Hostname: page-component-84b7d79bbc-g78kv Total loading time: 0 Render date: 2024-07-26T12:23:58.621Z Has data issue: false hasContentIssue false
  • English
  • Français

Novel Beings and Assisted Nonexistence

Published online by Cambridge University Press:  10 June 2021

Isra Black
Isra Black*
Affiliation:
York Law School University of York, YorkYO10 5GD, United Kingdom
*
Corresponding author. Email: isra.black@york.ac.uk
Get access
Rights & Permissions [Opens in a new window]

Abstract

This article engages with the legal regulation of end-of-existence decisionmaking for novel beings, specifically assisted nonexistence for such entities. The author explains the concept of a legal model for assisted death by reference to the substantive features of legal regimes in three jurisdictions in which assisted suicide or euthanasia is lawful. He considers how these models might fit novel beings who may require or prefer assistance to end their own existence by reference to the constituent features—abstract legal ingredients—that models for assisted death share. The author argues that extant models may block some novel beings’ access to end-of-existence assistance or fail to track what matters to them. He then examines the merits of adopting a universal model for assisted nonexistence, that is, a legal framework whose substantive features capture the end-of-existence concerns of both human and novel beings. Consideration of a unified legal framework may illuminate the discussion of assisted nonexistence for humans and novel beings. However, the paper proposes that whereas novel beings may have similar interests to humans, they may be relevantly different also. The prima facie case for adopting a one regime to rule us all approach to assisted nonexistence may be defeated by reasons for divergent regulation.

Keywords

end of lifeassisted deathnovel beingsartificial intelligence
Type
Commentary
Information
Cambridge Quarterly of Healthcare Ethics , Volume 30 , Issue 3 , July 2021 , pp. 543 - 555
Copyright
© The Author(s), 2021. Published by Cambridge University Press

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Introduction

This article engages with the legal regulation of end-of-existence decisionmaking for novel beings, specifically assisted nonexistence for such entities. Although this may appear somewhat esoteric, in my view, it is one whose analysis pays, for two reasons: first, the topic is of substantive, if speculative, interest—it is not implausible that some novel beings would possess a wish to end their own existence; second, analysis of assisted nonexistence for novel beings may facilitate our understanding of the concerns that underpin such choices to end existence in general and the adequacy of legal responses to these concerns.

By way of plan: following some preliminaries, I explain the concept of a legal model for assisted death by reference to the substantive features of legal regimes in three jurisdictions in which assisted suicide or euthanasia is lawful. I consider how these models might fit novel beings who may require or prefer assistance to end their own existence by reference to the constituent features—abstract legal ingredients—that models for assisted death share. I argue that extant models may block some novel beings’ access to end-of-existence assistance or fail to track what matters to them. I then examine the merits of adopting a universal model for assisted nonexistence, that is, a legal framework whose substantive features capture the end-of-existence concerns of both human and novel beings. Consideration of a unified legal framework may illuminate the discussion of assisted nonexistence for humans and novel beings. However, I argue that whereas novel beings may have similar interests to humans, they may be relevantly different also. The prima facie case for adopting a one regime to rule us all approach to assisted nonexistence may be defeated by reasons for divergent regulation.

Preliminaries

First, I should make clear who are the target of this article. David Lawrence and Margaret Brazier define novel beings as “intelligent, conscious life-forms sapient in the same way or greater than are human beings.”Footnote 1 Lawrence and Brazier understand sapience to carry “an implication of wisdom, reason, and insight.”Footnote 2 This is a helpful starting point. As I understand it, the status of novel being refers to entities who are nonhuman yet are possessed of a conception of the good, as well as the capacities of appreciation (which implies knowledge of relevant aspects of the external world and how it affects subjective experience) and reason. These features or properties are important for the purposes of discussing end-of-existence decisionmaking, insofar as I take it that a legally valid decision, at least in this context, presupposes, inter alia, the capacity for (minimally) rational choice, which requires the aforementioned properties. I also take it that the novel beings in which I am interested are relevantly sentient, that is, they possess, inter alia, the capacity to fare well or to fare poorly from their own point of view.

What kinds of novel beings do I have in mind specifically? Like Lawrence and Brazier, I suggest that two candidate sets of novel beings would be—they currently are not known to exist—those who possess artificial general intelligence (AGI) or who are the product of synthetic biology and endowed with sapience and sentience.Footnote 3 Indeed, there may be an overlap between these two categories. Whereas any attempt to provide examples from science fiction will doubtless prove controversial, we might think good contenders for the status of novel beings are HAL 9000 from 2001: a space odyssey,Footnote 4 Samantha in Her,Footnote 5 the Machine and Samaritan in Person of Interest,Footnote 6 the replicants in Blade Runner,Footnote 7 the humanoid Cylons in Battlestar Galactica,Footnote 8 and the replicators in Stargate SG-1.Footnote 9 A further candidate set of novel beings for discussion are posthumans,Footnote 10 that is, individuals subject to “radical enhancement” that has turned them into “fundamentally different kinds of beings, so different that [they] will no longer… be called human.”Footnote 11 Such beings may, for example, enjoy the prospect of radically extended (and healthy) life-spans.Footnote 12

Second, to aid the discussion of assisted nonexistence, it is perhaps useful to add a further property to the novel beings under consideration. I stipulate that the target novel beings have reason to require or to prefer assistance to end their own existence. In my view, this makes the discussion of assisted nonexistence more salient, since it excludes those novel beings, for example, certain “unembodied” AGI, for whom there may exist no practical impediment or experiential cost to “suicide” without assistance. It includes novel beings, for example, who perhaps consistent with the third Asimov Law,Footnote 13 or in virtue of replication or storage processes, will lack access to their own “kill switch.” It also includes those novel beings, for example, whose embodiment will be such that “suicide” without assistance would be possible, but less preferable, than assisted nonexistence.

Third, I should explain the terminology used. In academic discussion of human end-of-life decisionmaking, it is common to see assisted dying used to refer to assisted suicide or voluntary euthanasia. Footnote 14 Unfortunately, within the public debate on the legalization of end-of-life decisionmaking and assistance, certain organizations and politicians have attempted to redefine assisted dying as (physician) assisted suicide only.Footnote 15 Whatever its political merits, this move has not served the ends of conceptual clarity. Therefore, I generally prefer the term assisted death to refer to assisted suicide or voluntary euthanasia, which in any event is less euphemistic than assisted dying.Footnote 16 However, for the purposes of this article, I have chosen to use the terms end-of-existence decisionmaking and assisted nonexistence in respect of novel beings. This is to avoid begging the question whether the novel beings under consideration are alive. Footnote 17

Fourth, and as a final stipulation, I assume that the novel beings in question possess legal personality. Of course, the basis and acquisition of legal personality for novel beings are not necessarily simple matters.Footnote 18 However, to the extent that I wish to discuss novel beings’ enjoyment of legal rights, such as access to lawful assisted nonexistence, it is necessary either to assume legal personality or to argue for it. The former move enables me to proceed directly to my principal matters of interest. In addition, to be clear, assuming legal personality need not determine the substantive issue of assisted nonexistence. It is trite that the legal rights of entities endowed with legal subjectivity may differ.Footnote 19

Having set the scene, we can now proceed to consideration of legal models for assisted death for human beings.

Legal Models for Assisted Death

In this section, I explain the concept of a legal model for assisted death. At the highest level of abstraction, I take a model for assisted death to be a legal framework that regulates assisted suicide or voluntary euthanasia in a permissive way. Of course, there may be some difficulty in determining whether any one jurisdiction regulates assisted death permissively,Footnote 20 but there clearly are some jurisdictions that permit assisted death in some form and clearly are other jurisdictions that take a prohibitive stance. Still at a high level of abstraction, a legal model for assisted death regulates assisted suicide or euthanasia in a permissive way. Important to note is that different approaches to the legal regulation of assisted death are possible.

I do not propose an exhaustive presentation of the legal regimes for assisted death in permissive jurisdictions—not least because their number is now numerous and ever-increasing. Rather, I shall focus on three regimes: Oregon,Footnote 21 the Netherlands,Footnote 22 and Switzerland.Footnote 23 I have chosen these regimes, because they are representative of the way in which legal models for assisted death may differ.Footnote 24 Differences notwithstanding, it is possible to identify a number of constituent features—abstract legal ingredients—that assisted dying laws share. My interest here is in the substantive features that bear on individual eligibility for assistance to die and the legal defaults that are, all else being equal, applicable in the event of noncompliance. These features provide the basis for later discussion and include: underlying prohibition, autonomous decision, age, type of assistance, qualifying condition, and institutionalization. Footnote 25 For simplicity, I shall not press this point here and shall instead present the law jurisdiction by jurisdiction. For reasons of brevity, I shall not detail (or later discuss) the procedural features of the Oregonian, Dutch, and Swiss laws,Footnote 26 such as waiting periods,Footnote 27 or reporting and scrutiny arrangements.Footnote 28

In Oregon, the offence of “assisting another person to commit suicide” is inapplicable to a physician who complies with the conditions set out in Death with Dignity Act.Footnote 29 Under the Act, a terminally ill adult—aged 18 years or aboveFootnote 30—resident who has capacity to take an informed decision and is acting voluntarily may request a prescription for lethal medication from their attending physician.Footnote 31 Terminal illness is defined as “an incurable and irreversible disease that has been medically confirmed and will, within reasonable medical judgment, produce death within 6 months.”Footnote 32

In the Netherlands, euthanasia and assisted suicide are prohibited by the Dutch Criminal Code, articles 293 and 294, respectively. However, physicians who comply with the due care criteria set out in the Euthanasia Act (Netherlands) 2002, section 2 have a defense to these offences.Footnote 33 The attending physicianFootnote 34 may, all else being equal, provide voluntary euthanasia or assisted suicide to an individual who possesses information “about his situation and his prospects,” who has “made voluntary and carefully considered request,”Footnote 35 who has “unbearable” suffering for which there is “no prospect of improvement,” and for whom there exists “no reasonable alternative” to assisted death.Footnote 36 The suffering requirement is less broad than might first appear; although the courts have interpreted the due care criteria to permit assisted death for somatic and nonsomatic suffering,Footnote 37 existential suffering—suffering whose principal source is not clinical illness or impairment, for example, arising from being “tired of life”—is not a permissible ground, for want of a “medically recognizable condition.”Footnote 38 The Dutch regime enables access to assisted death for adults, minors aged 16 and 17 years following consultation with those exercising parental responsibility,Footnote 39 and minors aged 12–15 years on the agreement of those exercising parental responsibility.Footnote 40

The Swiss Criminal Code, article 115 criminalizes only the conduct of individuals who, acting on selfish motives, encourage or assist suicide. This prima facie permissive regime has no age limitation and enables, inter alia, the existence of “right to die” organizations who provide not-for-profit suicide assistance.Footnote 41 However, as I note elsewhere:

Physician-assisted suicide… is subject to additional regulation. The prescription of sodium pentobarbital, the preferred lethal medication… in Switzerland, is subject to federal narcotics and [federal] therapeutic products law.Footnote 42

The federal narcotics regime and therapeutic products regimes require that lethal medication is prescribed in accordance with good medical practice.Footnote 43 Misuse or negligence is backed by criminal sanction.Footnote 44 In addition, Cantonal health law requires that physicians comply with good medical practice, which is backed by administrative sanction.Footnote 45

At present, there is a lack of legal guidance as to the permissible bounds of physician assisted suicide in Switzerland. The problem lies in determining what individual circumstances are consistent with good medical practice for the purposes of prescribing lethal medication. The Swiss Federal Supreme Court in the Haas case held a “medical indication” to be a requirement for lawful prescription of lethal medication under the federal narcotics and therapeutic products regimes.Footnote 46 The court accepted that, in addition to physician assisted suicide when an individual with a somatic disease is “approaching the end of life,” the prescription of sodium pentobarbital to mentally disordered individuals may be consonant with good medical practice, subject to the further condition that the individual requesting assistance undergoes an extensive psychiatric evaluation.Footnote 47 These statements might be thought to constitute legal authority that possibly includes other nonfatal medical conditions but excludes physician assisted suicide for existential suffering. The latter claim finds support in the Federal Supreme Court judgment in Gross. Footnote 48 However, in Gross v. Switzerland,Footnote 49 the second section of the European Court of Human Rights took issue with Federal Supreme Court’s reliance on the Swiss Academy of Medical Sciences guidance on Care of patients at the end of life as an authoritative statement of good medical practice.Footnote 50 The section noted that these guidelines emanate from a “nongovernmental organization” and thus lack the “formal quality of law” necessary to satisfy the European Convention on Human Rights (ECHR), article 8(2) requirement that interference with the ECHR, article 8(1) right to decide how and when to die be in accordance with the law.Footnote 51 As such:

Swiss law, while providing the possibility of obtaining a lethal dose of sodium pentobarbital on medical prescription, does not provide sufficient guidelines ensuring clarity as to the extent of this right.Footnote 52

Specifically, it lacks “clear, State-approved guidelines” to govern cases in which “death is not imminent as a result of a specific medical condition.”Footnote 53 Notwithstanding that the section judgment is now moot,Footnote 54 there is force in this criticism. It is difficult to know with certainty what qualifying conditions fall within the ambit of lawful physician assisted suicide in Switzerland.Footnote 55

Having outlined the substantive features of three legal models for assisted death for human beings, we may proceed to consider the extent to which these legal models might enable novel beings to access assisted nonexistence.

Assisted Nonexistence for Novel Beings within Assisted Death Law?

My aim in this section is to consider the fittingness of human assisted death models to novel beings. I shall discuss fit of model by reference to constituent features I identified in the previous section: underlying prohibition, age, type of assistance, qualifying condition, and institutionalization. Footnote 56 I shall pursue two claims: first, substantive features of assisted death regimes may block access to assisted nonexistence; second, accessible assisted death regimes may lack salience—they may fail to capture the kinds of concerns that might plausibly matter to novel beings.

Underlying Prohibition

We can contrast models, like Oregon and the Netherlands, that carve the lawfulness of physician assisted death out of universal prohibitions on suicide assistance or consensual killing, with models that, like Switzerland, contain a limited prohibition and a more rigorous regime for physician assisted death. This may matter to novel beings to the extent that in the Swiss-type model, there is legal space for the growth of an assisted nonexistence regime.Footnote 57 In the carveout-type model, we find no default space for assisted nonexistence for novel beings that is relevantly different to assisted death for human beings. That is, in the absence of permissive legal change, novel beings must satisfy the substantive criteria of human assisted death regimes. I now show how this might be problematic in the ways suggested above.

Age

It is plausible that some novel beings may come into existence endowed with capacities and experience commensurate with adult human beings. This may be the case, for example, for “initially programmed” novel beings such as droids.Footnote 58 Alternatively, novel beings’ development may elapse over a substantially shorter duration than humans. For example, we may create or encounter AGIs who have very high processing capacities and access to large quantities of energy that enable them to develop at speed. The issue is that age—set according to a human chronological baseline—may be a poor proxy for the development of the capacities that enable an individual rationally to evaluate her own existence. Relatedly, age may fail to track the acquisition of experience commensurate with the development of a stable conception of the good. It may also fail to mark the point at which others interested in an individual’s welfare ought no longer to shield the latter from the consequences of her action, for reasons of her own prudential good. In addition, it may be that some novel beings lead complete “lives” over short spaces of time, such that justifiable reluctance to permit certain activities in the case of young humans may be inappropriate in the case of chronologically young novel beings. Thus, the restriction of assisted death to adults or even to minors of a certain age—understood in a human sense—may result in novel beings who meet the other substantive criteria for assisted death, being ineligible in virtue of chronological age and facing a significant wait.

Type of Assistance

The kind of assisted death available to human beings may be thought to be of diminishing practical relevance. This is because the existence of technology, such as the Deliverance Machine,Footnote 59 the Thanatron,Footnote 60 and most recently the Sarco,Footnote 61 may enable physically impaired individuals to perform suicide in a way that entails reduced risk of experiential trauma for the individual who dies or witnesses,Footnote 62 and a reduced risk of failure.Footnote 63 As such, in the future, it may not matter to human beings whether euthanasia is lawful, provided assisted suicide is. However, whether the legal regime requires novel beings to end their own existence or permits third-party existence-ending conduct may matter to novel beings. If a jurisdiction only permits assisted suicide, this may exclude access to some novel beings. In the preliminaries, I stated that some of the novel beings in which I take an interest may have reason to require assistance to end their own existence, because their nature is such that they cannot perform suicide. For example, if code that would permit an AGI to self-destroy is kept air-gapped and physically inaccessible, its only route to nonexistence is via third-party assistance.

Qualifying Condition

The requirement that an individual have a specific condition, such as terminal illness or mental disorder, in order to be eligible for lawful assisted death may make some novel beings ineligible for assisted nonexistence. In addition, the comparative prevalence of certain conditions among novel beings may entail that the legal regime for assisted death fails to address their salient end-of-existence concerns. For example, although it is not totally implausible to think of analogues to terminal illness or physical or psychological decline for some novel beings, for example, viruses for AGIs, disease for synthetic life-forms, or even mechanical decline for droids, it is an open question whether such analogues would represent what matters most to novel beings as a population. Arguably, it would not. If we imagine future entities with extended life-spans—perhaps because of the ability to embody and disembody, their end-of-existence concerns might plausibly be existential; there may come a time when the experience of such beings equates to several full human lives and are simply “tired of existence.” If we consider that existential suffering may also arise in virtue of an individual’s social circumstances, we might encounter novel beings who experience obsolescence, lack of opportunity, loss of self-esteem, isolation, or social stigma, all of which may give rise to grave suffering. Yet these ills are not the product of a loss or decline in functioning in the individual. As such, models for assisted death with suffering criteria may capture the existential end-of-existence concerns of novel beings. But they will fail to do so if—as is the case in the Netherlands—there is the additional requirement that suffering have a health-related origin. Only suffering-oriented models with less causally prescriptive suffering criteria or a liberal underlying criminal and regulatory framework can accommodate assisted nonexistence for existential reasons.

Institutionalization

All the models discussed in the previous section contain some degree of ex ante institutionalization. The Oregonian and Dutch models entrust provision of assisted death exclusively to a social institution, viz., the medical profession. The Swiss model permits anyone to assist suicide but also permits medical involvement, subject to more stringent conditions. In jurisdictions in which the provision of assisted death is exclusively institutionalized, this may have the consequence of bounding the regime along professional lines. As we have seen, for example, in the Netherlands, a medical condition is required in order for physician assisted death to be lawful. Institutionalization may therefore pose a problem for novel beings who require or desire assistance to end their own existence. Medicine might plausibly expand to accommodate the concerns of some novel beings, particularly those who are biological in a relevant sense. However, the care of other novel beings, for example, entirely robotic AGIs, may fall outside the domain of medicine. Medicalization may be one way in which the institutionalization of assisted death may block novel beings’ access to assisted nonexistence.

The foregoing analysis suggests that the substantive criteria of models for assisted death may be a poor fit for novel beings. To accommodate the claims of novel beings to assisted nonexistence, the legal regimes for assisted death may require reconceptualization. However, in the next section, I shall argue that any such reconceptualization requires careful and critical treatment.

Lawful Assisted Nonexistence for All?

My analysis of models for assisted death exposes apparent shortcomings in some or all of the legal regimes discussed. Assessed critically, the content of these models may fail to capture the potential diversity in novel beings or the range of reasons why such entities may suffer and, in consequence, form a preference for assisted nonexistence. Concretely, background prohibitions on assisted death, the exclusion of euthanasia, eligibility restrictions going to chronological age and qualifying condition, and medicalization may all serve to deny overlapping classes of novel being access to assisted nonexistence.

In light of these forecasts, we might plausibly respond by formulating a legal model for assisted nonexistence that is, in principle, applicable to both human and novel beings, that is, a model that captures the end-of-existence concerns of both populations. Indeed, to the extent that the preceding discussion of the fittingness of assisted death models to novel beings is analogous to potentially unattended human claims to assisted death, a rethink may appear welcome.

It is likely that a unified legal regime for assisted nonexistence would be frugal in its substantive criteria. We might envisage in terms of its features: the absence of any chronological age criteria; the absence of restriction on the means employed—euthanasia and assisted suicide would be lawful; a suffering criterion without health-related origin restriction; and the absence of medicalization—or institutionalization generally. On this model, it would, all else being equal, be lawful for anyone to provide assisted nonexistence to humans and novel beings when they possess relevant knowledge, have decisionmaking capacity, and are free of autonomy-undermining third-party influence, and their existence entails grave suffering to which assisted nonexistence is a proportionate response.

A pared-down model for assisted nonexistence potentially has much to commend to it. To cite one argument that might be offered in its support, a narrow qualifying condition restriction, such as terminal illness, may seem unprincipled. Terminal illness is at best a proxy for the kind of suffering that gives rise to a wish for nonexistence. We might infer, for example, from the low absolute numbers of individuals who have availed themselves of physician assisted suicide in Oregon and the fact that more than a third of individuals issued a prescription for lethal medication have not used it,Footnote 64 that not all terminally ill individuals suffer gravely.Footnote 65 Moreover, insofar as the terminally ill are, by definition, not long for this world, they may suffer less than other individuals who face the prospect of many years of existence.Footnote 66 As Lord Neuberger argued in R (oao Nicklinson and another) v. Ministry of Justice:

there seems to me to be significantly more justification in assisting people to die if they have the prospect of living for many years a life that they regarded as valueless, miserable and often painful, than if they have only a few months left to live.Footnote 67

There may be a compelling argument, therefore, for abandoning terminal illness as a criterion for assisted nonexistence in favor of a suffering-based criterion. Moreover, it may be difficult to limit, as a matter of principle, any suffering criterion to a health-related origin. As Richard Huxtable and Maaike Möller note, “‘Suffering’ itself is not a medical term and… although ‘illness’ is a significant cause thereof, it by no means commands a monopoly.”Footnote 68 As noted above, an entity may suffer gravely in virtue of being tired of existence or fare very poorly in virtue of social factors that are in practice as intractable as terminal or chronic disease. As such, it may seem arbitrary to permit assisted nonexistence for suffering with a health-related origin while excluding all suffering that is existential.Footnote 69

Arguments from “demedicalization” may also tell in favor of limited regulation of assisted nonexistence. Tania Salem argues that physician assisted suicide “transforms a private act (suicide) into a medical event,” and in so doing, individuals cede personal autonomy to the medical profession.Footnote 70 For it is physicians, in virtue of “the social and symbolic power… conferred on medicine and medical professionals in our societies,”Footnote 71 “who are in charge of freeing patients from medicine” at the end of life through the requirement that individuals submit “to medical norms and scrutiny” in order to gain access to assisted death.Footnote 72 Under medicalization, Salem observes, “people… have physician assisted suicide not only because they want it, but because physicians agree they can have it.”Footnote 73 A demedicalized model for assisted nonexistence might be thought partially to return control of the decision how and when no longer to exist to individuals. The important point is not that individuals would gain total control over assisted nonexistence; this is false—a wish to die with assistance would remain subject to the willing cooperation of others. Rather, the claim is that greater scope for self-determination that would exist were control over assisted nonexistence to be wrested from the social institution of medicine and decentralized to individuals and assistors.

No doubt there are further examples of how a legal model designed to capture the concerns of novel beings might lead to a legal regime that is better for all—that is better for human and novel beings. However, at this juncture, I would like to problematize a one regime to rule us all approach to assisted nonexistence. I shall argue using the examples discussed—qualifying condition and medicalization—that we have cause to question whether unified treatment of assisted nonexistence is the optimal way to regulate end of existence for human and novel beings.

The argument offered for an expansive qualifying condition criterion as presented fails to account for a potentially important factor: that the reason for restriction on eligibility is less about individuals who possess a wish for assisted nonexistence, and more about individuals who have no desire for the latter but who might be exposed to pressure to request it, or who might be inclined imprudently to request assisted nonexistence. As Suzanne Ost observes, “a potential danger of loosening the medical criteria for assisted death is that it becomes harder to identify and maintain boundaries.”Footnote 74 Of course, it is in part an empirical and in part a normative matter, which among terminal illness or suffering (health origin or less restrictive) would produce an acceptable distribution of access to assisted nonexistence and exposure to risk. Resolving the issue is not necessary for our purposes.Footnote 75 The relevant point is that there may exist different risk profiles between human and novel beings and indeed between classes of novel being. It is not implausible to think that some novel beings may be impervious to autonomy-undermining third-party influence, irrational decisionmaking, or weakness of will (akrasia), whereas others may share similar “vulnerabilities” to humans. As such, we may have reason to believe that some classes of beings ought to be subject to more relaxed or restrictive eligibility for assisted nonexistence, respectively, because of the other-regarding (but within class) implications of a qualifying condition criterion. The legal model for assisted nonexistence ought to vary according to the nature of the beings subject to the measure. In short, novel beings may have similar end-of-existence concerns to humans, but they may be relevantly unlike us for the purposes of regulatory response.

In respect of demedicalization, let us accept, for the sake of argument, Salem’s general claim that the medicalization impacts on individual self-determination.Footnote 76 I also accept that the demedicalization of assisted nonexistence would be a requirement for novel beings whose care stood outside the medical domain. However, I submit that there are principled arguments for medicalization,Footnote 77 or at least the institutionalization of assisted nonexistence. As such, we need not abandon medicalization for human beings or commit to deinstitutionalized assisted nonexistence for novel beings.

In my view, the strongest, albeit contingent, argument for medicalization (as opposed to institutionalization generally) stems from the claim that the purpose of medicine, properly understood, is to alleviate suffering. As Eric Cassell argues, “the mandate for the existence of a profession of medicine in society is its obligation to relieve the suffering caused by human sickness.”Footnote 78 Conceiving of the goals of medicine in this way permits the observation that, in models that contain a suffering criterion, the provision of assisted nonexistence by physicians rests in part on familiar reasons. As such, physicians ought to possess the interpretive and analogical skills that would permit the exercise of discretion in respect of requests for assisted nonexistence.Footnote 79 I submit that this may be true regardless of whether the suffering in question is of a health-related origin or existential in nature, for the skills in question go to the recognition of suffering and of the appropriate response thereto. My argument is that medicalization entrusts the operation of assisted nonexistence to a profession that is, in principle, equipped with the analytical tools to discharge it well. It would seem unnecessary to demedicalize assisted nonexistence for human beings merely because demedicalization for novel beings were necessary, for example, if the means employed to end existence were not medical in nature. Instead, the legal models for human and novel beings plausibly ought to diverge.

It is possible to argue for deinstitutionalized assisted nonexistence for everyone, however. This might be because the absence of institutionalization would entail the abandonment of the suffering criterion, because of the absence of involvement of social institutions with substantive commitments to the good,Footnote 80 such as the duty to act in an individual’s best interests. As such, deinstitutionalized assisted nonexistence might find support on liberal neutrality grounds, the argument being that the State should not interfere in the autonomous choices of its citizens.Footnote 81

I would argue, however, that rather than commit to a deinstitutionalized regime for assisted nonexistence for everyone, we ought to maintain medicalized assisted nonexistence for human beings and, if possible, to find an institutional home for assisted nonexistence for novel beings. First, it is desirable to institutionalize assisted nonexistence, to the extent that it permits ex ante scrutiny of requests for end-of-existence assistance,Footnote 82 which may attenuate the ex post involvement of the coronial, police, and prosecutorial authorities.Footnote 83 Second, and relatedly, institutionalization may be necessary in order to prevent, as opposed merely to punish, improper assisted nonexistence. Third, there is a sense of good—conceived as achieving one’s own ends well—that is compatible with liberal neutrality and that institutionalization of assisted nonexistence might promote. Institutionalization may help entities to receive competent assistance to end their own existence, that is, assistance that has a high probability of success and that involves the minimum amount of suffering; deinstitutionalized assisted end of existence cannot ensure this.Footnote 84 Institutionalization may also assist in weeding out irrational or akratic requests for assisted nonexistence. Fourth, and perhaps controversially, I would suggest that the State ought to institutionalize assisted nonexistence precisely, because it permits the imposition of a substantive conception of the good. In my view, it is important, both for the sake of the individual who will cease to exist and for the sake of their assistor, that the reasons for providing assisted nonexistence go beyond mere respect for autonomy or even respect for an individual’s own conception of the good. To this end, I would respectfully endorse the dictum of Lady Hale in R (oao Purdy) v. Director of Public Prosecutions:

It is not for society to tell people what to value about their own lives. But it may be justifiable for society to insist that we value their lives even if they do not.Footnote 85

The upshot of these arguments is that it may be preferable not to unify the legal regulation of assisted nonexistence, notwithstanding that human and novel beings may share the concerns that motivate a desire for assisted end of existence.

Conclusion: Similar Interests, Different Models?

In this article, I have attempted to engage with the issue of how the law might regulate assisted nonexistence for novel beings. I outlined three models governing assisted death for human beings. Taking the constituent features of these models as a frame, I considered the potential obstacles novel beings might encounter were they to seek access to assisted nonexistence; these were manifold. I subsequently considered a one regime to rule us all approach to assisted nonexistence, that is, a legal model applicable to both human and novel beings. While prima facie attractive, I argued that a unified legal model for assisted nonexistence may fail to take into account relevant differences between human and novel beings; it may lead us to choose or abandon legal criteria that serve a useful purpose for human or novel beings. In sum, there may be merit in divergent legal regulation of assisted nonexistence.

If I may end on a brief methodological note, my approach to the study of end of existence involving novel beings has been to attend to potential common ground between human and novel beings, and also relevant differences between them. I suggest that both human and novel beings share an interest in deciding how and when no longer to exist. But how we respond to that interest in legal form ought sometimes to diverge. To derive one potentially generalizable aid for the study of novel beings, there is merit in maintaining separation between the issue and its means of resolution. We ought not to expect the regulation of novel beings to mirror the regulation of ourselves.

Acknowledgements

I would like to thank an anonymous peer reviewer for CQHE and the audience at the Regulating the Tyrell Corporation: Company Law and the Emergence of Novel Beings, Wellcome Trust, 2018 for helpful comments a previous version of this article. Responsibility for errors is mine alone.

References

Notes

1. Lawrence, DR, Brazier, M. Legally human? “Novel beings” and English law. Medical Law Review 2018;26(2):309–27, at 309.CrossRefGoogle ScholarPubMed

2. See note 1, Lawrence, Brazier 2018, at 312.

3. See note 1, Lawrence, Brazier 2018, at 314–7.

4. Clarke, AC. 2001: A space odyssey. Hutchinson (London); 1968.Google Scholar

5. Jonze, S. Her. Warner Bros (United States); 2013.Google Scholar

6. Nolan J. Person of Interest. Warner Bros (United States);2011–2016.

7. Scott, R. Blade Runner. Warner Bros (United States); 1982.Google Scholar

8. Moore RD. Battlestar Galactica. NBC Universal (United States);2004–2009.

9. Wright B, Glassner J. Stargate SG-1. MGM (Canada; United States); 1997–2007.

10. See note 2.

11. Agar, N. Humanity’s end: Why we should reject radical enhancement. MIT Press (Cambridge, Mass); 2010:2.CrossRefGoogle Scholar

12. See note 11, Agar 2010, at chap. 5 and 6.

13. Asimov I. I, Robot. Gnome Press (New York); 1950. I see nothing incoherent in the idea that a novel being could, all else being equal, be hardwired to protect its own existence whereas at the same time be possessed of the capacity to regret that it exists.

14. See, e.g., Coggon, J. Assisted dying and the context of debate: “Medical law” v. “end-of-life” law. Medical Law Review 2010;18(4):541–63;CrossRefGoogle Scholar Huxtable, R. Splitting the difference? Principled compromise and assisted dying. Bioethics 2014;28(9):472–80;CrossRefGoogle ScholarPubMed Lewis, P, Black, I. Adherence to the request criterion in jurisdictions where assisted dying is lawful? A review of the criteria and evidence in the Netherlands, Belgium, Oregon, and Switzerland. The Journal of Law, Medicine & Ethics 2013;41(4):885–98.CrossRefGoogle ScholarPubMed

15. See, e.g., Assisted Dying (No 2) Bill; 2015; available at http://www.publications.parliament.uk/pa/bills/cbill/2015-2016/0007/16007.pdf (last accessed 15 Dec 2020); Dignity in Dying, “Our Position”; available at https://www.dignityindying.org.uk/assisted-dying/our-position/ (last accessed 15 Dec 2020).

16. When one provides euthanasia or supplies medication that is subsequently used to fatal effect, one assists an individual to die, i.e., one assists death, not dying.

17. See note 1, Lawrence, Brazier 2018, at 318 et seq.

18. See note 1, Lawrence, Brazier 2018.

19. Consider the case of companies, adolescents, noncitizens, etc.

20. For example, there exists disagreement whether the legal regulation of assisted suicide in England and Wales, that is, whether the universal statutory prohibition on encouraging or assisting suicide contained in the Suicide Act 1961, s 2, subject to the requirement of prosecutorial consent to prosecution contained in s 2(4) of the Act (supported by the Director of Public Prosecutions, Policy for Prosecutors in Respect of Cases of Encouraging or Assisting Suicide (2010, amended 2014); available at https://www.cps.gov.uk/sites/default/files/documents/legal_guidance/assisted-suicide-policy.pdf (last accessed 15 Dec 2020), constitutes a permissive legal regime. Cf., Lewis P. “[t]he DPP has… implicitly describe[d] (albeit imperfectly through the use of factors for and against prosecution) a class of assisted suicides which are permissible.” Informal legal change on assisted suicide: The policy for prosecutors. Legal Studies 2011;31(1):119–34, at 133; Montgomery J. “the Director has skilfully negotiated a path that clarifies his prosecution policy without stepping into legislative territory.” Guarding the gates of St Peter: Life, death and law making. Legal Studies 2011;31(4):644–66, at 664–5.

21. Death with Dignity Act (Oregon); ORS §127.800–127.995.

22. Wet toetsing levensbeëindiging op verzoek en hulp bij zelfdoding (Termination of Life on Request and Assisted Suicide [Review Procedures] Act 2002) (Euthanasia Act [Netherlands] 2002). English language translation: House of Lords, Select Committee on the Assisted Dying for the Terminally Ill Bill, Volume II: Evidence. HL Paper 86-II; 2005:396; available at http://www.publications.parliament.uk/pa/ld200405/ldselect/ldasdy/86/86ii.pdf (last accessed 15 Dec 2020).

23. Code pénal suisse du 21 décembre 1937 [Swiss penal code of December 21, 1937], art 115 partially criminalises suicide assistance. We shall see that the full legal picture is more complex: see Black, I. Existential suffering and the extent of the right to physician-assisted suicide in Switzerland: Gross v. Switzerland [2013] ECHR 67810/10. Medical Law Review 2014;22(1):109–18.CrossRefGoogle Scholar

24. The Oregonian and Dutch models are also important, because they have been transferred (more or less) into other jurisdictions. Regarding Oregon, see Death with Dignity Act (Washington); RCW 70.245.010; Patient Choice and Control at the End of Life Act (Vermont); End of Life Option Act (California); End of Life Options Act (Colorado); Death with Dignity Act (District of Columbia); Our Care, Our Choices Act (Hawaii); and Aid in Dying for the Terminally Ill Act (New Jersey). The Oregonian regime also provided the basis for recent attempts to legalize assisted suicide in England and Wales and Scotland, respectively: Assisted Dying (No 2) Bill 2015; Assisted Suicide (Scotland) Bill 2013. Regarding the Netherlands, see Loi relative à l’euthanasie du 28 mai 2002 [Law on euthanasia of May 28, 2002] (Euthanasia Law (Belgium) 2002) and Loi du 16 mars 2009 sur l’euthanasie et l’assistance au suicide [Law of March 16, 2009 on euthanasia and assisted suicide] (Euthanasia Law (Luxembourg) 2009). The influence of Dutch regime is arguably visible in both the Act Respecting End-of-Life Care (Québec) and Bill C-14 (medical assistance in dying) (Canada).

25. Penney Lewis and I employ the idea of features or legal ingredients in The effectiveness of legal safeguards in jurisdictions that allow assisted dying, Briefing Paper for the Commission on Assisted Dying. Demos (London); 2012; available at http://philpapers.org/rec/LEWTEO-8 (last accessed 15 Dec 2020). The underlying prohibition feature is new, and I reframe what we called “identity of the assistor” as institutionalisation.

26. Of course, individual provisions may contain both substantive and procedural elements, e.g., ORS 127.830 §3.04 requires that a person take an “informed decision,” which is defined in ORS 127.800 §1.01.(7) as a “decision… based on an appreciation of the relevant facts and after being fully informed by the attending physician of… [relevant matters pursuant to ORS 127.816 §3.01.].”

27. See, e.g., ORS 127.840 §3.06.

28. See Lewis, P, Black, I. Reporting and scrutiny of reported cases in four jurisdictions where assisted dying is lawful: A review of the evidence in the Netherlands, Belgium, Oregon, and Switzerland. King’s Law Journal 2013;4:MLI 221–39.Google Scholar

29. ORS 163.193.

30. ORS 127.800 §1.01.(1).

31. ORS 127.810 §2.02.(1). “Attending physician” means “the physician who has primary responsibility for the care of the patient and treatment of the patient’s terminal disease”: ORS 127.800 §1.01.(2). ORS 127.810 §2.02.(1). Capacity is defined as “the ability to make and communicate healthcare decisions to healthcare providers”: ORS 127.800 §1.01.(3). Voluntariness is specified as a requirement but not defined in the Act, although 127.810 §2.02 requires additionally that witnesses attest to the absence of coercion. An informed decision requires that the requestor bases their decision on an appreciation of “relevant facts,” including diagnosis, prognosis, potential risks, and probable consequences associated with taking lethal medication, as well as “feasible alternatives” to suicide: ORS 127.800 §1.01.(7).

32. ORS 127.800 §1.01.(12).

33. Euthanasia Act (Netherlands) 2002, at s 20.

34. “[T]he physician who, according to the notification, has terminated life on request or has provided assistance with suicide”: see note 33, Euthanasia Act (Netherlands) 2002, at s 1(c).

35. Making a “carefully considered” request requires the possession of decisionmaking capacity: Lewis, P, Black, I. Adherence to the request criterion in jurisdictions where assisted dying is lawful? A review of the criteria and evidence in the Netherlands, Belgium, Oregon, and Switzerland. Journal of Law, Medicine & Ethics 2013;41(4), at 888.CrossRefGoogle ScholarPubMed

36. See note 33, Euthanasia Act (Netherlands) 2002, at s 2(1).

37. Chabot NJ 1994, no 656 (NL Supreme Court), i.e., assistance is limited to neither terminal illness nor “physical” conditions.

38. Brongersma NJ 2003, no 167 (NL Supreme Court).

39. See note 33, Euthanasia Act (Netherlands) 2002, at s 2(3).

40. See note 33, Euthanasia Act (Netherlands) 2002, at s 2(4).

41. Bosshard, G. Switzerland. In: Griffiths, J, Weyers, H, Adams, M, eds. Euthanasia and Law in Europe. Hart (Oxford); 2008:474.Google Scholar

42. See note 23, Black 2014, at 110. The relevant statutes are: Loi fédérale sur les stupéfiants et les substances psychotropes du 3 octobre 1951 [Federal law on narcotics and psychotropic substances of October 3, 1951] (LStup); and Loi fédérale sur les médicaments et les dispositifs médicaux du 15 décembre 2000 [Federal law on medicines and medical devices of December 15, 2000] (LPTh).

43. LStup, at art 11(1); LPTh, at art 26.

44. LStup, at art 20(1)(e); LPTh, at art 86(1)(a).

45. See note 41, Bosshard 2008, at 473.

46. Haas Entscheid 2A48/2006 (3 November 2006) (BGer) [6.3.2]. English language translated extracts: Haas v. Switzerland (2011) 53 EHRR 33 (ECtHR) [16]. See Black, I. Suicide assistance for mentally disordered individuals in Switzerland and the state’s positive obligation to facilitate dignified suicide. Medical Law Review 2012;20(1):157–66.CrossRefGoogle ScholarPubMed

47. See note 46, Haas 2006, at [6.3.4.]–[6.3.5.2.].

48. Gross Entscheid 2C_9/2010 (12 April 2010) (BGer). English language translated extracts: Gross v. Switzerland (2014) 58 EHRR 7 (ECtHR) [32]–[33].

49. See note 48, Gross v. Switzerland 2014.

50. Swiss Academy of Medical Sciences. Care of patients in the end of life (2004, revised 2013); available at https://www.samw.ch/dam/jcr:de64e102-1495-4c48-9fbd-1c7d4d45932f/guidelines_sams_end_of_life_2012.pdf (last accessed 15 Dec 2020). The new SAMS guidance replaces the terminal illness requirement with a condition that “[t]he symptoms of disease and/or functional impairments are a source of intolerable suffering for the patient [from their own perspective],” which appears inclusive of nonfatal conditions and possibly some cases of existential suffering: Medical-ethical guidelines: Management of death and dying 2018:23; available at https://www.samw.ch/dam/jcr:25f44f69-a679-45a0-9b34-5926b848924c/guidelines_sams_dying_and_death.pdf (last accessed 15 Dec 2020). I have added the text in square brackets to clarify the English language version, based on the French language guidance, which reads: «Les symptômes de la maladie et/ou les limitations fonctionnelles du patient lui causent une souffrance qu’il juge insupportable» (emphasis added). The shift to this condition has proved controversial, with the Swiss Medical Association (FMH), for the first time, refusing to adopt the guidance into its own Code of Ethics, on grounds that the term (subjectively appreciated) intolerable suffering was an indeterminate legal notion that gave rise to uncertainty for physicians: “La FMH ne reprend pas les directives de l’ASSM «Attitude face à la fin de vie et à la mort» dans son Code de déontologie” 2018; available at https://www.fmh.ch/files/pdf23/communique_de_presse_la_chambre_medicale_est_favorable_a_une_revision_du_tarif_en_partenariat.pdf (last accessed 15 Dec 2020). In consequence, the 2004 SAMS guidance remains in the FMH Code of Ethics.

51. See note 48, Gross v. Switzerland 2014, at [65]–[67].

52. See note 48, Gross v. Switzerland 2014, at [67].

53. See note 48, Gross v. Switzerland 2014, at [66].

54. The Grand Chamber declared the case inadmissible for abuse of rights: Gross v. Switzerland [2015] 60 EHRR 18 (ECtHR). See Black, I. A postscript to Gross v. Switzerland. Medical Law Review 2014;22(4):656.Google Scholar

55. I take it that the fact that physician assisted suicide has been provided without sanction to individuals who are neither terminally ill nor mentally ill is not determinative of its legality: See note 23, Black 2014, at 110–11.

56. I shall not discuss the autonomous decision criterion, since I doubt that—unlike the others—this criterion poses any problems for novel beings.

57. See note 41, Bosshard 2008, at 472 for an overview of the history of organized assisted suicide in Switzerland.

58. The thought here is that some novel beings may start off with a predetermined set of features but possess the ability to develop their own subjectivity over time.

59. Science Museum. “Euthanasia machine, Australia, 1995–1996” http://www.sciencemuseum.org.uk/broughttolife/objects/display.aspx?id.

60. Wired. “The Thanatron, Jack Kevorkian’s Death Machine” 2007; available at https://www.wired.com/2007/06/the-thanatron-j/ (last accessed 15 Dec 2020).

61. Exit International. “The Sarco” 2018; available at https://exitinternational.net/sarco/ (last accessed 15 Dec 2020).

62. For example, arising from the need to ingest lethal medication, or voluntarily stopping eating and drinking.

63. In Oregon, there have been a very small number of cases (n = 8) in which individuals have regained consciousness after an attempt to utilize prescribed lethal medication: Oregon Public Health Division. Oregon Death with Dignity Act: 2018 Data Summary 2019, Table 2; available at https://www.oregon.gov/oha/ph/providerpartnerresources/evaluationresearch/deathwithdignityact/Documents/year21.pdf ( last accessed 15 Dec 2020).

64. A total of 2,217 prescriptions were written for individuals under the Death with Dignity Act (Oregon) between 1997 and 2018; 1,459 were used: Oregon Public Health Division (see note 63), at 5, Table 2.

65. I appreciate that some terminally ill individuals would suffer gravely without the option of physician assisted suicide, i.e., the provision of a prescription for lethal medication itself provides succour.

66. Of course, I am not suggesting that the suffering of terminally ill individuals is never sufficient to qualify for assisted nonexistence.

67. R (oao Nicklinson and another) v. Ministry of Justice; R (oao AM) v. DPP [2014] UKSC 38 [122].

68. Huxtable, R, Möller, M. Setting a principled boundary? Euthanasia as a response to ‘life fatigue.’ Bioethics 2007;21(3):117–26.CrossRefGoogle ScholarPubMed

69. I take it that if such social factors are not intractable in principle or in some reasonable practical sense, this grounds a general reason against assisted death, for both humans and novel beings.

70. Salem, T. Physician-assisted suicide: Promoting autonomy or medicalizing suicide? Hastings Center Report 1999;29(3):30–6, at 30.CrossRefGoogle ScholarPubMed

71. See note 70, Salem 1999, at 33.

72. See note 70, Salem 1999, at 35.

73. See note 70, Salem 1999, at 35.

74. Ost, S. The de-medicalisation of assisted dying: Is a less medicalised model the way forward? Medical Law Review 2010;18(4):497–540, at 526;CrossRefGoogle ScholarPubMed see note 14, Coggon 2010.

75. See Black I. Refusing life-prolonging medical treatment and the ECHR. Oxford Journal of Legal Studies 2018;38(2):299–327, at 316–8 for discussion in the context of refusal of life-prolonging medical treatment.

76. Cf., Huxtable, R. Whatever you want? Beyond the patient in medical law. Health Care Analysis 2008;16(3):288–301;CrossRefGoogle ScholarPubMed see note 14, Coggon 2010, at 543–4.

77. There are pragmatic arguments for medicalisation too. As Ost (see note 74), at 539, no 196 notes: “it is arguably better to advocate the model that is more likely to be acceptable to the legislature and public… advocates of legal reform should unite their claims with widely accepted cultural values, the process of ‘frame alignment.’”

78. Cassell EJ. The Nature of Suffering and the Goals of Medicine. 2nd ed. OUP (New York); 2004:61.

79. I am not suggesting that physicians necessarily possess these skills. Indeed, Cassell’s critique is that modern medicine has lost its connection to suffering: see note 78, Cassell 2004, at 61.

80. Coggon (see note 14), 543–4.

81. See Dworkin R. Life’s Dominion: An Argument About Abortion, Euthanasia, and Individual Freedom. 1st ed. Knopf (New York); 1993.

82. See Brownsword, R, Lewis, P, Richardson, G. Prospective legal immunity and assistance with dying: Submission to the Commission on Assisted Dying. King’s Law Journal 2012;23(2):KLJ 181–93;CrossRefGoogle Scholar Ost (see note 74), at 537.

83. My claim here is not that there would be no ex post review of assisted nonexistence, but that such review may be lighter touch, e.g., if there is a “buffer” between assistors and the institutions of criminal or administrative justice: see note 28, Lewis, Black, at 238.

84. It is perhaps telling that in Switzerland, almost all assisted suicides involve physicians and a right-to-die association: van der Heide, A, Deliens, L, Faisst, K, Nilstun, T, Norup, M, Paci, E, et al. End-of-life decision-making in six European countries: Descriptive study. Lancet 2003;362(9381):345–50, at 347. Of course, the persuasiveness of this claim depends on the particularities of ending the existence of classes of novel beings.CrossRefGoogle ScholarPubMed

85. R (oao Purdy) v. Director of Public Prosecutions [2009] UKHL 45 [68]. I think this last claim has less force for entities who are unable to perform suicide in some way.