Douglas W. Portmore

This paper concerns Warren Quinn’s famous “The Puzzle of the Self-Torturer.” I argue that even if we accept his assumption that practical rationality is purely instrumental such that what he ought to do is simply a function of how the relevant options compare to each other in terms of satisfying his actual preferences that doesn’t mean that every explanation as to why he shouldn’t advance to the next level must appeal to the idea that so advancing would be suboptimal in terms of the satisfaction of his actual preferences. Rather, we can admit that his advancing would always be optimal, but argue that advancing isn’t always what he ought to do given that advancing sometimes fails to meet some necessary condition for being what he ought to do. For instance, something can be what he ought to do only if it’s an option for him. What’s more, something can be what he ought to do only if it’s something that he can do without responding inappropriately to his reasons—or, so, I argue. Thus, the solution to the puzzle is, I argue, to realize that, in certain circumstances, advancing is not what the self-torturer ought to do given that he can do so only by responding inappropriately to his reasons.

Many philosophers hold that the achievement of one's goals can contribute to one's welfare apart from whatever independent contributions that the objects of those goals or the processes by which they are achieved make. Call this the Achievement View, and call those who accept it achievementists. In this paper, I argue that achievementists should accept both (a) that one factor that affects how much the achievement of a goal contributes to one’s welfare is the amount that one has invested in that goal and (b) that the amount that one has invested in a goal is a function of how much one has personally sacrificed for its sake, not a function of how much effort one has put into achieving it. So I will, contrary to at least one achievementist (Keller), be arguing against the view that the greater the amount of productive effort that goes into achieving a goal, the more its achievement contributes to one's welfare. Furthermore, I argue that the reason that the achievement of those goals for which one has personally sacrificed matters more to one’s welfare is that, in general, the redemption of one's self-sacrifices in itself contributes to one’s welfare. Lastly, I argue that the view that the redemption of one's self-sacrifices in itself contributes to one's welfare is plausible independent of whether or not we find the Achievement View plausible. We should accept this view so as to account both for the Shape of a Life Phenomenon and for the rationality of honoring "sunk costs."

In this paper, I take it for granted both that there are two types of blameworthiness—accountability blameworthiness and attributability blameworthiness—and that avoidability is necessary only for the former. My task, then, is to explain why avoidability is necessary for accountability blameworthiness but not for attributability blameworthiness. I argue that what explains this is both the fact that these two types of blameworthiness make different sorts of reactive attitudes fitting and that only one of these two types of attitudes requires having been able to refrain from φ-ing in order for them to be fitting.

On what I take to be the standard account of supererogation, an act is supererogatory if and only if it is morally optional and there is more moral reason to perform it than to perform some permissible alternative. And, on this account, an agent has more moral reason to perform one act than to perform another if and only if she morally ought to prefer how things would be if she were to perform the one to how things would be if she were to perform the other. I argue that this account has two serious problems. The first, which I call the latitude problem, is that it has counterintuitive implications in cases where the duty to be exceeded is one that allows for significant latitude in how to comply with it. The second, which I call the transitivity problem, is that it runs afoul of the plausible idea that the one-reason-morally-justifies-acting-against-another relation is transitive. What’s more, I argue that both problems can be overcome by an alternative account, which I call the maximalist account.

It is through our actions that we affect the way the world goes. Whenever we face a choice of what to do, we also face a choice of which of various possible worlds to actualize. Moreover, whenever we act intentionally, we act with the aim of making the world go a certain way. It is only natural, then, to suppose that an agent's reasons for action are a function of her reasons for preferring some of these possible worlds to others, such that what she has most reason to do is to bring about the possible world which, of all those available to her, is the one that she has most reason to want to obtain. This is what is known as the `teleological conception of practical reasons'. Whether this is the correct conception of practical reasons is important not only in its own right, but also in virtue of its potential implications for what sort of moral theory we should accept. Below, I argue that the teleological conception is indeed the correct conception of practical reasons

Maximalism is the view that an agent is permitted to perform a certain type of action if and only if she is permitted to perform some instance of this type, where φ-ing is an instance of ψ-ing if and only if φ-ing entails ψ-ing but not vice versa. Now, the aim of this paper is not to defend maximalism, but to defend a certain account of our options that when combined with maximalism results in a theory that accommodates the idea that a moral theory ought to be morally harmonious—that is, ought to be such that the agents who satisfy the theory, whoever and however numerous they may be, are guaranteed to produce the morally best world that they have the option of producing. I argue that, for something to count as an option for an agent, it must, in the relevant sense, be under her control. And I argue that the relevant sort of control is the sort that we exercise over our reasons-responsive attitudes by being both receptive and reactive to reasons. I call this sort of control rational control, and I call the view that φ-ing is an option for a subject if and only if she has rational control over whether she φs rationalism. When we combine this view with maximalism, we get rationalist maximalism, which I argue is a promising moral theory.

It seems that we can be directly accountable for our reasons-responsive attitudes—e.g., our beliefs, desires, and intentions. Yet, we rarely, if ever, have volitional control over such attitudes, volitional control being the sort of control that we exert over our intentional actions. This presents a trilemma: (Horn 1) deny that we can be directly accountable for our reasons-responsive attitudes, (Horn 2) deny that φ’s being under our control is necessary for our being directly accountable for φ-ing, or (Horn 3) deny that the relevant sort of control is volitional control. This paper argues that we should take Horn 3.

In this dissertation, I argue that commonsense morality is best understood as an agent-relative consequentialist theory, that is, as a theory according to which agents ought always to bring about what is, from their own individual perspective, the best available state of affairs. I argue that the agent-relative consequentialist can provide the most plausible explanation for why it is wrong to commit a rights violation even in order to prevent a number of other agents from committing comparable rights violations: agents bear a special responsibility for their own actions and consequently a state of affairs where an agent has herself committed a rights violation is worse, from her perspective, than a state of affairs where a number of other agents have committed comparable rights violations. I also argue that agent-relative consequentialism can accommodate moral options despite being a theory which requires agents always to do the best they can

The book concerns what I take to be the least controversial normative principle concerning action: you ought to perform your best option—best, that is, in terms of whatever ultimately matters. The book sets aside the question of what ultimately matters so as to focus on more basic issues, such as: What are our options? Do I have the option of typing out the cure for cancer if that’s what I would in fact do if I had the right intentions at the right times (e.g., the intention to type the letter T at t1, the intention to type the letter H at t2, the intention to type the letter E at t3, etc.)? If I can’t form intentions voluntarily, does that mean that I don’t have the option of forming the intention that I ought to form? Which options do we assess directly in terms of their own goodness and which do we assess in terms of their relations to the goodness of other options? What do we hold fixed when assessing how good an option is? Do we, for instance, hold fixed the agent’s future beliefs, desires, and intentions? And do we hold fixed the agent’s predictable future misbehavior? Lastly, how do the things that ultimately matter determine the goodness of an option? If one of the things that ultimately matters in determining the goodness of an option is that the option doesn’t involve violating anyone’s rights, do we evaluate the option itself in terms of whether it involves violating anyone’s rights or do we evaluate the option’s prospect in terms of this and then the option in terms of its prospect? And what if there is indeterminacy or uncertainty with regards to whether an option would involve violating someone’s rights?

This is Chapter 5 of my Commonsense Consequentialism: Wherein Morality Meets Rationality. In this chapter, I argue that those who wish to accommodate typical instances of supererogation and agent-centered options must deny that moral reasons are morally overriding and accept both that the reason that agents have to promote their own self-interest is a non-moral reason and that this reason can, and sometimes does, prevent the moral reason that they have to sacrifice their self-interest so as to do more to promote the interests of others from generating a moral requirement. Furthermore, I argue that given that an act’s deontic status of both moral and non-moral reasons, the consequentialist must adopt dual-ranking act-consequentialism. I then defend dual-ranking act-consequentialism against a number of objections.

We ought to perform our best option—that is, the option that we have most reason, all things considered, to perform. This is perhaps the most fundamental and least controversial of all normative principles concerning action. Yet, it is not, I believe, well understood. For even setting aside questions about what our options are and what our reasons are, there are prior questions concerning how best to formulate the principle. In this paper, I address these questions. One of the more interesting upshots of this inquiry is that the deontic statuses (e.g., obligatory, optional, and impermissible) of individual actions are determined by the deontic statuses of the larger sets of actions of which they are a part. And, as I show, this has a number of interesting implications both for normative theory and for our understanding of practical reasons.

In this paper, I argue that we have obligations not only to perform certain actions, but also to have certain attitudes (such as desires, beliefs, and intentions), and this despite the fact that we rarely, if ever, have direct voluntary control over our attitudes. Moreover, I argue that whatever obligations we have with respect to actions derive from our obligations with respect to attitudes. More specifically, I argue that an agent is obligated to perform an action if and only if it’s the action that she would perform if she were to have the attitudes that she ought to have. This view, which I call attitudism, has three important implications. First, it implies that an adequate practical theory must not be exclusively act-orientated. That is, it must require more of us than just the performance of certain voluntary acts. Second, it implies that an adequate practical theory must be attitude-dependent. That is, it must hold that what we ought to do depends on what attitudes we ought to have. Third, it implies that no adequate practical theory can require us to perform acts that we would not perform even if we were to have the attitudes that we ought to have. I then show how these implications can help us both to address certain puzzling cases of rational choice and to understand why most typical practical theories (utilitarianism, rational egoism, virtue ethics, Rossian deontology, etc.) are mistaken.

Consequentialism is an agent-neutral teleological theory, and deontology is an agent-relative non-teleological theory. I argue that a certain hybrid of the two—namely, non-egoistic agent-relative teleological ethics (NATE)—is quite promising. This hybrid takes what is best from both consequentialism and deontology while leaving behind the problems associated with each. Like consequentialism and unlike deontology, NATE can accommodate the compelling idea that it is always permissible to bring about the best available state of affairs. Yet unlike consequentialism and like deontology, NATE accords well with our commonsense moral intuitions.

In this paper, I argue that maximizing act-consequentialism (MAC)—the theory that holds that agents ought always to act so as to produce the best available state of affairs—can accommodate both agent-centered options and supererogatory acts. Thus I will show that MAC can accommodate the view that agents often have the moral option of either pursuing their own personal interests or sacrificing those interests for the sake of the impersonal good. And I will show that MAC can accommodate the idea that certain acts are supererogatory in the sense of not being morally required even though they are what the agent has most moral reason to do. These two theses are surprising in themselves, but even more surprising is how I arrive at them. I argue that anyone generally concerned to accommodate, in some coherent fashion, our pre-theoretical moral intuitions at both the normative and meta-ethical levels will have to give a certain account of agent-centered options and supererogatory acts and that this account is the very one that allows for the maximizing act-consequentialist to accommodate both. So my paper will not only be of interest to those concerned with the tenability of consequentialism, but also to anyone interested in giving a coherent account of our pre-theoretical moral intuitions.

The performance of one option can entail the performance of another. For instance, baking an apple pie entails baking a pie. Now, suppose that both of these options—baking a pie and baking an apple pie—are permissible. This raises the issue of which, if either, is more fundamental than the other. Is baking a pie permissible because it’s permissible to bake an apple pie? Or is baking an apple pie permissible because it’s permissible to bake a pie? Or are they equally fundamental, as they would be if they were both permissible because, say, they both accord with Kant’s categorical imperative? I defend the view that the permissibility of an option that entails another is more fundamental than the permissibility of the option that it entails. That is, I defend maximalism: the view that if an agent is permitted to perform a certain type of action (say, baking a pie), this is in virtue of the fact that she is permitted to perform some instance of this type (say, baking an apple pie), where φ-ing is an instance of ψ-ing if and only if φ-ing entails ψ-ing but not vice versa. If maximalism is correct, then, as I show, most theories of morality and rationality must be revised.

Following Shelly Kagan’s useful terminology, foundational consequentialists are those who hold that the ranking of outcomes is at the foundation of all moral assessment. That is, they hold that moral assessments of right and wrong, virtuous and vicious, morally good and morally bad, etc. are all ultimately a function of how outcomes rank. But foundational consequentialists disagree on what is to be directly evaluated in terms of the ranking of outcomes, which is to say that they disagree on what the primary evaluative focal point is. Act-consequentialists take acts to be the primary evaluative focal point. They evaluate acts in terms of how their outcomes rank (the higher ranked the outcome, the morally better the act), but evaluate everything else in terms of the morally best acts. Thus, the morally best rules are those that would, if internalized, most reliably lead us to perform the morally best acts. Rule-consequentialists, by contrast, take rules to be the primary evaluative focal point. They evaluate rules according to how their outcomes rank and then assess everything else in terms of the morally best rules. Thus, the morally best acts are those that conform to the morally best rules. In this paper, I argue that foundational consequentialists should not take the primary evaluative focal point (or points) to be acts, rules, virtues, or even everything. In so doing, I argue against act-consequentialism, rule-consequentialism, and global consequentialism. But my project is not entirely negative, for I argue that the primary evaluative focal point should be a complex of acts and attitudes. In the end, then, I claim that foundational consequentialists should accept a new kind of consequentialism, which I call attitude-consequentialism.

On commonsense morality, there are two types of situations where an agent is not required to maximize the impersonal good. First, there are those situations where the agent is prohibited from doing so--constraints. Second, there are those situations where the agent is permitted to do so but also has the option of doing something else--options. I argue that there are three possible explanations for the absence of a moral requirement to maximize the impersonal good and that the commonsense moralist must appeal to all three in order to account for the vast array of constraints and options we take there be.

In this paper, I present an argument that poses the following dilemma for moral theorists: either (a) reject at least one of three of our most firmly held moral convictions or (b) reject the view that moral reasons are morally overriding, that is, reject the view that moral reasons override non-moral reasons such that even the weakest moral reason defeats the strongest non-moral reason in determining an act’s moral status (e.g., morally permissible). I then argue that we should opt for the second horn of this dilemma, in part because we should be loath to reject such firmly held moral convictions, but also because doing so allows us to dissolve an apparent paradox regarding supererogation. If I’m right, if non-moral reasons are relevant to determining what is and isn’t morally permissible, then it would seem that moral theorists have their work cut out for them. Not only will they need to determine what the fundamental right-making and wrong-making features of actions are (i.e., what moral reasons there are), but they will also need to determine what non-moral reasons there are and which of these are relevant to determining an act’s deontic status. And moral theorists will have to account for how these two very different sorts of reasons—moral and non-moral reasons—”come together” to determine an act’s deontic status. I will not attempt to do this work here, but rather only to argue that the work needs to be done.

I argue that when determining whether an agent ought to perform an act, we should not hold fixed the fact that she’s going to form certain attitudes (and, here, I’m concerned with only reasons-responsive attitudes such as beliefs, desires, and intentions). For, as I argue, agents have, in the relevant sense, just as much control over which attitudes they form as which acts they perform. This is important because what effect an act will have on the world depends not only on which acts the agent will simultaneously and subsequently perform, but also on which attitudes she will simultaneously and subsequently form. And this all leads me to adopt a new type of practical theory, which I call rational possibilism. On this theory, we first evaluate the entire set of things over which the agent exerts control, where this includes the formation of certain attitudes as well as the performance of certain acts. And, then, we evaluate individual acts as being permissible if and only if, and because, there is such a set that is itself permissible and that includes that act as a proper part. Importantly, this theory has two unusual features. First, it is not exclusively act-orientated, for it requires more from us than just the performance of certain voluntary acts. It requires, in addition, that we involuntarily form certain attitudes. Second, it is attitude-dependent in that it holds that which acts we’re required to perform depends on which attitudes we’re required to form. I then show how these two features can help us both to address certain puzzling cases of rational choice and to understand why most typical practical theories (utilitarianism, virtue ethics, rational egoism, Rossian deontology, etc.) are problematic.

Dual-ranking act-consequentialism (DRAC) is a rather peculiar version of act-consequentialism. Unlike more traditional forms of act-consequentialism, DRAC doesn’t take the deontic status of an action to be a function of some evaluative ranking of outcomes. Rather, it takes the deontic status of an action to be a function of some non-evaluative ranking that is in turn a function of two auxiliary rankings that are evaluative. I argue that DRAC is promising in that it can accommodate certain features of commonsense morality that no single-ranking version of act-consequentialism can: supererogation, agent-centered options, and the self-other asymmetry. I also defend DRAC against three objections: (1) that its dual-ranking structure is ad hoc, (2) that it denies (putatively implausibly) that it is always permissible to make self-sacrifices that don’t make things worse for others, and (3) that it violates certain axioms of expected utility theory, viz., transitivity and independence.