Concern, Respect and Cooperation sets out what is standardly called a normative moral theory. I prefer to call it a substantive moral theory, because it covers both the normative and evaluative content of morality (14).11 Numbers in parentheses are page-references to Concern, Respect, and Cooperation. It is helpful to distinguish these, and one of the theory's main aims is to explain the relationship between them.

The book's central idea is very simple: that morality—thought of as comprising the ways in which a proper regard for others should be reflected in the way we respond and relate to them—consists fundamentally in three things. The first is treating others’ welfare as important: having concern for them as patients whose interests we can affect. The second is accommodating their independent self-expression: respecting them as agents with their own lives to lead. And the third is joining with others in worthwhile collective action: cooperating with them as partners.

Many theories of morality try to combine these three elements in some way, often treating one of them as primary (15-20). Broadly conceived, the welfarist tradition of theorizing about morality can be seen as offering a picture of morality that has the first of these at its core, the Kantian tradition as emphasizing the second and the contractualist tradition the third. A fourth prominent tradition—the perfectionist one—structures its theory of morality around another core idea: that morality consists in those forms of excellence in interacting with each other through which we flourish as human beings (18-19). My approach is instead to treat the first three of these elements as independent foundations in a plural foundation theory of the type pioneered by W.D. Ross. These, as I see it, are three independent sources from which we derive morality's normative content—the normative reasons it gives us that bear on what we do, feel, think and say in response to each other. The forms of moral excellence emphasized by the fourth tradition can then be explained as excellent forms of responsiveness to the reasons deriving from these three sources—supplying the aretaic part of morality's evaluative content (151).

The sense in which I claim these are foundations for morality is this. Some parts of the normative and evaluative content of morality derive their moral significance from others. For example, negligence is wrong because it is a failure of due regard for others’ welfare; one reason for being tolerant is that it enables cooperation; and self-reliance is good (at least in part) because it allows others to lead their own lives instead of having to look after you—so it derives its value (in part) from the value of respect. Morality's foundations are the parts of its content that are not derived from other parts of its content (21). In value theory, one of the things that can be meant by saying that something's value is “intrinsic” is that it does not derive its value from the value of anything else (257 n. 23). The claim that morality has foundations is a generalization of the claim that some moral values are intrinsic in that sense: a generalization, because it applies to the normative as well as the evaluative parts of morality's content. Foundations are not stopping points for “Why?” questions. It can still make sense to ask “Why is this a foundational part of the content of morality?” (162-4). But they are stopping points for relations of derivation of one part of the content of morality from another.

One of the attractions of Ross's approach is that it gives us a compositional model for moral theorizing (3). It allows that the derived content of morality is inexhaustibly complex, but explains this as emerging from a relatively simple set of underlying elements and compositional principles, with the complexity being generated by their repeated combination (168-9). A Ross-style theory builds its explanations of the content of morality from three main sorts of ingredients: the foundations it attributes to morality, the principles of derivation through which morality's derived content is generated from its foundations, and the principles of interaction through which the various components of its derived content interact with each other to determine the overall moral quality of the various objects of moral assessment: actions, attitudes, persons, states of affairs, and so on (4-5). In Ross's own theory, the foundational elements are prima facie duties (which, as now standardly interpreted, correspond to pro tanto reasons for action); other prima facie duties are derived by subsuming them as instances of one or more of the foundational ones; and prima facie duties (whether basic or derived) interact with each other by having different strengths or weights, the overall balance of which determines the overall moral quality of one's act—one's ‘duty proper’ (5).

What I offer has, very broadly speaking, the same sort of structure, but with three large differences. Morality's foundations are not prima facie duties; there are other forms of derivation apart from subsumption; and the interaction of the derived elements is not confined to the balancing of comparative weights, but also includes forms of undermining, through which one part of the content of morality can deprive another of the reason-giving force it would otherwise have had (5). Of the book's three parts, the first is devoted to setting out my view of the foundations of morality: its task is to explain what kinds of things these are, if not prima facie duties, and what exactly is their content. Part II explains the forms of derivation and interaction through which the content of morality derives from these foundations. Part III then offers three extended applications of the theory.

But acting to promote others’ welfare is only one of a variety of morally important responses to it; what can be subsumptively derived from this is limited to other prima facie duties, and does not cover the further content of morality, such as its evaluative content; and it also invites counterexamples (33-40). If, as it appears, it can be in a bad person's interests to get bad things, then there can be parts of a person's welfare that there is no prima facie duty to promote (39).

The relationships that this principle reports I call “norms of presumptive fitness”.

Here, “self-expression” is a term of art covering all those forms of conduct that are attributable to a person. It includes actions that are the product of autonomous choice but extends more broadly: we can have reasons to respect what someone else is doing that do not require that she has autonomously chosen to do it (48-9). These reasons, I claim, do not derive from another part of the content of morality—in particular, they do not derive from concern for her welfare. The presumptive fitness of respecting a person's self-expression is morally fundamental, alongside the presumptive fitness of promoting her welfare. But again, there are exceptions: not every form of self-expression is respectworthy (51-2).

When a collective action is merely pointless or incompetent, it is sufficient to say that it is not worthwhile, and therefore does not call for the responses in (C+). But when it is directed towards ends that are morally bad, it calls for the opposite responses.

Vindicating these claims—the principal claims in Part I—is a matter of demonstrating the explanatory adequacy of a theory with these foundations, showing that it can make sense of the full content of morality. On the face of it, these four principles may seem an unpromising basis from which to work. Can the full content and nuance of morality be derived from just four principles? And how can exception-hedged principles like (W) and (S) really be a foundation from which the rest of the content of morality can be derived? To apply principles like these, we need to be told when their exception-conditions are and are not triggered, and whatever determines that must it seems occupy a more fundamental level within the theory (42).

The overall aim of Part II is to supply the materials with which to address those two worries. In response to the first, I agree that the full content of morality cannot be derived from these four foundational principles, if derivation is limited to the subsumptive form recognized by Ross. But derivation can take other forms: there are further kinds of subsumption (70-71), as well as non-subsumptive derivation-relations of enabling (72-4) and responsiveness (74-9). Derivations of these different kinds can then combine in simple and more complex ways to generate various familiar parts of the non-foundational content of morality. Rights and justice, on this picture, are prominent parts of morality, but they are not fundamental: their content is determined by reasons deriving from all of the foundations of morality. Our moral rights can be thought of as the moral standings we possess in virtue of a morally bounded allocation of freedom—where the “moral boundaries” consist in permissions of enforcement and/or requirements of redress (82-6). Justice is accommodated as the quality that a social structure has when it is regulated by an authority as it morally should be (86-8).

The other main departure from Ross is my account of how moral norms (both fundamental and derived) interact with each other. This extends beyond simply providing considerations of different weights that must be balanced against each other, to include relationships of undermining through in which one moral consideration deprives another of the weight it would otherwise have had.

The book's central chapters examine two main forms that undermining can take (again distinguishing subvarieties of each). The first is “content-undermining” (89). This is my name for a class of cases from which examples are often taken to support holism in the theory of practical reasons: examples such as sadistic pleasure. Pleasure, as a constituent of welfare, presumptively calls for promotion; but when its content is misdirected towards others’ suffering, the fitness of promoting it is undermined (92-4). Redeploying some ideas from Brentano's value-theory, I argue that a theory with exception-hedged principles at its foundations can explain this (98-103). We can explain the misdirection of the sadist's pleasure by first invoking an application of the principle (W) to the suffering-states that are the object of his pleasure, and can draw on this to explain how the exception-clause in (W) is satisfied when applied to his own sadistic pleasure-state (92-8). It is in this way that the second worry about the structure of the theory—the worry that exception-hedged principles cannot be foundational—is addressed. A Brentano-style theory containing exception-hedged principles can explain attributions of value without needing to be supplemented by any further more fundamental principles, and without any circularity (103-5).

This puts us in a position to explain when it is that we can derive norms of fitness from the presumptive fitness-relationships that lie at the foundations of morality—namely, when there is no content-undermining (108-9). However, to derive normative reasons from norms of fitness, several further conditions must be satisfied. In order to have a reason to respond to others’ welfare with concern, or their self-expression with respect, you must have the capacity to make the response. You must also satisfy a condition of personal relevance: the expression of condolences is a fit response to bereavement, but this is not enough to give you a reason to write a condolence letter to a bereaved person you have never met (35). And a further condition comes from what I call “context-undermining”. An example is this: if you are acting as a trustee for some third party, the interests of your relatives are disqualified from being reasons for you to use the fund for their benefit (111). This is not because their interests have the wrong content to be fitly promoted, like those of the sadist; it is because they fail to qualify as reasons in this context. Here, the context is one in which you are exercising agency on behalf of the beneficiary of the trust, and the duties that attach to this role constrain the legitimate exercise of your agency, restricting the considerations that can qualify as reasons for and against the actions you perform. This “agency-based” kind of context-undermining can be distinguished from other, “domain-based” and “meaning-based”, kinds (115-19).

What results is a theory with the following overall structure (153-4). Morality's foundations are the norms of presumptive fitness (W), (S), (C+) and (C-). From these, norms of fitness are derived (when there is no content-undermining). And from norms of fitness, reasons derive (when there is no context-undermining, and conditions of capacity and personal relevance are satisfied). Deontic judgements—judgements about rightness and wrongness, and what ought all things considered to be done—amount to verdicts about the overall bearing of reasons on our responses. Then, finally, aretaic judgements can be understood as verdicts about the quality of our responsiveness to those reasons—allowing us to generate a helpful taxonomy of moral virtues, reflecting the relationships between the reasons to which the different virtues respond (149-51).

If a substantive moral theory such as this is to be convincing, it must give an account of the content of morality that is not only recognizable as corresponding to what we do, on reflection, see as morally important, but can also explain why it should matter to us as much as, and in the way that, it does (2). In this spirit, Part II finishes by explaining how, through following norms of concern, respect, and cooperation, we can create and sustain three valuable kinds of interpersonal relationship—three valuable ways of connecting ourselves to each other (160-61). Through concern for each other, we forge a relationship in which the import of others’ welfare is shared: it matters not just to the other person but to you too. Through respecting each other's self-expression, we share a recognition of the dignity we each have as independent recognizers of reasons who form our own decisions and opinions, and the equal status that this confers on us. And through cooperative relationships, we share our agency, and our authorship of the actions we perform together. This, I suggest, supplies a non-derivational justification for taking morality seriously: a justification of its importance that does not derive that importance from a deeper set of normative foundations (162-4).

Finally, Part III offers three extended applications of the theory—discussing three contentious types of action for which reasons of concern, respect and cooperation seem to pull in different directions. These are, respectively, actions of paternalism, in which a person's self-expression is restricted for her own benefit (174); actions of using one person as a means of benefiting another (195); and the actions we take as consumers, participating in markets that have bad as well as good effects—which I see as an important application of reasons of cooperation (217). The emphasis in these three chapters is on showing how the materials provided earlier in the book allow us to say something more illuminating than simply that we must balance reasons from these three sources against each other. A Rossian theory does not give you principles telling you what verdicts to reach in particular cases; but I try to show that it can still give us useful guidance, by clarifying which considerations are important and how to think about the relationship between them (173).

As will be clear from this description, Concern, Respect, and Cooperation is an ambitious book. In trying to present an overall moral theory within a book of readable length, it attempts to cover a lot of ground in a short space, and there is much room for supplying further detail. Whether that is worth pursuing depends on the strength of the overall structure set out so far. To assess that, some serious stress-testing is called for: I am grateful to my fellow symposiasts for their willingness to take this on.

关注、尊重和合作提出了标准上所谓的规范道德理论。我更愿意称其为实质性道德理论，因为它涵盖了道德的规范和评价内容（14）.11括号中的数字是对关注、尊重和合作的页面引用。区分这些是有帮助的，该理论的主要目的之一是解释它们之间的关系。

这本书的中心思想非常简单：道德——被认为包括对他人的适当尊重应该反映在我们对他人的回应和联系方式中的方式——基本上由三件事组成。首先是将他人的福利视为重要：将他们视为我们可以影响其利益的患者。第二是适应他们独立的自我表达：尊重他们作为有自己的生活的代理人。第三个是与他人一起进行有价值的集体行动：作为伙伴与他们合作。

许多道德理论试图以某种方式将这三个要素结合起来，通常将其中之一视为主要的（15-20）。从广义上讲，福利主义的道德理论化传统可以被视为提供了一幅以第一个为核心的道德图景，康德主义传统强调第二个，契约主义传统是第三个。第四个突出的传统——完美主义——围绕另一个核心思想构建其道德理论：道德在于那些我们作为人类蓬勃发展的相互作用中的卓越形式（18-19）。相反，我的方法是将这些元素中的前三个元素视为 WD Ross 开创的那种类型的多元基础理论中的独立基础。这些，在我看来，是我们从中获得道德规范内容的三个独立来源——道德规范给我们的规范理由，这些理由与我们相互回应的行为、感受、思想和言论有关。然后，第四传统所强调的道德卓越形式可以解释为对源自这三个来源的原因的出色回应形式——提供道德评价内容的戏剧部分（151）。

我声称这些是基础的意义因为道德就是这样。道德规范和评价内容的某些部分从其他部分获得其道德意义。例如，疏忽是错误的，因为它没有适当考虑他人的福利；宽容的一个原因是它可以促成合作；自力更生是好的（至少部分地），因为它允许他人过自己的生活，而不必照顾你——所以它的价值（部分地）来自于尊重的价值。道德的基础是其内容的一部分，而不是从其内容的其他部分派生出来的（21）。在价值理论中，说某物的价值是“内在”的，其中一件事是它不会从其他任何东西的价值中获得价值（257 n. 23）。道德有基础的主张是对某些道德价值在这个意义上是内在的主张的概括：概括，因为它适用于道德内容的规范部分和评估部分。基金会不是“为什么？”的终点。问题。仍然可以问“为什么这是道德内容的基础部分？” (162-4)。但它们是道德内容的一部分从另一部分派生的关系的停止点。

罗斯方法的魅力之一是它为我们提供了一个道德理论化的组合模型（3）。它允许衍生的道德内容是无穷无尽的复杂，但将其解释为来自一组相对简单的基本元素和构成原则，而复杂性是由它们的重复组合产生的（168-9）。罗斯风格的理论从三种主要成分构建其对道德内容的解释：它赋予道德的基础，道德的派生内容从其基础产生的派生原则，以及相互作用的原则，其派生内容的各个组成部分相互影响，以确定各种道德评估对象的整体道德质量：行为、态度、人、事态等（4-5）。在罗斯自己的理论中，基本要素是表面上的义务（按照现在的标准解释，对应于行动的原由）；其他表面上的义务是通过将它们归入一个或多个基本义务的实例而得出的；表面上的义务（无论是基本义务还是派生义务）通过具有不同的优势或权重而相互作用，其整体平衡决定了一个人行为的整体道德品质——一个人的“适当义务”（5）。

从广义上讲，我提供的内容具有相同的结构，但有三个很大的不同。道德的基础不是表面上的义务；除了包含之外，还有其他形式的推导；派生要素的相互作用不仅限于比较权重的平衡，还包括破坏形式，通过破坏形式，道德内容的一部分可以剥夺另一部分本来应该具有的推理力（5） . 在这本书的三个部分中，第一部分致力于阐述我对道德基础的看法：它的任务是解释这些东西是什么类型的东西，如果不是表面上的义务，它们的内容到底是什么。第二部分解释了道德内容从这些基础中衍生出来的衍生和相互作用的形式。

但采取行动促进他人的福利只是对它的多种道德上重要的回应之一。可以由此推导出的内容仅限于其他表面上的义务，不包括道德的进一步内容，例如其评价内容；它还引入了反例（33-40）。如果看起来，得到坏东西可能符合坏人的利益，那么个人福利的某些部分可能没有表面上明显的义务来促进（39）。

我将这一原则所报告的关系称为“假定适用性规范”。

在这里，“自我表达”是一个涵盖所有可归因于一个人的行为形式的艺术术语。它包括作为自主选择的产物但更广泛的行为：我们可以有理由尊重其他人正在做的事情，而不需要她自主选择去做（48-9）。我声称，这些理由并非源于道德内容的另一部分——特别是，它们并非源于对她福利的关注。尊重一个人的自我表达的假定适用性在道德上是基本的，同时促进她的福利的假定适用性也是如此。但同样，也有例外：并非每种形式的自我表达都是值得尊重的（51-2）。

当集体行动只是无意义或无能时，说它不值得就足够了，因此不需要（C+）中的回应。但是，当它指向道德上不好的目的时，它会要求相反的反应。

证明这些主张——第一部分的主要主张——是证明具有这些基础的理论的解释充分性的问题，表明它可以理解道德的全部内容。从表面上看，这四项原则似乎是一个没有希望的工作基础。道德的全部内容和细微差别能否仅来自四个原则？像 (W) 和 (S) 这样的例外对冲原则如何才能真正成为衍生道德其他内容的基础？为了应用这样的原则，我们需要被告知它们的异常条件何时被触发和未被触发，以及任何决定它似乎必须在理论中占据更基本水平的东西（42）。

第二部分的总体目标是提供解决这两个问题的材料。针对第一个问题，我同意，如果推导仅限于罗斯所承认的概括形式，道德的全部内容不能从这四个基本原则推导出来。但是推导可以采取其他形式：还有更多种类的包含（70-71），以及启用（72-4）和响应（74-9）的非包含推导关系。然后，这些不同种类的派生可以以简单和更复杂的方式组合起来，生成道德的非基础内容的各种熟悉部分。在这张图上，权利和正义是道德的突出部分，但它们不是根本的：它们的内容是由源自所有道德基础的原因决定的。我们的道德权利可以被认为是我们凭借对自由的道德约束分配而拥有的道德地位——其中“道德边界”包括执法许可和/或补救要求（82-6）。正义被容纳为社会结构在道德上应该由权威监管时所具有的品质（86-8）。

与罗斯的另一个主要区别是我对道德规范（基本的和衍生的）如何相互作用的描述。这不仅限于提供必须相互平衡的不同权重的考虑，还包括破坏关系，在这种关系中，一种道德考虑剥夺了另一种道德考虑本来应该具有的权重。

本书的中心章节研究了破坏可以采取的两种主要形式（再次区分每种形式的子变体）。第一个是“内容破坏”（89）。这是我为一类案例命名的，在实践原因理论中，这些案例经常被用来支持整体论：比如虐待狂快感。快乐，作为福利的组成部分，推定要求提升；但是，当它的内容被误导为他人的痛苦时，宣传它的适宜性就会受到破坏（92-4）。重新部署布伦塔诺价值理论中的一些观点，我认为以异常对冲原则为基础的理论可以解释这一点（98-103）。我们可以通过首先将原理（W）应用于作为他快乐对象的痛苦状态来解释施虐者快乐的误导，并且可以以此来解释（W）中的例外条款在应用于他自己的虐待狂快乐状态时如何得到满足（92-8）。正是通过这种方式解决了对理论结构的第二个担忧——即异常对冲原则不能成为基础的担忧——得到了解决。包含异常对冲原则的布伦塔诺式理论可以解释价值属性，无需任何更基本的原则补充，也没有任何循环性 (103-5)。

这使我们能够解释什么时候我们可以从位于道德基础的假定的适合度关系中推导出适合度的规范——即，当没有内容破坏时（108-9）。然而，要从适应度规范中推导出规范性理由，还必须满足几个进一步的条件。为了有理由关切地回应他人的福利，或尊重他们的自我表达，你必须有能力做出回应。您还必须满足一个与个人相关的条件：表示哀悼是对丧亲之痛的恰当回应，但这还不足以给您给一个你从未见过的失去亲人的人写一封吊唁信的理由 (35)。另一个条件来自我所说的“破坏语境”。一个例子是：如果您是某个第三方的受托人，那么您亲属的利益就不能成为您为他们的利益使用基金的理由 (111)。这并不是因为他们的利益有错误的内容可以适当地促进，就像虐待狂一样；这是因为在这种情况下，它们不能作为理由。在这里，背景是您代表信托受益人行使代理权，而与此角色相关的职责限制了您的代理权的合法行使，限制了可以作为支持和反对行动的理由的考虑因素你表演。

结果是具有以下总体结构的理论（153-4）。道德的基础是推定适合度（W）、（S）、（C+）和（C-）的规范。从这些中导出适合度规范（当没有内容破坏时）。并且从适合的规范中得出原因（当没有破坏背景，并且满足能力和个人相关性的条件时）。道义判断——关于对与错的判断，以及所有事情都应该做什么——相当于对理由对我们的反应的总体影响的判断。然后，最后，阿里泰式判断可以被理解为对我们对这些原因的反应质量的判断——使我们能够生成一个有用的道德美德分类法，反映不同美德所回应的原因之间的关系（149-51）。

如果像这样的实质性道德理论要令人信服，它必须说明道德的内容，不仅可以识别为对应于我们所做的事情，经过反思，认为在道德上很重要，而且还可以解释为什么它应该对我们来说，它和它一样重要，并且以它的方式（2）。本着这种精神，第二部分最后解释了通过遵循关注、尊重和合作的规范，我们如何能够创造和维持三种有价值的人际关系——三种将我们彼此联系起来的有价值的方式（160-61）。通过相互关心，我们建立了一种关系，在这种关系中，他人的福利是共享​​的：这不仅对对方很重要，对你也很重要。通过尊重彼此的自我表达，我们都承认我们每个人都拥有独立承认形成我们自己的决定和意见的理由的尊严，以及这赋予我们的平等地位。通过合作关系，我们分享我们的代理权，以及我们共同执行的行动的作者身份。我建议，这为认真对待道德提供了一个非衍生的理由：它的重要性的理由不从更深层次的规范基础中得出这种重要性（162-4）。

最后，第三部分提供了该理论的三个扩展应用——讨论了三种有争议的行动类型，关注、尊重和合作的原因似乎朝着不同的方向发展。这些分别是家长式的行为，其中一个人的自我表达被限制为自己的利益（174）；将一个人用作使另一个人受益的手段的行为（195）；以及我们作为消费者所采取的行动，参与有坏影响的市场，也有好的影响——我认为这是合作理由的一个重要应用（217）。这三章的重点是展示本书前面提供的材料如何让我们说一些更有启发性的东西，而不仅仅是我们必须平衡这三个来源的理由。罗西理论并没有给你原则告诉你在特定情况下要达到什么判决；但我试图通过阐明哪些考虑因素是重要的以及如何考虑它们之间的关系来表明它仍然可以为我们提供有用的指导（173）。

从这个描述中可以清楚地看出，《关注、尊重和合作》是一本雄心勃勃的书。试图在一本通俗易懂的书中呈现一个整体的道德理论，它试图在很短的篇幅中涵盖很多领域，并且有很大的空间可以提供更多的细节。这是否值得追求，取决于目前所制定的整体结构的强度。为了评估这一点，需要进行一些严肃的压力测试：我感谢我的研讨会同事愿意承担这一点。