In "Agency and Responsibility" I argued that Christine Korsgaard's account of agency in her works including The Sources of Normativity fails as a comprehensive account. Korsgaard's approach does not adequately allow for responsibility for doing morally wrong or prudentially irrational actions.1 In the same article, I contended that Kantians need to consider basing their conception of responsibility on Harry Frankfurt's view, namely his compatibilist approach, on which a person who holds moral responsibility is defined in terms of having certain kinds of effective higher order states of desire.2 This, I thought, was an improvement both on Korsgaard's account as well as on Immanuel Kant's transcendental idealist account of the conditions for responsibility. The aim of this essay here is to reconsider my earlier endorsement of Frankfurt's definition of personhood and agency in favor of an account of personhood that draws on the work of Ludwig Wittgenstein, in light of recent and historical discussions. This includes discussions of personhood and the self, and of understanding "person" in Locke's forensic sense as the bearer of responsibility, that is, as an agent. I will consider a broader range of conceptions of responsible personhood. I will also draw on the applied ethics literature regarding the notion of a person, a literature that has focused somewhat more on moral patiency, rather than on moral agency, in contexts of life and death, such as abortion.

There are several issues in the literature on responsibility, freedom of the will, and personal identity that are often treated in isolation from each other, in sub-disciplinary silos. There is often little connection between the work of philosophers who define free will primarily as an issue in metaphysics and the work of those philosophers who treat conditions of freedom and responsibility solely as a practical problem. Personal identity, as a metaphysical issue, is its own line of research, separate from research into what makes a person a person from the practical point of view, or through the metaphysics of freedom. This essay attempts to remedy this situation in novel ways. I contend that Frankfurt's conception of freedom of the will serves Kantians well, by tying the conditions of agency to a certain kind of rational self-control, while acknowledging the relevance and pull of other conceptions of selfhood, personhood, and responsibility. Frankfurt provides valuable insight into the conditions of responsibility, in that he gives a psychological account of when an individual is a responsible agent. However, I contend that it is not the case that there is or even can be precise delineation, in terms of an exact account of necessary and sufficient conditions, between those individuals that have or lack the conditions of personhood, of moral agency and relevant kinds of moral patiency. For this reason, precision must be set aside in the moral context, and insight is required to properly settle what makes a person be a person. As Aristotle stressed in the Nicomachean Ethics, strict precision in ethics is the wrong approach. He writes:

Our discussion will be adequate if it has as much clearness as the subject-matter admits of, for precision is not to be sought for alike in all discussions, any more than in all the products of the crafts. Now noble and just actions, which political science investigates, exhibit much variety and fluctuation, so that they may be thought to exist only by convention, and not by nature. But goods exhibit a similar fluctuation because they bring harm to many people; for before now men have been undone by reason of their wealth, and others by reason of their courage. We must be content, then, in speaking of such subjects and with such premises to indicate the truth roughly and in outline, and in speaking about things which are only for the most part true, and with premises of the same kind, to reach conclusions that are no better.3

Some humility on the part of philosophers is in order in considering who is a target of moral judgment, or who it would be right or wrong to treat in certain ways, for instance by benefitting or harming oneself or others.

I want, first, to address some preliminary matters. In discussing as to what constitutes a person, when understood in terms of an agent, of an individual who is capable of responsibility, I am drawing on John Locke who introduced a forensic conception of personhood in which a person is the bearer of responsibility for action. In more contemporary terms, one would say that Locke considered it a key element of personhood that persons are moral agents. Locke writes:

Person is a Forensick Term appropriating Actions and their Merit; and so belongs only to intelligent Agents capable of a Law, and Happiness and Misery.4

Harry Frankfort too relates personhood to responsibility. His essay "Freedom of the Will and the Concept of a Person" greatly influenced subsequent discussions on both subjects. For example, Frankfurt is critical of P. F. Strawson's definition of "person," a definition on which a person is an individual with both consciousness and bodily characteristics. Frankfurt rightly points out that this applies to nonhuman animals as well, and so does not fit the notion of a person in the sense Frankfurt thinks is the most relevant one. Namely, a sense that captures

those attributes which are the subject of our most humane concern with ourselves and the source of what we regard as most important and most problematical in our lives. [FWC 6]

The distinguishing characteristic of persons, in Frankfurt's view, is the ability to form second-order desires, desires regarding what sort of desires one would want to have and what sort of desires one would not want to have. These second order desires are, according to Frankfurt, not present in nonhuman animals. If this is correct, Frankfurt has improved on Strawson's distinction, for he has found a trait that distinguishes persons from other creatures. This conception of personhood is related to Frankfurt's account of freedom of the will, and by extension to the conditions under which an individual is a moral agent. So, I think it would be fair to consider Frankfurt's account as a part of Locke's tradition of considering personhood to be a forensic term.

For the aim of investigating relevant conceptions of personhood further, I see no good reason to make a distinction between a self and a person. When talking about "personal identity," one is talking about what makes an individual the same self over time. It seems thus that there is an established philosophical practice of treating these notions as being more or less the same. Undoubtedly, one could stipulate a more specific meaning for "self" or "person" in order to distinguish between the two. However, I contend that any such distinction would be based on terminological stipulation rather than an analysis of the ordinary meaning of either term.

Viewing the person as being the bearer of responsibility is a central idea in this essay. I contend that there is a close relationship between the matter as to whether an individual is a person and as to whether an individual meets the conditions for being responsible. To this end, I am following the terminological distinctions of philosophers dating back to Locke. While in ordinary life words such as "person" or "human" are used interchangeably, philosophers have made a distinction here. An infant might be a human without being a person in the Lockean philosophical tradition, since infants do not bear moral responsibility for their actions. An alien from outer space might be a person without being a human so long as this alien meets the conditions for responsibility, whatever they may be.

Philosophers frequently use "person" in ways that do not commonly capture ordinary uses of the term, notwithstanding its broad technical and non-technical usage. Each one of these uses of "person" is plausibly technical. I argue that there is no one precise noncontroversial conception of a "person," on account of this it is the vacillation between these two that I am concerned with here.

While my argument concerns the relation between Kantian ethics and the conditions of personhood or selfhood, I do not endorse a Kantian ethical framework. Instead, this framework is used in order to assess Kant's own conception of the conditions of moral responsibility. One key element of Kant's ethical framework, as I see it in agreement with most Kant scholars, is that there are rational moral principles. Certainly, there are nuances in interpretation; however, that there are rational moral principles is an assertion all Kantians can agree upon. Kant himself, in addition to holding that there are rational moral principles, spells out what he sees as the most fundamental moral principle, namely the Categorical Imperative, and he goes in great length to expound this principle in a variety of ways, claiming they are all equivalent.

In order to identify which conditions of personhood best fit a Kantian framework, I will survey some of the interpretations that are currently being offered. To this end it is helpful to assess whether these conditions are meant as (a) sufficient conditions for personhood, (b) necessary conditions for personhood, or (c) both necessary and sufficient conditions for personhood. This is not always clear and obvious. In what follows, I discuss possible conditions commonly used to define moral personhood:

Libertarian Perspective. Several conditions for personhood are based on certain conceptions of free will. These conditions focus on moral agency. In the corresponding literature, one of the main proposed ideas is a libertarian conception of freedom. Accordingly, to be capable of being responsible, to be a person in the sense of being a moral agent, one has to have the capacity for doing otherwise. This notion of an "ability to do otherwise" is itself controversial, and compatibilist philosophers of freedom of the will have offered their own interpretations of the notion of ability to do otherwise. In any case, ability to do otherwise, read in an appropriate libertarian fashion, is taken by libertarians to be a necessary condition for the capacity to have moral personhood, qua agency.

In my interpretation, Kant's categorical imperative can be classified as belonging to a libertarian approach. Kant offers two different perspectives with regard to thinking about the self: the phenomenal self and the noumenal self. Were there just a phenomenal self, bound by the laws of nature, Kant would claim that there is no genuine moral responsibility. It is only a noumenal self, not bound by natural law, that can be capable of true moral action. I think it would be misleading to class Kant as a compatibilist. He can allow that, in a sense, the world follows deterministic laws, while in another sense it may not. Kant is a libertarian, because it is the possibility of a world outside of the causal order that allows, for him, moral responsibility.

Compatibilist Perspective. In contrast to the libertarian conception, no compatibilist philosopher takes the ability to do otherwise, construed in a libertarian way, to be a necessary condition for personhood. This makes clear that the conditions for personhood are far from obvious, at least to philosophers: a millennia-old philosophical debate divides us quite sharply in camps, with common sense having little pull for or against either view.

What I will call the early compatibilist view, in the tradition of Thomas Hobbes, David Hume, and A. J. Ayer, holds that the distinction between free, responsible action and action that is not free or responsible turns on the kind of causal chain leading to a person's action. It does not turn on the libertarian-style ability to do otherwise. I will use Ayer's version of this theory as the exemplar. Ayer distinguishes between ordinary causes and constraints. While in every case, in a deterministic universe, there is some causal chain leading to an individual's action, some of those causes are such that a person lacks the ability to do what the individual wants to do. In these cases, that individual lacks freedom. So, a person who walks from home to work, without impediment, is doing so freely. Someone who happens to be chained in a dungeon, incapable of movement, is constrained in such a way that such a person cannot walk freely. To lack this ability to do as one wishes is to lack freedom.

Having certain abilities is key to Frankfurt's account of being a person, but the ability is spelled out in terms of effective control not over one's environment but over one's psychological attitudes. In his view, freedom and personhood is a matter of desiring to do something, while also having higher order desires that line up with that lower-order desire. If I desire to do an act, while also desiring to have that desire, I am doing the act on my own free will. For Frankfurt, there is no need to appeal to libertarian freedom in order to capture the idea of responsible action. He argues in "Alternate Possibilities and Moral Responsibility," that libertarian freedom is not necessary for moral responsibility.5 The key aspect of Frankfurt's view I hope to advance here is the idea that responsible action and personhood can be conceived of in terms of these higher order desires, desires to have the desires one has. For example, in the case of an addict, who does not desire to have addictive desires, one can see where an individual may act in a fashion that is not best considered as being free or responsible.

Intellectual Capacity. Free will is not the only criterion on offer for personhood and responsibility. Several philosophers define personhood in terms of intellectual capacity. An instance for the intellectual capacity view can be found in An Essay Concerning Human Understanding where Locke draws a distinction between "man" (human being) and "person." Continuous existence of an individual as the same human being differs from continuous existence as the same person. There is of course a vast literature on personal identity and quite a lot can be said about this distinction and the way Locke draws it. The main issue I want to focus on is Locke's use of an intellectual capacity, namely the one of consciousness, in order to define personhood. To be the same person over time is to have continuity of consciousness, defined in terms of the presence of links in memory that are connecting the current moment to the past. Locke famously claims that an individual whose consciousness is continuous from the present to the past might be the same person, responsible for his or her actions, even though the individual does not occupy the same body and is not, in Locke's terms, the same human being. What I want to stress for the purpose of this essay is the relevance of humans' cognitive capacities with regard to consciousness and memory when discussing this highly influential account of personhood.

One contemporary variation of Locke's view, put forward by Peter Singer, holds that the central characteristics of personhood are "rationality and self-consciousness."6 Singer uses this kind of understanding of personhood for discussing differences in moral standing between nonhuman animals, human fetuses and embryos, infants and children, and adult human beings. This approach focuses on the conditions under which someone is a moral patient, rather than a moral agent. Using intellectual capacities such as rationality and self-consciousness for distinguishing between persons and nonpersons differs from distinguishing them in terms of free will. In an important way, it seems neutral in respect of different conceptions of what free will is and what it might amount to.

One way of bridging the gap between Locke's and Singer's intellectual conception of what makes a person a person to more recent normative accounts is by looking at the characteristics for establishing personhood as presented by Mary Anne Warren. In order to argue in support of the permissibility to abort fetuses, Warren names the following characteristics that for her are central to the concept of personhood: consciousness, emotionality, reasoning, the capacity to communicate, self-awareness, and moral agency.7 As all of these qualities are lacking in the early developing fetus, according to Warren, abortion of fetuses is, in her view, permissible.

Warren does not intend to take this list of characteristics as a set of necessary and sufficient conditions. Her main point revolves around the idea that none of these characteristics is present in a human fetus. To a greater extent than some of the narrow definitions of personhood on offer, Warren gives some criteria that ordinary people might have in mind when attributing personhood to humans.

More recently, there has been a greater focus on the role of what Warren called "moral agency." For these philosophers, normative reasons figure in what constitutes morally responsible personhood. For example, R. Jay Wallace has deemphasized freedom of the will, stressing instead the ability to govern one's actions based on moral reasons. This ability is what makes one a morally responsible individual, according to Wallace.8

Personhood and Reasons

A more complicated version of a similar view is Christine Korsgaard's theory of self-constitution. For Korsgaard too, the criterion of what makes a person a person is governing oneself through moral principles, being a responsible individual, an agent. Korsgaard is more specific regarding the principles that make one responsible, namely it is required to obey to Kantian hypothetical and categorical imperatives. Not only is following these principles a kind of condition for rationality as it is both necessary and sufficient, given that, in Korsgaard's view, following the hypothetical and categorical imperatives constitutes an individual as a self, or person.9

These examples raise the question as to whether various conceptions of selfhood or personhood are part of the practice of holding individuals responsible. The mere act of holding individuals responsible is a common practice. It is not in the purview of philosophers alone. For this reason, I suspect that the more technical the account of personhood on offer, the less likely it is to be part of the common everyday conception of personhood we implicitly apply when holding people responsible.

Anyone who has critically thought about free will, or has taught this topic to students who were previously unfamiliar with it, will recognize that the complexity of this topic can dawn on someone in a way that is not merely a reflection of previously attained knowledge. In other words, the ideas that are being presented in libertarianism, compatibilism, and hard determinism are not obvious. They only seem so to some of those versed in the relevant philosophical issues. For this reason, I am doubtful that the libertarian conception of the person, regardless whether it is spelled out in an event-causal or agent-causal way, is part of the conditions that ordinary people have in mind when they consider each other as persons.

On the other hand, the idea that those who lack intellectual capacity of some sort cannot be responsible for their actions is to me solid common sense. One rules out holding infants, or the very young, or mentally ill people responsible for their actions. Yet one does not make the same allowances for adults. P. F. Strawson puts this point well: among the considerations that lead us to deny that a person is fully responsible is "He's only a child."10 Strawson argues in favor of adding a variety of circumstances that would legitimately identify someone as not being capable of morally responsible action, or, in other words, to not be a person in the Lockean forensic sense.

In my view, Korsgaard's and Wallace's more recent views of personhood are convincing. The ability to act based on moral considerations is certainly part of the paradigm of personhood one recognizes in ordinary life. The idea that the individuals who are being held responsible are the ones that can act responsibly also appears to have some plausibility. However, Korsgaard's more detailed account, framed in terms of the hypothetical and categorical imperatives, seems less likely to be part of the ordinary practice of holding people accountable. The categorical imperative is already in and by itself controversial, given that philosophers have disagreed over it for centuries. How could such a controversial principle be part of what is involved in the practice of holding humans responsible, a practice that is suitable for people to engage with in everyday life?

I suggest that an approach inspired by Wittgenstein is in order, regarding the folk conception of the person. Borrowing from Warren's account, which is in some ways philosophically unsatisfying for it is a list rather than a strict set of necessary and sufficient conditions, might be a promising approach toward capturing what people really mean when they consider someone to be a responsible individual. A person or a self, or as in Wittgenstein's famous examples, a language or a game, might best be defined in terms of a set of loosely connected family resemblances among different capacities, rather than in terms of strict necessary and sufficient conditions. If one takes Warren's conception of a person to include both theoretical and practical reason, this approach would endorse some of the insights of Wallace and Korsgaard as well.

Frankfurt's Kantian account of responsibility, which I have previously argued for, is highly theoretical and is not obviously based in common sense. However, I still think it offers great insight. This is so because each one of the different capacities for designating personhood, such as the ones cited by Warren, needs some kind of psychological account that underpins it. With this end in view, I want to focus on the last item in Warren's list, namely moral agency. What is needed to have rational agency? Under which circumstances does one have the ability to act based on moral considerations? I contend that Frankfurt's higher-order desire account of freedom brings a psychological theory of responsible action that other accounts of agency are lacking. Any talk of governing oneself by reason, be it for practical or theoretical applications, needs a psychological account. It otherwise risks appearing to be mysterious—as one would ask oneself what are reasons anyway, and how do human minds in a naturalistic universe link up with them? Frankfurt offers a clear psychological narrative by advancing the thesis that when a person identifies with one's own first order desires, that person endorses something that may be what morally ought to or may not to be done. When that person fosters an attitude that is the morally obligatory one, one can say that the person acts on a moral reason. The actions resulting from it are to ones for which a rational person can hold another one responsible.

Moral responsibility is not attributed in a blanket fashion, where a person always has moral responsibility given one's capacities. Instead, people might have moral responsibility on one occasion and lack it on another. A person might lack responsibility when it comes, for example, to substance abuse, while at the same time this person might freely do responsible charitable work. Responsibility can be recognized on a case-by-case basis by using a Frankfurtian account, and thereby noting that an individual can be responsible for those actions that derive from desires with which the individual identifies, and one may be lenient in cases less so responsible for those pesky desires one cannot control.

Frankfurt offers great insight, and a suitable theoretical account, of responsible action, one that lines up well with Kantian rationalism. It is probably not best understood as a strict definition of the vague concept of personhood, given its commitment to a theoretical conception of desires that is controversial and that is likely not part of the common concept of person. By way of contrast, I think Mary Anne Warren puts forward an argument that has previously not been sufficiently recognized. For what makes a person a person might not be the kind of thing one can give the sort of definition that can be given to a term such as "bachelor" or "doe." Warren's list of characteristics, understood not as a strict definition but instead as a loosely connected set of criteria, is an improvement on most philosophical conceptions of person. As Wittgenstein has stressed, there are perfectly useful concepts such as "game" that do not need a strict definition to be useful in everyday life.

These uses of what constitutes a person, in moral judgment, should encourage scholars to rethink any attempt to precisely delineate a distinction between persons and non-persons. For this reason, Aristotle's advice that one should not expect precision in ethics is correct. Using insight and judgment, rather than definitions, is preferable in order to consider who should and should not be held morally accountable, and to consider who should and should not be treated in the ways we should treat persons. A bit of humility is in order in making moral judgments, as it is only with experience and careful thought that one can gain such insight.