When Suicide is not a Self-Killing: Advance Decisions and Psychological Discontinuity—Part I

Abstract

Derek Parfit’s view of ‘personal identity’ raises questions about whether advance decisions refusing life-saving treatment should be honored in cases where a patient loses psychological continuity; it implies that these advance decisions would not be self-determining at all. Part I of this paper argues that this assessment of personal identity undermines the distinction between suicide and homicide. However, rather than accept that an unknown metaphysical ‘further fact’ underpins agential unity, one can accept Parfit’s view but offer a different account of what it implies morally: that the social and legal bases for ascribing a persisting ‘personal identity’ maintain the distinction between homicide and suicide.

Introduction

Suicide sometimes involves killing other people, as with suicide bombers, for example. People can lead or coerce others into suicide, and such deaths may also constitute homicide, such as when Jim Jones led his followers into a mass suicide at Jonestown. There may be another class of suicide, however, that is both suicide and homicide.¹ If we accept that over time the same human body can ‘house’ a series of distinct selves or persons, then an agent who antecedently initiates the causal chain that results in her death may not be the same person, metaphysically, who dies, although they inhabit the same body. This raises the question of when what seems to be suicide actually turns out to be homicide, killing another person. This is not a purely theoretical philosophical question—it applies to real-world cases of advance medical decisions. For example, if someone issues an advance decision refusing life-saving or life-sustaining treatment and later loses mental capacity due to severe brain trauma, she may be killing a different ‘person’ since this future self has profoundly different mental properties.

An advance decision is a written statement specifying in advance the medical interventions the author (a) gives consent for; (b) withholds consent for; (c) according to specific clinical circumstances; and (d) in the event of lacking the capacity to do so in the future. Although we may not ordinarily think of advance decisions to refuse future medical treatment as suicide, consider the following example: In 2007, Kerrie Wooltorton, who had a long-standing history of depression and suicide attempts, drank anti-freeze to kill herself and then called an ambulance because she did not want to die alone or in pain. She consistently refused life-saving treatment, both verbally and in a written advance decision that she had with her for when she lost consciousness. The doctors assessed her as having the mental capacity to make such a decision, so they did not treat her. She died as a result. It was clear to medical staff that her death was a suicide at least partly because of her treatment refusal in advance of losing consciousness.² (On the question of whether all such refusals would constitute suicide, see Dowie (2020, 2022)^{3, 4} on the role of intention in suicide.) By contrast, if someone makes an advance decision long before her eventual death, it is not clear that her prior decision should apply; her opinions and commitments could have changed dramatically. This is why advance directives ought to be periodically updated, especially upon the diagnosis of a terminal or debilitating illness.

To claim that someone is the same person over time, we must establish what makes him the same, keeping in mind that our physical properties change routinely; even all of our body’s cells are replaced periodically. Derek Parfit’s account of this problem is renowned; he was “among the first contemporary theorists to explore the relation between identity and ethics explicitly.”⁵ Parfit thinks that we should not consider our own body as a substratum to base our individual personhood on. He thinks that what is important in ‘personal identity’ is survival in terms of psychological continuity, in which we are a series of successive ‘selves,’ sets of psychological properties such as memories and states, that are connected. Parfit’s view may be intuitively plausible when we consider such things as cell replacement or the philosophical problem of identity posed by such examples as the Ship of Theseus. Many authors on the subject of identity in normative ethics offer accounts of what his pivotal view implies regarding various aspects of life, including healthcare. Allen Buchanan and Alasdair Maclean, for instance, employ Parfit’s conceptual framework to advocate at least a degree of paternalism in advance decisions refusing life-sustaining treatment. My aim, however, is to show that even if one accepts Parfit’s somewhat contentious metaphysical view, it may have different moral implications that do not justify such paternalism.

To determine people’s obligations to themselves in the future, in the section “Parfit’s view of psychological continuity, applied to advance decisions,” I explain Parfit’s view of ‘personal identity.’ I discuss how advance decisions refusing life-sustaining treatment may become problematic because the whole point of the advance decision is to establish what someone wants when she loses her mental capacity or her ability to communicate. On Parfit’s view, it looks as though such an advance decision should not be binding. It would be morally equivalent to making a decision about my neighbor Margo’s life, for example, should she be suffering from Alzheimer’s disease, something we would not normally think is permissible. I compare cases where there is psychological continuity with those where there is no continuity between the self who issues the advance decision and the self to whom that decision is applied, although the selves inhabit the same body—for the purposes of this paper, one can regard this psychological discontinuity. ⁶ I argue that Parfit’s view of psychological continuity undermines the distinction between suicide and homicide generally, unless that view is amended.

In the section “Psychological relations between selves and persons,” I assess Allen Buchanan’s suggestion that as we lose psychological continuity the moral force of advance decisions correspondingly diminishes. I also consider Alasdair Maclean’s proposal that we should regard the relationship between our current self and our future discontinuous self as roughly analogous to a parent–child relationship. I argue against these proposals because they imply that most people who are dying would not have the right to refuse treatment since there is generally, perhaps even always, a period of psychological discontinuity between selves while dying, even if it is very short. I then make concluding remarks.

Parfit’s view of psychological continuity, applied to advance decisions

Parfit’s distinction between selves and persons is sometimes confusing, and these terms are often used interchangeably. Therefore, adopting David Shoemaker’s distinction, going forward, I use ‘person’ to mean entities that are psychologically continuous, made up of connected ‘selves,’ unless otherwise noted (as ‘meant colloquially,’ for instance). By contrast, the ‘selves’ along this chain are the bundled sets of mental properties. Being a ‘person’ in this metaphysical sense is not necessarily the same as having ‘personal identity’ in the colloquial sense;⁷ going forward, I use ‘personal identity’ in the colloquial sense throughout this paper. The bundled mental properties constituting a self are connected to the next self with its own set of psychological properties because they share some of these properties in common, such as direct memories (see Figure 1). Self A and Self B are connected because they share overlapping psychological properties. An adult is the same person he was at 3 years old because the selves are continuous, even if not directly connected, like connected links in a continuous chain. This psychological continuity is the metaphysical basis for personhood surviving over time, Parfit maintains.

Figure 1 My schematic of Parfit’s psychological continuity model.

Parfit stipulates that uniqueness, or ‘non-branching,’ is integral to one’s personal identity surviving because we cannot have individual personhood if others are identical to us.⁸ To illustrate this, Parfit imagines a case where a brain is divided into half and one of the hemispheres is transplanted to another body (we can stipulate that the new body is an exact copy of the first).⁹ If we assume that the symmetrical hemispheres start off with identical properties (for the sake of argument), including the same memories and preferences, then if these two minds persist independently they will develop into two distinctly different minds, each becoming unique due to its separate experiences. On this scenario, the original mind would cease to survive, as the first “and the resulting two people cannot be one and the same person.”¹⁰ This is similar to cell division, where a single cell divides into two; in a sense, they both are the first cell, and yet neither is. Therefore, he stipulates that personal identity cannot assume a branching form.¹¹ Eric Olson, for one, discusses this sort of ‘branching’ in the form of ‘brain transfers’ that I address in Part II.

According to Parfit’s view, one’s lack of direct connection to a future self may be morally equivalent to being another person (meant colloquially) altogether; the separateness of my various selves is similar to the separateness of other people generally since personal identity holds by degrees. In his famous example, Parfit discusses the Russian Nobleman with strong socialist ideals who signs a document giving the land to peasants on his inheritance, that can only be revoked with his wife’s consent. Knowing that his views may change as he grows old, he makes his wife promise that she would honor the views of his younger self and regard his older self as not him but someone else who has replaced him.¹² Parfit suggests that she has reason to regard her promise as binding to the young man, rather than the older man with changed ideals, and that this older man is not the husband she loved and married.¹³ Parfit’s view implies that we may have reason to shift our commitments and that contracts may not be binding as people change.

Parfit’s view leads to the position that if one asks, “Am I the self that I was at time t?” then at some point there may not be a right or wrong answer—the answer may be indeterminate if we are sufficiently far removed from a previous bundled set of mental properties yet still continuous with that self. Parfit claims that ‘personal identity’ is less deep, involves less, and is less significant morally than people commonly take it to be. This leads to the position that since it would be impermissible to act against other people’s best interests, we ought only to do what is in our overall best interests and avoid imprudence, since acting imprudently toward another person would be wrong.¹⁴ Therefore, he writes, “We ought to prevent anyone from doing to his future self what it would be wrong to do to other people.”¹⁵ This is his basis to endorse paternalism.

On this view, when someone decides for a future self, this could be like deciding for someone else altogether. This becomes clearer when we consider the following example: If I issue an advance decision that refuses life-saving and life-sustaining medical treatment and at some later time I lose my capacity to make decisions because of severe brain damage, I may be causing the death of someone else who is not continuous with me and is therefore a different person—I could be killing someone who is not me. This implies that this death may turn out to be a homicide, a person killing another person, since under some circumstances passively allowing someone else to die can amount to killing her.¹⁶ (Note, however, that not all such resulting deaths qualify as killings, nor are all killings wrongful, as we have justifiable homicides and suicides. See Dowie (2020, 2022)^{17, 18} on causing deaths in homicides and suicides.) Parfit’s position would limit what we could permissibly outline in an advance decision. However, despite his claim that the moral difference between doing something to oneself and doing it to someone else is less significant than we take it to be, we could have morally significant reasons for regarding ourselves as having a persisting personal identity in such circumstances that grounds our decisions about the future, as I argue in Part II.

Advance decisions applied to five different cases

Advance decisions (and advance directives and living wills, according to jurisdiction) specify in advance what patients do and do not consent to relating to medical care. This is to prepare for when they have severely diminished capacity to make consent decisions contemporaneously. Advance decisions must be made while the patient has sufficient capacity to make consent decisions.¹⁹ To illustrate the problem of applying Parfit’s view of psychological continuity to advance decisions, I provide five different real-world cases showing varying degrees of connectedness and continuity between selves:

1. 1. Someone decides to shoot himself and does so immediately after making this decision. The death takes place within seconds, so there seems to be little or no significant psychological change in the person between when the decision is made and when he dies. On Parfit’s view, the self who kills and the one who dies seem to be connected and continuous. (However, in the section “Psychological relations between selves and persons,” I argue that even these selves may not be continuous, undermining the distinction between suicide and homicide, because the self who initiates the causal chain that causes death is not the same self or person who dies.)

2. 2. Someone makes an advance decision outlining what medical treatments she would refuse were she to be hospitalized with kidney failure. Several months later, she does suffer such an event. Having previously specified no antibiotic treatment in these circumstances, the advance decision is upheld, and as a result of refusing a transplant or dialysis, she falls unconscious for several days and dies without regaining consciousness. On Parfit’s view, there seems to be psychological continuity between the self who wrote the advance decision and the self who died, but there may be no direct connection between them because she has been unconscious. (Again, I dispute this in the section “Psychological relations between selves and persons.”)

3. 3. Someone makes an advance decision outlining what medical treatments he would refuse were he to be diagnosed with a severe brain injury. Later, he does come to suffer from this condition through an accident, and although he has psychological states, he is now not psychologically connected to or continuous with the self that used to inhabit that body. He does not retain the memories or personality of the prior person, though the self who now inhabits that body is deemed to have at least some residual capacity to make some basic consent decisions.

4. 4. Someone has suffered a severe hemorrhagic stroke. She is unconscious and unlikely to recover. There is no apparent psychological entity at all, and there are no signs of persisting personhood. She had an advance decision made up before her stroke, and although there remains a living body, there is no self connected to the person who issued that decision. She is no longer capable of making, much less communicating, any decisions.

5. 5. A woman had drawn up a will detailing her final instructions concerning her estate and bodily remains. There is no psychological continuity or connectedness between the remaining, but now deceased, body and the various selves and persons that had once inhabited it.

Parfit’s view raises the question, if psychological continuity is lost, do we lose our right to make choices for what remains, a discontinuous self? Alasdair Maclean explains that Parfit’s view of psychological continuity allows us to claim that “one body may house more than one morally relevant entity.”²⁰ This gives rise to the question of “why the self who currently inhabits the body should be bound by the decisions of the self who previously inhabited that body.”²¹ As Maclean points out, in cases of psychological discontinuity between selves, the advance decisions would not be self-determining at all.

Our alternatives are as follows:

1. a) We could accept Parfit’s view of selfhood and personhood as it stands. This would render many advance decisions morally unacceptable and equivalent to a type of homicide because the self who issued the advance decision may not be the person to whom the advance decision applies; he may no longer have personhood at all. Furthermore, I maintain that Parfit’s view undermines the notion of suicide generally since in killing oneself there would be a period of psychological discontinuity between the self who shoots and the entity that dies, even if that time is very short, merely a matter of seconds (or fractions thereof). A Parfitian may respond by claiming that what we take to be suicide is actually self-inflicted psychological discontinuity. This implies that taking a drug that causes one’s psychological states to be sufficiently permanently altered such that there is no connection between selves would be suicide. However, since the remaining self is not dead (or not killed) this in turn undermines what we think of as suicide and killing; they could amount to altering someone’s psychological state, but I reject this for reasons outlined in the section “Psychological relations between selves and persons.” In Parfit’s sense, it is not clear that we can kill a person because once psychological continuity is lost between selves, no person remains, though we may have killed a body. (I discuss this more in “My Response to Buchanan’s Proposal.”)

2. b) We could reject Parfit’s view of psychological continuity as a basis for personhood and defend a position that relies on a metaphysical ‘further fact’ of personal identity.

3. c) We could accept Parfit’s metaphysical claim but offer a different account of what it implies morally; we may be able to establish a ‘commonsense’ morality of treating our future selves differently from other people generally.

For this paper, I put aside b and focus on c to show that even if we accept a Parfitian account of psychological continuity, the moral implications do not necessitate paternalism in medical consent or advance decisions.

Psychological relations between selves and persons

Despite the complications of psychological continuity and personal identity, we nevertheless may determine a basis for honoring advance decisions. Although Parfit’s understanding of personal identity may have interesting metaphysical significance, his position may have different moral implications from those he suggests. Rather than claim that Parfit’s view shows that honoring advance decisions is unjustified, Buchanan and Maclean propose option c. They aim to provide a moral basis for honoring at least some advance decisions, though they both limit the force of advance decisions in various ways.

Maclean and Buchanan focus on situations such as that in case 4, but I maintain that this is not the only case that would be problematic on Parfit’s view. Even in cases where dying does not take a significant amount of time, such as in case 1, the self that initiates the causal chain that causes death is not the same self or person, according to Parfit’s view of personal identity, as who dies, so the agent could not kill himself. On Parfit’s view, suicides would not be self-killing at all; this is clearer when we consider that the process of dying affects brain functioning, and at some point before brainstem death, brain damage is severe and irreparable, inevitably leading to a complete and permanent loss of mental states. One may suppose that an exception to this might be where the person who kills dies instantly because he has shot himself in the brainstem, but even then we would have to revert to a non-Parfitian view of identity, or modify Parfit’s view, to claim that the corpse left behind was the same person (meant colloquially) who initiated the killing; the body is not the self or person on Parfit’s view.

Selves and persons in Parfit’s sense require psychological states. Even if we suppose a dead body can have a self (one devoid of mental states), then on Parfit’s view, it is not the same as the one who was alive; this would entail that if I shot Jack what remains is not psychologically continuous with him, so I have not killed Jack, I have killed a body that housed Jack. Parfit’s view undermines what we think of as death and killing. To clarify, suppose we gave someone a psychotropic drug that altered mental states completely, such that the person had no memories or other mental states connected with a prior self. On a Parfitian view, we will have destroyed the prior person without destroying the body that housed it, and this would have to be tantamount to killing, even if the post-drug self could continue to live and create new memories and so forth. In any event, adopting an unmodified Parfitian approach would require changing many aspects of our social and legal system, including how we identify people.

Therefore, I take the psychological continuity view’s failure to account for suicide as self-killing as a good reason not to accept prima facie Parfit’s view; we could reject option a. However, rather than defend a position that relies on an unidentified metaphysical ‘further fact’ of personal identity—option b—we could accept Parfit’s understanding of personhood in metaphysical terms, but offer a different moral assessment from Parfit’s understanding, as in option c. Next, I consider Buchanan’s view that endorses option c.

Buchanan’s proposal for honoring advance decisions

Allen Buchanan accepts Parfit’s view of psychological continuity. He thinks that we can nevertheless defend a commonsense morality of treating our future selves differently from other people and honoring some advance decisions. He acknowledges that psychological continuity is not an all or nothing position, but rather is one of gradation, and therefore, he thinks that the level to which we honor advance decisions should be along a continuum.

If the criterion for determining psychological continuity is set at a high threshold, then even in cases where there are continuous connections between selves we would regard someone as a different person (colloquially) by virtue of the fact that the self who was is no longer the same as the self he becomes. Here, Buchanan refers to Parfit’s Russian Nobleman example, where the young Nobleman with socialist ideals is a different self, with a very different character, from the later self.²² The Nobleman’s selves have psychological continuity, so he is the same person in the Shoemaker’s usage of the word,²³ but his later self has little in common with his younger self in his distant past. On Buchanan’s analysis, this threshold is too high for considering the younger and later selves the same person (meant colloquially); he thinks that we need not have a direct connection (or close relation) between the selves. He thinks that since there is psychological continuity between selves this is sufficient to consider these selves the same person (meant colloquially) and therefore a promise that the younger one makes may be binding later.

He suggests that if psychological continuity is present at a moderate threshold, and someone has at least some psychological states in common with a prior self, then her advance decision could be honored even though her psychological properties may have changed. The moral authority of the advance decision diminishes with lessening degrees of psychological continuity. If the criterion for determining whether two selves are the same person (meant colloquially) is set at a low threshold—someone whose psychological states have only a very little in common with a prior self—then he thinks cases where a prior self has ceased to exist will be cases where neurological damage is so extreme that we would not consider him as having personhood at all.²⁴ He thinks that we should not honor advance decisions in cases where the threshold for continuity is very low or if continuity between selves is lost.²⁵

Buchanan says, “There would be no conflict between honouring one person’s advance decision and preserving the life of the different person who succeeded him because the authority of the advance directive would be understood by all concerned to begin with the onset of incompetence and to terminate with the loss of personal identity.”²⁶ On his view, if someone suffers from sufficiently severe brain damage, she would not be the same person (meant colloquially) she was before since she no longer has personhood at all, and her advance decision would not be honored.

My response to Buchanan’s proposal

I disagree with Buchanan’s proposal of honoring advance decisions on the basis of the level of psychological continuity for the following reasons:

1. 1) I maintain that even in cases such as case 1, where intuitively the act seems to be contemporaneous with the decision, Parfit’s view of psychological continuity is problematic for advance decisions if it is accepted that the process of dying normally involves discontinuity between selves, even if only for a very short time. This becomes clearer if case 2 is amended slightly, and someone with an advance decision refusing treatment falls into post-coma unresponsiveness.²⁷ She is kept on life support for an extended period but finally dies of respiratory arrest. Then, either:

   1. a) She is psychologically continuous during that time but is equivalent to someone who is sleeping or unconscious. As in original case 2, let us assume for the sake of argument that her brain still functions normally, and if she were to wake, she could be regarded as psychologically connected to and continuous with herself before unconsciousness. If the woman in amended case 2 is psychologically continuous, then so too is the woman in the original case 2, who is in a short coma and dies without coming out of it. (However, note that on the approaches taken by Maclean and Buchanan, amended case 2 would be equivalent to case 4, one where the agent lacks personhood.)

      Alternatively, b) the woman in the amended case 2 is psychologically discontinuous with her prior selves. If we consider amended case 2 as psychologically discontinuous, then we should also do so with the original case 2, and with case 1, and in fact, most deaths of this type occur, where the person becomes unconscious (even if only for a very short time) and suffers brain damage, as organ failure or loss of blood pressure leads to cerebral hypoxia.²⁸ This means that if the woman’s selves in amended case 2 are psychologically discontinuous, then so too are most, perhaps all, people when they are in the process of dying. This would mean that Buchanan’s proposal of a moderate psychological continuity threshold would be inadequate for honoring an advance refusal in such circumstances.

2. 2) Buchanan’s proposal runs counter to the purpose of advance decisions, that is for such a time that the agent’s selves are no longer psychologically continuous and/or cannot communicate, for instance. Buchanan’s proposal requires treating people with locked-in syndrome differently from those with severe dementia, as in one case there is psychological continuity between selves and in the other there is discontinuity. We consider people with locked-in syndrome as having the same rights as people with severe dementia, however, and our duty of care to them would be similar.

3. 3) Buchanan’s proposal also undermines the point of requiring medical consent generally; the self or person who makes a consent decision, even a contemporaneous one, often is not the same self or person who must live with the consequences of that decision. In many medical procedures, there is a risk that a patient can become psychologically discontinuous due to brain damage caused by hemorrhaging, for example. If there is a basis for honoring consent decisions in cases of major surgery and prenatal plans, where there is a risk (albeit small) of becoming psychologically discontinuous, then so, too, would there be such a basis where people may become psychologically discontinuous due to mental or degenerative brain disorders.

   We may think that Buchanan’s proposal would be in place to protect a discontinuous future self from the decision of a prior self who is a different person, as in case 3. However, even though there is no psychological continuity between the self who issued the advance decision and the self now occupying that body, the self who remains still has the residual capacity to make his own decision, to either accept or reject the preexisting advance decision; this means that since he has the capacity to decide for himself, he is not bound to the prior decision, so Buchanan’s proposal is irrelevant in such cases.

4. 4) We need not ascribe a moderate threshold for being the same person (meant colloquially) over time to have a basis to honor such things as advance decisions, or indeed promises, contracts, and marriages. In his Russian Nobleman example, Parfit describes the Nobleman’s wife as plausibly having two different commitments to two different selves—one to the young Nobleman she first married and a conflicting one to her husband currently.²⁹ The Nobleman’s wife still has good reason to regard the man she married years prior as the same man to whom she is still married, even though his character is different. Her husband later in life occupies a persistent body, for instance. Her commitment to him has changed, not because he has changed or he has different selves, but because he has released her from the original promise. If her husband had died or lost personhood instead of simply changing character, then normally we would think that she is bound to her original promise unless he released her from it.

Therefore, I reject Buchanan’s proposal. Next, I address Alasdair Maclean’s proposal that we may still honor some advance decisions based on one’s relationship with one’s selves.

Maclean’s proposal for honoring advance decisions

Like Buchanan, Alasdair Maclean accepts Parfit’s view of personal identity and thinks that advance decisions should be limited. Maclean claims that if it is correct that personhood is lost in severe cases of neurological disorder, then autonomy cannot be the basis for honoring advance decisions since the agent would not be self-determining at all.³⁰ Even if someone has no personhood, however, one still has standing relationships with other people, such as relatives. He suggests that a normative justification for honoring advance decisions can be based on the obligations formed through these relationships. He points out, “Just as the external relationships may survive the change in [the patient who lacks capacity], so may the connection between the competent [self] and her now demented self.”³¹ This suggests that the selves inhabiting a single body have relationships with each other as well. He claims that the closest relationship conceptually between the competent and incapacitated selves would be the parent–child relationship—he takes this as intuitively obvious.³² He thinks that the competent self has decisional authority for future selves, acting with proxy consent on behalf of our future selves when we issue an advance decision.

He claims that it may be “better to err on the side of preserving life where it appears to be in the best interest of the patient as he or she currently is.”³³ This is because, he says, “parents have the authority to make decisions on behalf of their children where the children lack capacity to make their own decisions. This decisional authority is not absolute and must be in ‘the best interests’ or at least not be contrary to the child’s interests or welfare.”³⁴ Parents cannot, for example, refuse certain types of medical care for their children. As a result, he thinks that advance decisions, likewise, must not have absolute authority.³⁵

Maclean denies that advance decisions are equivalent to last wills because we would have to accept that the psychologically discontinuous future self is the property of the self who issues the advance decision, and we do not regard persons or selves as property. However, he claims that even if we accept this idea of selves as property, there is no clear reason to regard the self who issues the advance decision as having authority over the future self. Also, he maintains, regarding an advance decision as a statement of preference similar to a will “gives it too little weight in an arena where there are a number of morally relevant entities who have an interest in the decision to be made…”³⁶

My response to Maclean’s proposal

I disagree with Maclean’s view that advance decisions cannot be based on autonomy in cases of psychological discontinuity with future selves because, as Gerald Dworkin argues, someone can make autonomous decisions in advance of losing psychological continuity and those decisions are autonomous because they reflect long-standing commitments.³⁷ I endorse Gerald Dworkin’s view of autonomy because in the context of medical care it ensures that a doctor’s medical judgment is not privileged over what the patient values. However, even if we accept a view of autonomy that is more akin to liberty, as opposed to reflecting longer-term values and characteristics, I disagree with Maclean’s proposal for two reasons:

1) There are differences morally and legally between adults with diminished capacity and children that render Maclean’s proposal problematic. First, parents do not exercise proxy consent on behalf of their children, but are themselves responsible for their children’s welfare.³⁸ Second, the special status of children does not apply to adults with incapacity. In cases where someone is so badly disabled that as he matures he will never have capacity to make his own decisions, when he reaches the chronological age of maturity application to a court is required for someone to be empowered to make decisions on behalf of the adult.³⁹ Finally, although parents are limited in what they can refuse on behalf of their children, it does not follow that adults must do what is in their own future best interests. Maclean’s position assumes a symmetry between children and adult patients without capacity but he has not made a reasoned case supporting this adequately.

A relationship that may be a more plausible basis for honoring advance decisions would be a spousal one (or adult next of kin). A spouse or other person who possesses the appropriate power of attorney (or health-related equivalent, depending on the jurisdiction) may act as a proxy decisionmaker but (again, depending on the jurisdiction, as well as on best medical practice) “the proxy’s powers are limited by any advance decision made earlier by the donor [issuer of the decision].”⁴⁰ A proxy decision is meant to reflect what an adult has or would autonomously decide under the circumstances, according to her own settled values and commitments, not what the next of kin would want for the incapacitated patient or thinks is in her best interests. Thus, depending on the jurisdiction, the proxy decisionmaker is legally required to enact the patient’s wishes, respecting her autonomy. Limiting our future decisions to only being permissible if they are prudent undermines the point of requiring medical consent; that is, to respect patient autonomy even if doing so runs counter to what doctors think is in the patient’s interests medically. Maclean’s proposal to treat discontinuous selves that lack capacity as analogous to children supports medical paternalism and undermines patient autonomy and dignity.

The difference between advance decisions and other medical consent decisions is often thought to be the length of time between when the decision is made and when the treatment will be, but this is not necessarily relevant. For example, for prenatal plans of care it is recommended that doctors obtain their patient’s informed medical consent well in advance, to allow them time to gain information, ask questions, and consider their decisions. This is so that women do not have to make difficult choices while they are in labor and in case they suffer unexpected complications. If there is a basis for honoring consent decisions in advance of major surgery and prenatal plans, then so, too, should there be such a basis where people may become psychologically discontinuous in the future, due to mental or neurodegenerative brain disorders.

2) I disagree with Maclean about advance decisions not being analogous to last wills. Although last wills do distribute property, that is not their sole purpose. Last wills may include instructions about arrangements for one’s remains and for funeral services; we need not think of our bodies as equivalent to property if we regard advance decisions as similar to last wills. One also may stipulate who should have custody of one’s child, and that does not entail that the child is considered property. Similarly, someone may include a request in a last will pertaining to who should be a successor in a role, either as the new head of the family or as the head of a family-operated interest. They may also give advice and express preferences to loved ones, and although these may not always be legally binding, they may be seen as a last request. We honor such requests because we honor the person (meant colloquially) who has died.

Last wills often are not contracts because they amount to gifts and do not comprise an exchange, but will contracts are a type of last will that is a contract with an exchange. For example, a millionairess establishes a scholarship for a college, where it is understood that the college will name it after her in commemoration. Supreme Court Justice Benjamin Cardozo established that such last wills are contractual.⁴¹ Therefore, I suggest that Maclean’s view leads to unwarranted medical paternalism and his assessment of advance decisions not being like wills is unsupported. I return to the topic of advance decisions as contractual in Part II.

Continuity and discontinuity between homicide and suicide

There are suicides that have elements of both homicide and suicide, such as in cases where there is more than one person contributing to a death, such as with suicide bombers, suicide pacts, or in cases where a suicidal agent leads others to their own deaths, as with Ahab in Moby Dick. ⁴² There are other deaths that may straddle this division, such as when a pregnant woman commits suicide, but killing the fetus may not be viewed as homicide in some jurisdictions because it is merely a collection of cells that are part of the woman’s body. The woman is deemed to be acting autonomously in respect to herself, not in respect to another, depending on the stage of fetal development. There also may be more than one responsible party for a suicide/homicide; for example, a teenager who is bullied into suicide or the Athenians forcing Socrates to drink hemlock. Advance decisions that refuse life-saving medical treatments may have elements of both homicides and suicides metaphysically, but the person (meant colloquially) who dies is physically continuous with his other selves and has the same social/legal identity, making it a suicide in the pragmatic/legal sense.

We tend to think that homicide is morally worse if a death is premeditated and strictly intended, but with advance decisions the only self-killings that would be considered permissible are those that are premeditated (as they are outlined in the advance) and strictly intended by the person who issues them. Furthermore, they are not permissible if the person who issues them is suffering from severely diminished capacity to make consent decisions. Advance decisions seem to share this feature with suicides generally in terms of being morally permissible; one must have at least sufficient mental capacity when one chooses an assisted death (provided that choice is available), or allows oneself to die, for such a decision to be honored. Even if a given suicide is morally wrong, however, because the agent is not adequately considering his moral obligations to others, for example, this does not entitle a doctor to force treatment on him, though the doctor may have obligations to assess a patient’s capacity to make such a decision. We may, however, be justified in forcibly preventing a suicide whereby the agent also intends to harm others, as with suicide bombers, for example.⁴³

Conclusion

In Part I of this paper, I have explained Derek Parfit’s view of personal identity. I have argued that although his view undermines the distinction between homicide and suicide, we can nevertheless accept Parfit’s metaphysical claim but offer a different account of its moral implications. I rejected two accounts of his view that seek to limit the moral and legal force of advance decisions refusing lifesaving medical treatment. Rather, I maintain that accepting Parfit’s view of personal identity does not entail medical paternalism, but rather that we nevertheless have a moral and legal basis for honoring advance decisions refusing life-saving, life-sustaining, or life-extending treatment.

In Part II of this paper, I will proceed to show that rather than attribute an unknown metaphysical ‘further fact’ of agential unity, we could accept Parfit’s metaphysical claim but offer a different account of what it implies morally. I argue that we have social reasons for thinking that a dead human body has the same persisting identity as the psychologically distinct entity that once inhabited it and that the body is morally relevant. Michael Hardimon and others argue that normative roles, such as spousal roles, are contractual.⁴⁴ Likewise, I argue that healthcare professionals have a contractual relationship with patients, and advance decisions, as with other medical consent decisions, have similarities to contracts, such as life insurance policies and will contracts, the latter coming into effect when the psychological discontinuity is through death. If the doctor treats a patient against the patient’s decision to refuse consent for life-saving treatment, he has breached this contract.⁴⁵

In Part II, I return to the five cases outlined in the section “Parfit’s view of psychological continuity, applied to advance decisions” to show that we can give priority to the nearest self with sufficient capacity to make relevant decisions and extend the scope of that to his future incapacitated selves. If his selves are not connected but are still continuous, then we should ensure that his decision is current. If someone’s current self is not continuous with prior selves, but still has residual mental capacity (as in case 3), we should determine what his current self decides. If someone no longer has capacity to make this decision but has a properly drafted advance decision in place that was made when she did have capacity, then we should honor that advance decision.