This paper investigates the claim that artificial Intelligence Systems cannot be held morally responsible because they do not have an ability for agential self-awareness e.g. they cannot be aware that they are the agents of an action. The main suggestion is that if agential self-awareness and related first person representations presuppose an awareness of a self, the possibility of responsible artificial intelligence systems cannot be evaluated independently of research conducted on the nature of the self. Focusing on a specific account (...) of the self from the phenomenological tradition, this paper suggests that a minimal necessary condition that artificial intelligence systems must satisfy so that they have a capability for self-awareness, is having a minimal self defined as ‘a sense of ownership’. As this sense of ownership is usually associated with having a living body, one suggestion is that artificial intelligence systems must have similar living bodies so they can have a sense of self. Discussing cases of robotic animals as examples of the possibility of artificial intelligence systems having a sense of self, the paper concludes that the possibility of artificial intelligence systems having a ‘sense of ownership’ or a sense of self may be a necessary condition for having responsibility. (shrink)

One way of being responsible for an action is being praiseworthy for it. But what is the “praise” of which the praiseworthy agent is worthy? This paper provides a survey of answers to this question, i.e. a survey of possible accounts of praise’s nature. It then presents an overview of candidate norms governing our responses of praise. By attending to praise’s nature and appropriateness conditions, we stand to acquire a richer conception of what it is to be, and to regard (...) another as, a responsible agent. (shrink)

It is widely believed that implicit bias is common and that it contributes, in part, to the perpetuation of systemic injustice. Hence, the existence of implicit bias raises the question: can individuals be blameworthy for their implicit bias? Here, I consider what it is about implicit bias that renders agents blameworthy. I defend the claim that, when individuals omit to engage in activities that could prevent the influence of implicit bias on their behavior, they may be blamed for their implicit (...) bias. The plausibility of my proposal depends on whether individuals are able to engage freely in such activities and on whether there exists an obligation not to be biased (broadly put). I will answer positively to the former question and open the way for an answer to the latter, and tougher, question. (shrink)

In this paper I investigate which of the main conditions proposed in the moral responsibility literature are the ones that spell trouble for the idea that Artificial Intelligence Systems could ever be full-fledged responsible agents. After arguing that the standard construals of the control and epistemic conditions don’t impose any in-principle barrier to AISs being responsible agents, I identify the requirement that responsible agents must be aware of their own actions as the main locus of resistance to attribute that kind (...) of agency to AISs. This is because this type of awareness is thought to involve first-person or de se representations, which, in turn, are usually assumed to involve some form of consciousness. I clarify what this widespread assumption involves and conclude that the possibility of AISs’ moral responsibility hinges on what the correct theory of de se representations ultimately turns out to be. (shrink)

An encyclopedia article on the epistemic or knowledge condition for moral responsibility, written for the SEP.

I explore the question of when an agent is derivatively, rather than directly, culpable for an undesirable outcome. The undesirable outcome might be a harmful incompetent or unwitting act, or it might be a harmful event. By examining various cases, I develop a sophisticated account of indirect culpability that is neutral about controversies regarding normative ethical issues and the condition on direct culpability.

In this paper, we set out to test empirically an idea that many philosophers find intuitive, namely that non-moral ignorance can exculpate. Many philosophers find it intuitive that moral agents are responsible only if they know the particular facts surrounding their action. Our results show that whether moral agents are aware of the facts surrounding their action does have an effect on people’s attributions of blame, regardless of the consequences or side effects of the agent’s actions. In general, it was (...) more likely that a situationally aware agent will be blamed for failing to perform the obligatory action than a situationally unaware agent. We also tested attributions of forgiveness in addition to attributions of blame. In general, it was less likely that a situationally aware agent will be forgiven for failing to perform the obligatory action than a situationally unaware agent. When the agent is situationally unaware, it is more likely that the agent will be forgiven than blamed. We argue that these results provide some empirical support for the hypothesis that there is something intuitive about the idea that non-moral ignorance can exculpate. (shrink)

Ever since John Locke, philosophers have discussed the possibility of a normative epistemology: are there epistemic obligations binding the cognitive economy of belief and disbelief? Locke's influential answer was evidentialist: we have an epistemic obligation to believe in accordance with our evidence. In this dissertation, I place the contemporary literature on agency and reasons at the service of some such normative epistemology. I discuss the semantics of obligations, the connection between obligations and reasons to believe, the implausibility of Lockean evidentialism, (...) and some of the alleged connections between agency and justification. (shrink)

I defend the following version of the ought-implies-can principle: (OIC) by virtue of conceptual necessity, an agent at a given time has an (objective, pro tanto) obligation to do only what the agent at that time has the ability and opportunity to do. In short, obligations correspond to ability plus opportunity. My argument has three premises: (1) obligations correspond to reasons for action; (2) reasons for action correspond to potential actions; (3) potential actions correspond to ability plus opportunity. In the (...) bulk of the paper I address six objections to OIC: three objections based on putative counterexamples, and three objections based on arguments to the effect that OIC conflicts with the is/ought thesis, the possibility of hard determinism, and the denial of the Principle of Alternate Possibilities. (shrink)

Paul Silva has recently argued that doxastic justification does not have a basing requirement. An important part of his argument depends on the assumption that doxastic and moral permissibility have a parallel structure. I here reply to Silva's argument by challenging this assumption. I claim that moral permissibility is an agential notion, while doxastic permissibility is not. I then briefly explore the nature of these notions and briefly consider their implications for praise and blame.

Capacitarianism says that an agent can be non-derivatively blameworthy for wrongdoing if at the time of their conduct the agent lacked awareness of the wrong-making features of their conduct but had the capacity to be aware of those features. In this paper, I raise three objections to capacitarianism in relation to its verdict of the culpability of so-called “slips” and use these objections to support a rival (“accessibility internalist”) view which requires awareness of wrong-making features for non-derivative blameworthiness. The objections (...) are that (1) the rival internalist view is just as capable of explaining the culpability of paradigm cases of slips; (2) there are some cases of slips without dispositional awareness of wrong-making features that capacitarians wrongly deem culpable; and (3) capacitarianism cannot adequately ground the reasonable expectation to avoid slips (as Fernando Rudy-Hiller has argued). I then argue that an internalist can simultaneously ground a reasonable expectation to avoid slips and account for slips’ culpability by pointing to the satisfaction of a novel epistemic condition: it is (dispositionally) obvious to the slipping agent that they need to pay enough attention to what they are doing, and how to pay enough attention, to reliably achieve the goal that they have set out to achieve. (shrink)

This is a draft of my chapter on Negligence for the forthcoming Oxford Handbook in Moral Psychology. It discusses philosophical, psychological, and legal approaches to the attribution of culpability in cases of negligent wrongdoing.

Jonathan Bennett, Nomy Arpaly, and others see in Huckleberry Finn's apparent praiseworthiness for not turning Jim in (even though this goes against his own moral judgments in the matter) a model for an improved, non-intellectualist approach to moral appraisal. I try to show – both on Aristotelian and on independent grounds – that these positions are fundamentally flawed. In the process, I try to show how Huck may be blameless for lacking what would have been a praiseworthy belief (that I (...) should help Jim), hence, blameless for not acting on this belief; but being ‘blamelessly unpraiseworthy’ is not the same thing as being praiseworthy. (shrink)

Samuel Kerstein argues that an asymmetry between moral worth and maxims prevents Kant from accepting a category of acts that are impermissible, but have moral worth. Kerstein contends that an act performed from the motive of duty should be considered as a candidate for moral worth, even if the action's maxim turns out to be impermissible, since moral worth depends on the correct moral motivation of an act, rather than on the moral lightness of an act. I argue that Kant (...) cannot consistently maintain that there are morally forbidden, though good, acts since one of the conditions of acting from the moral law should be that one has a true belief about what the moral law requires. My project, then, rejects the possibility of morally impermissible, worthy acts for Kant, and qualifies the conditions for moral worth Kerstein gives with an epistemological constraint on moral worth. (shrink)

I start by using “Frankfurt-type” examples to cast preliminary doubt on the “Objective View” - that one is blameworthy for an action only if that action is objectively wrong, and follow by providing further arguments against this view. Then I sketch a replacement for the Objective View whose core is that one is to blame for performing an action, A, only if one has the belief that it is morally wrong for one to do A, and this belief plays an (...) appropriate role in the etiology of one's A-ing. I next defend this core against recently advanced objections and then show how it helps with defusing a skeptical challenge from the direction of causal determinism against blameworthiness. Finally, I exploit the core to isolate an analogous epistemic core for nonmoral but “normative” varieties of blameworthiness. (shrink)

On many prominent accounts, moral praise- and blameworthiness depend on the quality of will an agent manifests in their action. This paper draws attention to a pertinent but overlooked phenomenon: the manifestation of commendable and objectionable qualities of will in an action at once. By showing that all the manifested qualities of will have an effect on the agent's praise- and blameworthiness (i.e., on their action's moral worth), it is argued that quality of will theories need to provide an account (...) of this effect. An account according to which the manifested degrees of qualities of will cancel each other out and the praise- or blameworthiness is determined by the remaining quality of will is discussed and dismissed. Finally, it is argued that agents who manifest objectionable and commendable qualities of will at once are simultaneously both praise- and blameworthy for their actions. Actions can thus have ambivalent moral worth. (shrink)

Some philosophers – notably Bernard Williams, Martha Nussbaum and Ruth Barcan Marcus – argue that agents in moral dilemmas are blameworthy whatever they do. I begin by uncovering the connection these philosophers are presupposing between the agent’s judgement of wrongdoing and her tendency to self-blame. Next, I argue that while dilemmatic choosers cannot help but see themselves as wrongdoers, they both can and should divorce this judgement from an ascription of self-blame. As I argue, dilemmatic choosers are morally sui generis (...) in that their actions result in a diminishment of their personal integrity with no corresponding failure of character. It is this that makes them non-blameworthy wrongdoers. This way of seeing the problem should provide dilemmatic choosers with a novel conception of their own moral psychology, one that allows them to view their actions in a manner that is given neither to moral insensitivity nor to pathological self-accusation. (shrink)

Many philosophers think that a necessary condition on moral blameworthiness is that the wrongdoer can reasonably be expected to avoid the action for which she is blamed. Those who think so assume as a matter of course that the expectations at issue here are normative expectations that contrast with the non-normative or predictive expectations we form concerning the probable conduct of others, and they believe, or at least assume, that there is a clear-cut distinction between the two. In this paper (...) I put this widespread assumption under scrutiny and argue that it’s mistaken: although predictive and normative expectations are indeed distinct, there is no sharp separation between them. On the contrary, predictive expectations can have a substantial bearing on normative expectations in two related ways: they can recalibrate what is reasonable to expect of agents when responsibility attributions are at stake and they can help to uncover previously undetected excusing conditions. I illustrate my claims with the famous bystander effect from social psychology and show that it yields predictive expectations that affect normative expectations in these two ways. (shrink)

Thomas Hurka's brilliant study on the nature and value of virtue and vice opens with a puzzle. Hurka tells us that as consequentialism defines “all other moral properties in terms of goodness and evil”, a fully consequentialist characterization of virtue and vice should define these things by appeal to goodness and evil as well. However, it has traditionally been thought that the most promising analysis of virtue, in terms of what is intrinsically good or evil, embarks from the fundamental claim (...) that virtue is a disposition to promote good and prevent evil. But then this been concluded, the bedrock truth that virtue is intrinsically good cannot be captured successfully by any consequentialist approach. (shrink)

Thomas Hurka's brilliant study on the nature and value of virtue and vice opens with a puzzle. Hurka tells us that as consequentialism defines “all other moral properties in terms of goodness and evil”, a fully consequentialist characterization of virtue and vice should define these things by appeal to goodness and evil as well. However, it has traditionally been thought that the most promising analysis of virtue, in terms of what is intrinsically good or evil, embarks from the fundamental claim (...) that virtue is a disposition to promote good and prevent evil. But then this been concluded, the bedrock truth that virtue is intrinsically good cannot be captured successfully by any consequentialist approach. (shrink)