Loading field map
Moral responsibility is the study of when and why a person can be properly held accountable for their actions, character, or omissions. It concerns the conditions under which it is fair to praise or blame someone, to treat them as the author of their conduct, and to subject them to reactive attitudes such as resentment, gratitude, indignation, or guilt. The field is not primarily about legal liability, though the two overlap; it is about the moral status of persons as answerable agents.
The central question of the subfield is deceptively simple: what must be true of a person, and of the world, for them to be morally responsible for what they do? Answering this requires investigating the nature of agency, the role of reasons and intentions, the possibility of free choice, and the social practices through which we hold one another accountable. The stakes are high. If no one is ever morally responsible, then praise and blame, punishment and reward, and even the moral emotions that structure human relationships would seem to rest on an illusion. If everyone is responsible for everything they do, then the excuses we offer—ignorance, compulsion, childhood, madness—lose their force.
Philosophical work on moral responsibility typically begins by distinguishing two kinds of conditions. The first is an epistemic condition: the agent must know what they are doing, or at least be capable of knowing it. Acting in ignorance, or under a false belief about the nature of one’s act, normally excuses. The second is a control condition: the agent must have some capacity to choose otherwise, or at least to guide their behavior in light of reasons. Acting under duress, hypnosis, or a seizure typically excuses because the agent lacks control.
These two conditions are not always separable. A person who acts from a mistaken belief may be blameworthy if the ignorance itself is culpable—for example, if they failed to check whether the gun was loaded. A person who acts under an irresistible impulse may be excused even if they knew what they were doing. Much of the field’s work consists in refining these conditions and testing them against difficult cases: addiction, manipulation, implicit bias, childhood deprivation, and mental illness.
A further distinction is between attributability and accountability. Attributability concerns whether an action can be traced to the agent’s character or practical identity—whether it is genuinely theirs. Accountability concerns whether it is fair to demand a response from them, to sanction them, or to alter one’s relationship with them. A person may be attributable for an action (it reflects their values) without being accountable for it (they were coerced). Conversely, some theorists argue that accountability can extend to actions that are not deeply attributable, such as negligent omissions.
The most famous debate in moral responsibility concerns its relationship to determinism. Determinism is the thesis that every event, including every human choice and action, is necessitated by prior causes according to laws of nature. If determinism is true, then given the state of the universe in the past, only one future is possible. The question is whether moral responsibility is compatible with this picture.
Compatibilists argue that moral responsibility does not require the falsity of determinism. What matters, they say, is not whether the agent could have done otherwise in some cosmic sense, but whether their action was caused in the right way—by their own deliberation, desires, and character, rather than by external force or internal compulsion. The classic compatibilist move is to reinterpret “could have done otherwise” as a conditional: the agent would have done otherwise if they had chosen otherwise. On this view, a person is responsible when their action flows from their own practical reasoning, even if that reasoning is itself determined.
Incompatibilists deny that determinism and moral responsibility can coexist. They divide into two camps. Hard determinists accept determinism and conclude that moral responsibility is impossible. Libertarians accept that moral responsibility requires indeterminism and argue that human agency involves a genuine openness in choice that escapes deterministic causation. Libertarian accounts face a notorious difficulty: if an action is undetermined, it may seem random or arbitrary, which undermines rather than supports responsibility. Libertarians respond by distinguishing between undetermined events that are nonetheless authored by the agent—for example, through a process of self-forming choice that is not causally determined but is still reason-responsive.
A third position, skepticism about moral responsibility, holds that neither compatibilism nor libertarianism succeeds. Skeptics argue that the conditions we intuitively require for responsibility—being the ultimate source of one’s actions, or having a robust ability to do otherwise—are not met under either determinism or indeterminism. Some skeptics conclude that we should abolish blame and punishment; others argue that we should retain the practices but reframe them as forward-looking tools for moral improvement rather than backward-looking judgments of desert.
A major turning point in the field came with the work of P. F. Strawson, whose essay “Freedom and Resentment” (1962) reframed the debate. Strawson argued that the theoretical question of determinism is less important than the practical question of our reactive attitudes—the natural human responses of resentment, gratitude, forgiveness, and indignation that constitute holding someone responsible. These attitudes are not beliefs about a metaphysical fact; they are expressions of our involvement in interpersonal relationships.
On Strawson’s view, to hold someone responsible is to be susceptible to these attitudes. When we discover that someone acted from malice, we resent them; when we discover they acted under hypnosis, we suspend the attitude. The question of whether determinism threatens responsibility becomes the question of whether learning that determinism is true would rationally require us to abandon all reactive attitudes. Strawson argued that it would not, because these attitudes are so deeply woven into human life that giving them up is not a real option for us. This position is called reactive-attitude compatibilism or Strawsonian compatibilism.
Strawson’s approach shifted attention from the metaphysics of free will to the ethics of interpersonal life. It also introduced a distinction between the participant stance, in which we engage with others as responsible agents, and the objective stance, in which we treat someone as a thing to be managed, predicted, or handled rather than blamed. The objective stance is appropriate for children, the severely mentally ill, and sometimes for those who have wronged us so deeply that we can no longer engage with them as persons.
Critics of Strawson argue that he conflates the question of whether we do hold people responsible with the question of whether it is justified to do so. If determinism were true, they say, the reactive attitudes might be natural but still unjustified, like a fear of spiders that persists after we learn they are harmless. Defenders respond that the analogy fails: reactive attitudes are not beliefs about causation but constitutive features of moral life, and their justification is internal to that life.
A separate strand of the field asks not whether determinism threatens responsibility, but what kind of self an agent must have to be responsible. This is the problem of self-governance or deep self theories. The basic idea is that an agent is responsible for an action when the action expresses their true self—their values, commitments, or evaluative judgments—rather than forces alien to that self.
The most influential version of this view is Harry Frankfurt’s hierarchical account. Frankfurt distinguished between first-order desires (desires to do things) and second-order desires (desires about which first-order desires to have). A person acts freely when their first-order desire is aligned with their second-order volition—when they want to want what they do. An unwilling addict, for example, has a first-order desire for the drug but a second-order desire not to want it; their action does not express their true self, and they are not fully responsible. A willing addict, who embraces their addiction, acts in accordance with their self and is responsible.
Frankfurt’s account has been refined and criticized. The problem of manipulation is central: if a neuroscientist implants a desire in someone, and the person endorses that desire at the second order, are they responsible? Many theorists say no, because the desire is not truly theirs. This has led to historical accounts, which hold that responsibility depends not only on the current structure of the self but on how that structure was formed. A person who has been manipulated, brainwashed, or subjected to coercive conditioning is not responsible for the resulting desires, even if they now endorse them.
Other theorists emphasize reason-responsiveness. On this view, an agent is responsible when their action is governed by their capacity to recognize and respond to reasons. A responsible agent is one who could have been moved by the reasons against the action, had they been present or salient. This account is attractive because it explains why children, the mentally ill, and the severely impaired are not responsible: they lack the relevant capacity. It also explains why negligence can be blameworthy: the agent had the capacity to attend to reasons but failed to exercise it.
Recent work has moved beyond the individualistic focus on control and self to examine the social conditions of responsibility. This includes the question of who gets to hold whom responsible, and under what circumstances. Feminist philosophers have argued that our practices of blame and punishment are shaped by power relations: the same action may be excused in one person and condemned in another, depending on social standing, race, gender, or class. This raises the question of whether moral responsibility is a neutral fact about agents or a socially situated practice that can be unjust.
Relatedly, relational accounts of responsibility argue that being responsible is not primarily a property of an individual but a status conferred by relationships. To hold someone responsible is to enter into a normative relationship with them—to demand an account, to expect repair, to offer forgiveness. On this view, the question “is she responsible?” is less important than “is it fair for us to hold her responsible, given our relationship and the social context?” This approach connects moral responsibility to broader questions of justice, accountability, and moral repair.
A further development is the study of collective responsibility. When a corporation harms the environment, or a nation commits atrocities, can the group itself be morally responsible, or only its individual members? Some theorists argue that groups can be agents in their own right, with decision procedures and intentions that are not reducible to the intentions of their members. Others argue that collective responsibility is a useful fiction or a shorthand for the distributed responsibility of individuals. This subfield has grown in importance with the rise of institutional wrongdoing, artificial intelligence, and climate change.
The field today is characterized by a plurality of approaches that often overlap and borrow from one another. The old debate between compatibilists and incompatibilists has not been resolved, but it has been transformed. Many contemporary compatibilists are reasons-responsive or relational theorists who do not rest their case on the conditional analysis of “could have done otherwise.” Many incompatibilists have moved from defending libertarian metaphysics to developing source incompatibilism: the view that responsibility requires being the ultimate source of one’s actions, a condition that determinism violates regardless of whether one could have done otherwise.
The skeptical position has also become more sophisticated. Some skeptics argue not that responsibility is impossible, but that our ordinary concept of responsibility is confused and should be replaced with a more humane practice of moral engagement. Others defend a forward-looking account of responsibility, in which blame is justified not by desert but by its role in moral education, relationship repair, and social coordination.
Empirical work in psychology and neuroscience has entered the conversation, though its relevance is contested. Studies on the neuroscience of decision-making, the psychology of moral judgment, and the effects of childhood adversity on agency have been used to argue that responsibility is more limited than we think. Philosophers have responded that empirical findings cannot settle normative questions about what responsibility requires, though they can inform our understanding of when the conditions for responsibility are met.
The most durable contribution of the field is its insistence that moral responsibility is not a single thing but a cluster of related practices and concepts. We hold people responsible in different ways, for different reasons, and with different consequences. The task of the field is not to discover a single fact about the universe that makes responsibility possible, but to clarify the norms that govern these practices and to ask whether those norms are justifiable. This is why the field remains central to ethics: it asks not only what we should do, but who we are when we do it, and what we may fairly do to one another in response.