制条件的有限功利原则具有直觉主义的特征。从事实来看，使用两个正义原则的人也可能这样认为：社会应当在追求平均功利的最大值和维持适当的社会最小值之间维持一种平衡。社会最低受惠值如何确定以及它如何根据环境的变化而调整，罗尔斯认为功利主义都无法解释清楚，它只能依赖直觉。相反，从道德理由来看，社会最低受惠值事实上依赖差别原则。因为差别原则以最不利者作为正义的观察点，而社会最低受惠值根据最不利者来确定。社会最低受惠值的确定和变化在于正义观，而这个正义观只能是以差别原则为核心的“公平的正义”。更为关键的是，差别原则要比社会最低受惠值规定的要求要高，它不但有更高的社会最低受惠值标准，而且还要实现社会最低受惠值之上的更大平等。

Rawls devotes considerable effort to critiquing the second type. The second type advocates determining a certain distributive requirement, including a minimum, to maximize average utility. When everyone receives the same utility, the distributional deviation among people is smallest. Comparatively, this form of restricted utilitarianism cares more about the least advantaged than general average utilitarianism, and it has two notable advantages: first, the distributional deviation is small, satisfying the requirement of equality, which aligns with Rawls’s egalitarian demands; second, it concerns itself with the gains of the least advantaged, which converges with the intent of the difference principle. Even with these advantages, Rawls believes it still has three major problems. First, to prevent inequality in distribution, restricted utilitarianism proposes subtracting the standard deviation of the distribution (e.g., a fraction or multiple of the standard deviation). However, how the standard deviation of distribution that prevents excessive inequality is determined, and how this parameter should vary with changes in the mean, remains unclear. Second, while maintaining a certain minimum, restricted utilitarianism simultaneously affirms that both pursuing greater average welfare and pursuing a more equal distribution of shares are desirable goals. But how to balance these two goals, restricted utilitarianism either adopts commonsensical notions or relies on some underlying conception of justice. Third, in the actual operation of a democratic society, the formulation of social policy does not aim at maximizing some value, because the government cannot equally affect everyone’s interests and can only choose a standpoint that, from the perspective of the least advantaged, conforms to the normal practice of democratic politics. In Rawls’s view, all three factors can only depend on the prior difference principle.

罗尔斯对第二种形式的批判颇费笔墨。第二种形式主张确定某种分配要求包括最低受惠值来最大程度地追求平均功利。当每个人都获得相同的功利时，人们之间的分配偏差是最小的。比较而言，这种有限功利主义比一般的平均功利更关心最不利者，它有两个显著的优点：第一，分配偏差较小，满足平等的要求，这同罗尔斯的平等主义要求一致；第二，关心最不利者的获益，这又与差别原则的旨趣趋同。即使存在这些优点，罗尔斯认为它依然存在三大问题。第一，为了防止分配的不平等，有限功利主义主张减去分配的标准差（如标准差的分数或倍数）。但是，防止不平等过度分化的分配标准差是如何确定出来的，并且这个参数应该怎样随着平均数的变化而变化，这依然不清楚。第二，有限功利主义在保持某种最低额的前提下，同时肯定追求较大的平均福利和较平等的分配份额都是值得向往的目标。但是，如何平衡这两种目标，有限功利主义要么采用常识性的观念，要么依赖某种背后的正义观。第三，在民主社会的实际运作中，社会政策的制定不是追求某种最大值，因为政府不可能平等地影响每个人的利益，而只能选择从最不利者的观点看才符合民主政治的常规。在罗尔斯看来，这三个因素都只能依赖在先的差别原则。

Proceeding from this, Rawls argues that restricted utilitarianism cannot replace “justice as fairness” at all. The restrictive conditions of restricted utilitarianism either rely on intuition or on the difference principle. If they rely on intuition, these intuitive judgments often conflict with each other and contain theoretical contradictions. Restricted utilitarianism does not examine these contradictions but instead directly uses these judgments to criticize the difference principle. However, to replace the difference principle starting from these judgments requires clear theoretical reasoning and a definite object that can embody the replacement of the difference principle, which restricted utilitarianism evidently cannot achieve. Only the difference principle can balance the various intuitive judgments within restricted utilitarianism and determine the order of priority. If it relies on the difference principle, this is undoubtedly an admission of the difference principle; the restrictive conditions are merely the application of social policy under the conception of the difference principle and thus cannot form institutional principles that challenge the difference principle.

由此出发，罗尔斯认为有限功利主义根本不能代替“公平的正义”。有限功利主义对功利主义的限制条件要么依赖直觉，要么依赖差别原则。如果依赖直觉，这些直觉判断往往互相冲突且存在理论矛盾。有限功利主义并不审视这些矛盾，而是直接从这些判断出发来批评差别原则。但是，从这些判断出发来替换差别原则，这之间需要明确的理论推理以及能够体现代替差别原则的明确对象，而有限功利主义在这方面显然做不到。只有差别原则能够平衡有限功利主义中的各种直觉性判断和确定优先性秩序。如果依赖差别原则，这无疑是已经承认了差别原则，限制条件不过是差别原则观念下的社会政策运用，因而并不能形成挑战差别原则的制度原则。

The fundamental difficulty with restricted utilitarianism’s inability to replace “justice as fairness” lies in the vagueness of the “utilitarian standard” and the absence of “reasons for the utilitarian standard.” Restricted utilitarianism, under the premise of the welfare conception, needs to determine the average (or total) utility number. Due to the vagueness of utility itself, restricted utilitarianism shares the fatal defect common to all utilitarianism: it cannot establish a method for the interpersonal comparison of social utility. Moreover, regarding which factors should constitute the utility function for interpersonal comparison, restricted utilitarianism cannot provide clear reasons. Because of disagreements over the utility function and the problem of aggregating utility itself, the attempt by utilitarianism to form a standard for interpersonal comparison in cardinal form can never be established. It is for this reason that Rawls emphasizes: “The contract view holds that once we see this, we will understand that we had best abandon entirely the idea of measuring and calculating welfare.”

有限功利主义无法替代“公平的正义”，其根本困难在于“功利标准”的模糊和“功利标准理由”的缺失。有限功利主义在福利观念的前提下需要确定平均（或总体）功利数。由于功利本身的模糊性，有限功利主义存在着功利主义共有的致命缺陷：无法建立社会功利的人际比较的测度方法。而且，对于哪些因素应该成为人际比较的功利函数，有限功利主义也无法给出明确的理由。由于功利函数的分歧以及功利自身的加总问题，功利主义试图以基数形式形成人际比较的标准根本不可能建立起来。正因为如此，罗尔斯才强调：“契约论认为，只要我们看到这一点，我们就将明白：我们最好完全放弃测度和计算福利的观念。”

IV. Critique of the Critique of Utilitarianism

四、功利主义批判之批判

Through his critique of three different theoretical forms—classical utilitarianism, average utilitarianism, and limited utilitarianism—Rawls argues that “justice as fairness” can replace utilitarianism. Moreover, following Rawls’s theoretical critique, Western moral philosophy gradually formed a trend of criticizing utilitarianism, to the point that utilitarianism seemed to have reached a dead end. Even a rule-utilitarian philosopher like Brandt noted with some concern: “Many articulate philosophers are critics of utilitarianism. A recent article in the *Yale Law Journal* speculates that there are now only about five or six supporters of utilitarianism trying to save it from extinction.” Brandt’s assessment is somewhat alarmist. Whether utilitarianism is as lacking in sufficient justification and as riddled with theoretical flaws as Rawls criticizes it to be—this is precisely what we analyze and discuss as the theoretical intention behind Rawls’s critique of utilitarianism. Rawls’s triple critique of utilitarianism is a progressively deepening process, and also a process in which the difficulty of his theoretical critique continuously increases. Following Rawls’s line of thought, people need to re-examine his theoretical critique of utilitarianism, and through the critique and counter-critique of theory, obtain an equally better understanding of both “justice as fairness” and utilitarianism.

通过对古典功利主义、平均功利主义和有限功利主义三种不同理论形式的批判，罗尔斯认为“公平的正义”能够取代功利主义。不仅如此，经过罗尔斯的理论批判，西方道德哲学逐渐形成了一股批判功利主义的思潮，功利主义似乎走到了穷途末路，以至于像布兰特这样的规则功利主义哲学家不无担忧地指出：“许多善于表达的哲学家都是功利主义的批评者。最近发表在《耶鲁法律评论》上的一篇文章推测说，现在大约只有五六位功利主义的支持者试图使功利主义免于灭亡。”布兰特的评述有点危言耸听。功利主义是否就像罗尔斯批评的那样缺乏充足道理、理论体系千疮百孔，这正是我们分析和论述罗尔斯批判功利主义的理论初衷。罗尔斯对功利主义的三重批判是一个逐渐深入的过程，也是其理论批判难度不断增高的过程。跟随罗尔斯的思路，人们需要重新审视他对功利主义的理论批判，通过理论的批判与再批判，获得对“公平的正义”和功利主义同等的更好的理解。

First, Rawls’s critique of classical utilitarianism is primarily based on institutional value. In his critique of classical utilitarianism, Rawls elaborates a key idea: social justice is the institutional value of modern society, while classical utilitarianism cannot provide a strong explanation at the level of social institutions that is compatible with both justice and utility maximization. The institutional value that social justice upholds is freedom and equality. In the pursuit of utility maximization, classical utilitarianism often lacks a clear explanation of how to effectively protect individuals’ basic freedoms, rights, and equality, and may even sacrifice or violate freedom and rights for the sake of utility maximization. It is precisely based on this critique of institutional value that people feel utilitarianism cannot provide an appropriate justification for political values and principles of social justice. Utilitarianism loses out in the theoretical confrontation with “justice as fairness.” However, some utilitarian theorists argue that Rawls’s critique of classical utilitarianism does not represent the entirety of utilitarianism; classical utilitarianism is merely a simplified, extreme version of utilitarianism. In essence, classical utilitarianism is a form of act-utilitarianism, and act-utilitarianism indeed has many unavoidable theoretical difficulties.

首先，罗尔斯对古典功利主义的批判主要立足于制度价值。在对古典功利主义的批判中，罗尔斯阐释了关键性的思想：社会正义是现代社会的制度价值，而古典功利主义无法在社会制度层面上形成正义和功利最大化能够兼容的强有力解释。社会正义维护的制度价值是自由和平等，在追求功利最大化的同时如何有效地保护个人的基本自由、权利以及平等，古典功利主义往往缺乏清晰的解释，甚至会以功利最大化牺牲或侵犯自由和权利。正是立足于制度价值的批判，人们觉得功利主义不能为政治价值和社会正义原则提供恰当的证明。功利主义在与“公平的正义”之理论交锋中败下阵来。然而，一些功利主义理论家认为罗尔斯对古典功利主义的批判并不能代表功利主义的全部，古典功利主义只是一种简单化的极端的功利主义版本。就其本质而言，古典功利主义是一种行为功利主义，而行为功利主义的确存在许多不可避免的理论困难。

If utilitarian theorists could construct a complex and refined version of utilitarian theory that both better explains institutional values such as freedom and rights and promotes utility maximization, then this kind of utilitarianism would have considerable theoretical counterforce when facing Rawls’s criticism. Rule-utilitarianism belongs to this new theoretical paradigm. The claim of rule-utilitarianism is as follows: a rule-utilitarian holds that a right action is one permitted by a moral rule that is optimal for the society in which the agent lives. An optimal rule is one used to maximize welfare or the good (hence utilitarian). In short, rule-utilitarianism holds that people act according to the requirements of moral rules, and adhering to moral rules can produce utility maximization. In the general exposition of rule-utilitarianism, utilitarianism has a high degree of compatibility with the basic requirements of “justice as fairness.”

如果功利主义理论家能够构造一种复杂精致版本的功利主义理论，它既能更好地解释自由和权利等制度价值，又能推进功利最大化，那么，这种功利主义在面对罗尔斯的批评时将具有很大的理论回击力。规则功利主义就属于这种新的理论范型。规则功利主义的主张是这样的：一个规则功利主义者认为正确的行为是被道德规则允许的行为，这种道德规则对行为主体所处的社会来说是最优的。一个最优的规则就是被用来最大化福利或者善（因此是功利的）。简言之，规则功利主义认为人们的行为是按照道德规则的要求进行，而遵守道德规则就能够产生功利最大化。在规则功利主义的一般性论述中，功利主义同“公平的正义”之基本要求具有很高的契合度。

On the one hand, rule-utilitarianism, unlike act-utilitarianism, is not a direct form of utilitarianism but an indirect one; it is by following rules and acting according to them that maximum utility is produced. For Rawls, our primary task is to construct ideal principles of justice and act according to them. Rule-utilitarianism can likewise regard the institutional principles of justice, especially those concerning everyone’s equal freedom and rights, as moral rules that must be followed, because no moral rule would consider actions that violate individual freedom and rights as morally correct. Obeying moral rules produces the expected utility (the good). In explaining the opposition between deontology and teleology, Rawls emphasized that his justice is “non-teleological,” implying that principles of justice also concern consequences (the good), and adhering to them will necessarily bring about the good. In terms of consequences, rule-utilitarianism and “justice as fairness” have a high degree of theoretical similarity. Therefore, criticizing utilitarianism based on the tense conflict between freedom/rights and the good is, at least from the perspective of rule-utilitarianism, insufficient.

一方面，规则功利主义并不像行为功利主义那样属于直接的功利主义，而是间接的功利主义，它是遵守规则并按照规则行动产生出最大功利。对于罗尔斯来说，我们首要的任务是建构理想的正义原则并按照正义原则行事。规则功利主义一样可以把制度的正义原则尤其涉及每个人的平等的自由和权利原则看作是必须遵守的道德规则，因为没有哪个道德规则会认为侵犯个人自由和权利的行为在道德上是正确的。服从道德规则会产生预期的功利（善）。罗尔斯在解释义务论和目的论的对立时曾强调自己的正义是“非目的论的”，言外之意，正义原则同样关注结果（善），遵守正义原则必定会带来善。在后果的意义上，规则功利主义同“公平的正义”具有高度的理论相似性。因此，用自由、权利同善的紧张冲突来批评功利主义，至少在规则功利主义看来并不充分。

On the other hand, rule-utilitarianism is more consistent with moral intuition in explaining people’s behavior. Rule-utilitarianism advocates acting according to moral rules; in daily behavior, people do not directly calculate the magnitude of act-utility but act according to everyday moral rules. Rule-utilitarianism only calculates utility when faced with judging which moral rule is optimal. The difficulty of rule-utilitarianism also lies in how to determine the basis and reason for judging the optimal moral rule. Generally speaking, rule-utilitarianism holds that those ideal moral norms are optimal because they are universal in application and can produce the greatest good in regulating people’s behavior. From the perspective of a single moral rule, following it may not be consistent with utility maximization, but genuine, ideal, and universalized moral rules should maximize overall and long-term utility. Following ideal moral rules can be consistent with utility maximization. The largest and most universal moral rules are institutional principles; rule-utilitarianism would require people to follow institutional rules, respect people’s freedom and rights, and respect and recognize the value of individual personhood.

另一方面，规则功利主义在解释人们的行为时更加符合道德直觉。规则功利主义主张按照道德规则行事，人们在日常行为中并不是直接计算行为功利的大小，而是按照日常的道德规则行事。规则功利主义只有在面临判断哪一种道德规则是最优的时候才会计算功利大小，规则功利主义的难题也在于如何确定判断最优道德规则的依据和理由。一般来说，规则功利主义认为那些理想的道德规范是最优的，因为理想的道德规范在应用上具有普遍性，在规范人的行为时能够产生最大的善。从单个的道德规则来看，遵守道德规则未必与功利最大化相一致，但是，真正的、理想的和普遍化的道德规则应该是使整体的和长远的功利最大化。遵守理想的道德规则与功利最大化能够一致。最大最普遍的道德规则是制度原则，规则功利主义会要求人们遵守制度规则，尊重人们的自由和权利，尊重和承认人的个体性价值。

Second, Rawls’s critique of average utilitarianism does not target the basic concept of utilitarianism, but rather the moral reasoning in the original position. Rawls rejects average utilitarianism on the grounds of the difficulty of probability calculation and the unbearable consequences of risk-taking, while the difference principle can overcome these problems. In fact, under conditions of uncertainty, people have two different kinds of reasoning: one is what Rawls calls the “maximin rule,” and the other is “expected utility maximization.” Regarding the restrictive rules used in reasoning, both the “risk-taking” that Rawls criticizes and his own “risk-avoidance” are assumed psychological motives. Shielded by the veil of ignorance, Rawls believes that principles of justice should not consider the special fact of individual risk-taking; correspondingly, principles of justice should also not consider avoiding risk, because avoiding risk is also a special fact. In terms of what constitutes a special fact, Rawls clearly adopts an interpretation favorable to himself. In terms of reasoning, the maximin principle has obvious irrationality. For example, in a situation where a person chooses to fly to a distant place for work to live a decent life, or chooses to stay locally and live on a meager salary with difficulty, according to Rawls’s maximin reasoning, this person should stay locally because they can avoid the worst outcome of a plane crash. But maximin reasoning is clearly unreasonable, because the probability of a plane crash is very low. According to this reasoning, people should not even cross the street, nor drive over a bridge, nor even get married (marriage can also bring disaster to people). A rational chooser should choose to go to the distant place for work based on expected utility maximization.

其次，罗尔斯对平均功利主义的批判不在于功利主义的基本理念，而在于原初状态下的道德推理。罗尔斯拒斥平均功利主义的理由在于概率计算的困难以及冒险引发的不可承受之后果，而差别原则能够克服这些问题。实际上，在不确定条件下人们有两种不同的推理：一种是罗尔斯所说的“最大最小化规则”，另一种是“预期功利最大化”。就推理使用的限制规则来说，罗尔斯批评的“冒险”和自己的“不冒险”都属于假定的心理动机。由于受无知之幕的屏蔽，罗尔斯认为正义原则不应该考虑个人冒险这种特殊事实；与此相应，正义原则也不应该考虑避免冒险，因为避免冒险也是特殊事实。在何为特殊事实方面，罗尔斯明显采取了有利于自己的解释。就推理来说，最大最小化原则具有明显的不合理性。例如，在一个人选择坐飞机去很远的地方工作而过上体面的生活，或者是选择继续留在当地依靠微薄薪资而艰难活着这件事上，依据罗尔斯的最大最小化推理，这个人应该留在当地，因为他能避免飞机失事的最坏结果。但最大最小化推理明显不合理，因为飞机失事的概率很低。依据这种推理，人们甚至都不应该穿越马路，也不应该驾车过桥，甚至都不应该结婚（结婚也能给人带来灾难）。一个理性选择者应该依据预期功利最大化而选择去远方工作。

Finally, Rawls’s critique of utilitarianism reveals to us the following theoretical question: Is there a form of utilitarianism that both avoids the worst outcomes and can act in accordance with moral rules while pursuing the maximization of expected utility? Or, further, is there a more complex utilitarianism that both acts under moral rules and, in some appropriate way, ensures that people avoid the worst situations while promoting the maximization of utility—a new utilitarianism? We hold that limited utilitarianism is precisely this complex version of utilitarianism. Classical utilitarianism and average utilitarianism both have clear theoretical representatives: the former is represented by figures such as Henry Sidgwick, while the latter is supported by John Harsanyi. Rawls also provides theoretical annotations and citations when analyzing and critiquing these two forms of utilitarianism. However, as to who is the theoretical representative of limited utilitarianism, Rawls does not specify clearly. Limited utilitarianism is a theoretical model constructed by Rawls and a theoretical target for his critique.

最后，罗尔斯对功利主义的批判向我们揭示了这样的理论问题：是否存在一种既避免最坏状况而又可以按照道德规则要求行动、追求预期功利最大化的功利主义？或者进一步说，有没有更加复杂的功利主义，既在道德规则下行事，又可以以某种适当的方式保证人们避免最坏的情况，而且促进功利最大化的新功利主义？我们认为，有限功利主义就是这种复杂版本的功利主义。古典功利主义和平均功利主义两者都有明确的理论代表，前者是以西季威克等为代表，后者则以哈桑依为支持者，罗尔斯分析和批判这两种功利主义时也做了理论注释和引用。但对于谁是有限功利主义的理论代表，罗尔斯没有明确说明。有限功利主义是罗尔斯构造出来的理论范型和批判的理论靶子。

Considering the new developments in utilitarianism, the limited utilitarianism constructed by Rawls can be regarded as a special form of rule utilitarianism. First, it treats the principle of equal liberty and the principle of fair equality of opportunity as prior rules that people ought to follow, just as rule utilitarianism requires adherence to idealized universal moral rules rather than acting on the basis of utility calculations. Second, it takes the social minimum as an important limiting condition, meaning that limited utilitarianism guarantees a certain level of welfare for people and restricts extreme inequalities. A certain level of welfare and equality avoids the intolerable consequences that Rawls criticizes; on the contrary, it is acceptable. Third, it leaves considerable room for people’s choices, namely, under the constraints of the first two factors, it permits acting in accordance with the maximization of expected utility. It is precisely this special form that both foreshadows Rawls’s insight into the development of utilitarian theory and creates considerable theoretical pressure on “justice as fairness.”

考虑到功利主义的新发展，罗尔斯构造的有限功利主义可以视为规则功利主义的特殊形式。第一，它将平等的自由原则和公平的机会平等原则视为在先的规则，人们应该遵守，这就如同规则功利主义所要求的那样遵守理想的普遍化的道德规则，而不是依据功利大小而行动。第二，它将社会最低受惠值作为重要的限制条件，这意味着有限功利主义保证了人们一定程度的生活福利水平，限制了极端的不平等状况的出现。一定程度的生活福利水平和平等状态避免了罗尔斯批评的那种不可忍受的后果，相反，它是可接受的。第三，它为人们的选择留下了极大的空间，即在前两个因素的限制下，允许按照预期功利最大化行事。正是这种特殊的形式，既预示了罗尔斯对功利主义理论发展的洞见，也给“公平的正义”制造了相当大的理论压力。

Rawls’s first reason for criticizing limited utilitarianism concerns the nature of utility. Utility cannot be determined and cannot be calculated, and thus interpersonal comparisons are also impossible. If this is a theoretical problem, then “justice as fairness” faces the same issue. Limited utilitarianism is mainly intended to replace the difference principle; therefore, the real comparison is between the difference principle and the average utility principle. The principles of justice consist of three parts, and primary goods are also mainly divided into three groups. Broadly speaking, the first principle of justice corresponds to liberty and rights; the principle of fair equality of opportunity corresponds to powers and opportunities; and the difference principle primarily regulates wealth and income. Once the scope of the problem is delimited, the comparison between the difference principle and the limited utility principle becomes clear and precise: both regulate wealth and income. Undisputedly, the meaning of “utility” in limited utilitarianism at this point is clear—it is wealth and income, and these can be calculated. Rawls also uses wealth and income to determine the least advantaged, who are those “with less than half of the median income and wealth.” [13]13 John Rawls, *A Theory of Justice* (revised edition), p. 84. Rawls’s first criticism of limited utilitarianism does not hold.

罗尔斯对有限功利主义批评的第一个理由在于功利的性质。功利无法判定且无法计算，因而人际比较也不可能。如果这个问题是理论问题，那么，“公平的正义”也同样存在。有限功利主义主要是代替差别原则，因而，真正的对比是在差别原则和平均功利原则之间进行。正义原则由三部分构成，基本善也主要分为三组。大体上来说，第一个正义原则对应自由和权利；公平的机会平等原则对应权力和机会，而差别原则主要是调节财富和收入。当限定了问题的论域后，差别原则和有限功利原则的对比才清晰而明确，它们调节的对象都是财富和收入。无可争议的是，有限功利主义此时的“功利”含义是明确的，就是财富和收入，而它们可以计算。罗尔斯也是用财富和收入来确定最不利者，而最不利者是那些“所有达不到社会平均收入一半的人”。罗尔斯对有限功利主义的第一个批评不成立。

Rawls’s second reason for criticizing limited utilitarianism lies in the claim that the social minimum guaranteed by the difference principle is higher than that of the constrained average utility principle, meaning that the degree of equality achieved by the difference principle is also higher than that of the average utility principle. The purpose of limited utilitarianism in setting a social minimum is also to guarantee a certain level of welfare for everyone and to prevent severe inequality. The social minimum ensures that people do not fall into unacceptable circumstances; in this sense, the limited utility principle and the difference principle share the same overall theoretical aim. The key difference is that Rawls believes the social minimum required by limited utilitarianism is much lower than that required by the difference principle, and correspondingly, the degree of inequality is much higher than in “justice as fairness.” The question is: why can limited utilitarianism not set the social minimum as high as the difference principle does? At the very least, limited utilitarianism does not exclude a higher social minimum. Rawls provides no clear reason. Criticizing limited utilitarianism using the egalitarian requirements of the difference principle is insufficiently justified.

罗尔斯批判有限功利主义的第二个理由在于差别原则保障的社会最低受惠值比受到限制的平均功利原则高，这意味着差别原则实现的平等程度也比平均功利原则高。有限功利主义设定社会最低受惠值的目的也是保障每个人一定的福利水平，防止严重的不平等出现。社会最低受惠值确保人们不至于陷入不可接受的境况，在这种意义上，有限功利原则与差别原则总体理论旨趣一致。关键的区别在于，罗尔斯认为有限功利主义要求的社会最低受惠值要比差别原则要求的低很多，相应地，不平等的程度也比“公平的正义”高得多。问题是，有限功利主义为什么不可以将社会最低受惠者设定得同差别原则一样高呢？至少，有限功利主义并不排斥较高的社会最低受惠值。罗尔斯没有给出明确的理由。用差别原则的平等要求批判有限功利主义，理由不充分。

V. Conclusion

五、小结

Rawls’s critique of utilitarianism is an unavoidable theoretical background for anyone engaged in the study of utilitarianism. The purpose of examining Rawls’s multi-dimensional critique of utilitarianism is to understand the theoretical problems inherent in utilitarianism through theoretical critique. Most people focus their attention on the critique of classical utilitarianism, while far fewer pay attention to average utilitarianism, especially limited utilitarianism. It is precisely the limited utilitarianism constructed by Rawls that exerts considerable theoretical pressure on “justice as fairness,” drawing attention to the theoretical impact of this new utilitarianism. Viewing Rawls’s critique from the perspective of a complex version of utilitarianism, Rawls is not seeking to offer a theoretical defense of utilitarianism; rather, utilitarianism, as an important moral doctrine, has its own merits and better theoretical resources with which to respond. If “justice as fairness” is to completely replace utilitarianism, it must seek more sufficient reasons. For researchers, it is only through the theoretical confrontation formed by this mutual critique that we can treat both “justice as fairness” and utilitarianism—these two profoundly influential conceptions of justice—more fairly.

罗尔斯对功利主义的批判是人们从事功利主义研究无法回避的理论背景。考察罗尔斯对功利主义的多重维度批判，其目的是通过理论批判来认识功利主义存在的理论问题。人们大多将注意力集中在古典功利主义批判上面，而对于平均功利主义尤其是有限功利主义关注者甚少。正是罗尔斯构造的有限功利主义给“公平的正义”带来了相当大的理论压力，使得人们关注到新的功利主义带来的理论冲击。从复杂版本的功利主义理论来看待罗尔斯的批判，罗尔斯并不是要为功利主义作理论辩护，而是说，功利主义作为重要的道德学说有它自己的优点，也有更好的理论资源来回应。“公平的正义”要想完全取代功利主义，必须寻求更加充分的理由。对于研究者而言，在这种相互批判而形成的理论交锋中，我们才能够更加公平地对待“公平的正义”和功利主义这两种影响深刻的正义观。

(Author’s bio: Wang Li, Professor and Doctoral Supervisor at the School of Philosophy, Renmin University of China; Wu Yuzhang Distinguished Professor at Renmin University of China; Secretary-General of the Marxist Political Philosophy Research Association; Council Member of the Beijing Philosophy Society)

（作者简介：王立，中国人民大学哲学院教授，博士生导师，中国人民大学吴玉章特聘教授，马克思主义政治哲学研究会秘书长，北京市哲学学会理事）

(Fund project: Renmin University of China Central University Special Fund for Building World-Class Universities [Disciplines] and Characteristic Development Guidance Projects)

（基金项目：中国人民大学中央高校建设世界一流大学（学科）和特色发展引导专项资金项目）

Online editor: Tongxin

网络编辑：同心