Part B: Euthanasia, Assisted Death, and Their Critics
An Introduction to Bioethics.
In Part A every hard case was about whether to let dying happen — to stop or withhold treatment for a patient who could no longer decide. We ended on a puzzle: the law treats withholding and withdrawing treatment as the same, yet many caregivers feel that stopping a machine causes the death in a way that never starting it does not.
Part B takes up a related act: not letting a patient die, but actively helping them die — on their own clear request. This is where medicine, law, and philosophy disagree most deeply.
In early 2014, Brittany Maynard, 29, was diagnosed with an aggressive brain cancer (glioblastoma) and given about six months to live. Her doctors described the death she could expect: seizures, escalating pain, loss of speech and self.
She and her husband moved from California to Oregon, one of the few states where a terminally ill, mentally competent adult can ask a doctor for a prescription to end their own life. She set a date, spoke publicly about her choice, and on November 1, 2014, took the medication and died at home. (Maynard 2014)
Write or discuss:
Note
Brittany’s case differs from Nancy Cruzan’s in Part A.
Part A was about who decides for the silent. Part B is about what a person may ask for — and what others may do — when the patient is the one asking.
By the end of Part B, you will be able to:
This is a topic where people argue past each other because they are using the same words for different acts. Before any ethics, we need a clean map.
Three distinctions do all the work:
These three distinctions resolve most of the confusion before the ethics even starts.
Euthanasia means bringing about a death for the sake of the person who dies — a “good death.” It splits in two:
Almost every legal system treats these very differently: passive euthanasia is widely accepted, while active euthanasia is illegal in most of the world.
Most clinicians and courts insist that stopping treatment is not the same as killing.
On this view, a patient has a strong right to refuse treatment (so passive euthanasia is permissible), but no one has a right to be killed (so active euthanasia is not). Rachels attacks exactly this claim.
Physician-assisted death (PAD / MAID)
The doctor prescribes a lethal medication, but the patient takes it themselves. This is what Brittany Maynard did. The final act is the patient’s.
(Active, voluntary) euthanasia
The doctor administers the medication directly — for example, a lethal injection at the patient’s request. The final act is the doctor’s.
This line is legally decisive. US states that permit assisted death allow only the prescribing model; Canada and several European countries allow the doctor to administer as well. “MAID” (Medical Assistance in Dying) is the Canadian term and covers both.
| Term | Who acts last | Patient’s wish | Typical legal status |
|---|---|---|---|
| Refusing treatment | Patient (refuses) | Voluntary | Legal everywhere |
| Passive euthanasia | Doctor (withholds) | Usually voluntary/non-vol. | Widely accepted |
| Physician-assisted death | Patient (takes drug) | Voluntary only | Legal in some places |
| Active voluntary euthanasia | Doctor (administers) | Voluntary | Legal in fewer places |
| Non-voluntary euthanasia | Doctor | Patient can’t say | Highly contested |
| Involuntary euthanasia | Doctor | Against patient’s wish | Always wrong / murder |
The American Medical Association’s 1975 policy captured the standard view exactly: a doctor may stop “extraordinary means” of prolonging life (passive), but “mercy killing” (active) is forbidden — because killing is in itself worse than letting die. (Rachels 1975)
James Rachels argued that this position is confused, and often cruel in practice.
Rachels’ thought experiment
Smith killed; Jones merely let die. Is Jones any less monstrous? Almost no one thinks so. (Rachels 1975)
The case is built so everything is the same except killing vs. letting die — same motive, same outcome, same intention. Rachels calls what is left the bare difference.
If killing were worse in itself, Jones would be at least a little less guilty than Smith. He is not. So the bare difference carries no moral weight on its own — and “but that would be killing” cannot, by itself, settle the euthanasia debate.
Important
Be precise about the conclusion
Rachels does not show that active euthanasia is right. Consent, motive, consequences, and risk of abuse may still make killing usually worse — he denies only that the bare act-type does the work. Whether anything else can rescue the distinction is Foot’s question.
Philippa Foot agrees the bare physical difference is thin — but argues the distinction survives once we ground it in duties and rights. (Foot 1977)
“Killing vs. letting die” was a stand-in for “strong negative duty vs. weaker positive one” — tracking something real after all.
| Rachels (1975) | Foot (1977) | |
|---|---|---|
| Bare act-type matters? | No | No |
| So the distinction is empty? | Largely, yes | No — re-ground it |
| What does the real work? | Consequences, motive, consent | Negative vs. positive duties |
| Implication for euthanasia | Removes a key objection | Leaves a principled brake |
If the line between active and passive is not bedrock, the next question is whether the values underneath it actually support assisted death.
Dan Brock turns the discussion around: instead of asking whether killing is special, ask what values justify letting patients refuse treatment in the first place (Brock 1992). He finds two — and they are the same two that support voluntary active euthanasia.
You cannot consistently celebrate the right to refuse and flatly forbid voluntary euthanasia, Brock argues — the worries belong to consequences and safeguards, not the act itself.
Peter Singer pushes further, attacking the premise he thinks blocks clear thinking: the sanctity-of-life doctrine — the idea that human life is always equally precious simply because it is human. (Singer 1994)
Important
Singer defends some limited cases of neonatal euthanasia — deliberately ending the life of some newborns, a practice critics call infanticide. Two claims combine.
First: on the preference view, no newborn is yet a self-aware being with preferences about its own future, so the wrong that protects an adult does not yet attach to any infant. Second: there are some cases in which neonatal euthanasia produces better outcomes, even considering the eventual life this newborn would have had. (Singer 2011)
The argument’s weight rests on the quality-of-life prognosis, not on impairment as such. Even so, the newborns it identifies are largely disabled ones, which is why the position is most sharply contested among disability-rights advocates.
The cases are not all alike, and the difference marks where the disagreement actually is.
And Singer presses a consistency point: many people accept late-term abortion after a prenatal Down syndrome diagnosis, and he sees no morally significant line between a fetus shortly before birth and a newborn shortly after (Singer 2011). The premise he draws out is one much of society already accepts in another form.
A patient with ALS is on a ventilator and fully competent. She asks her physician to turn it off, knowing she will die within minutes. The physician does, giving sedation for comfort. The patient dies. Everyone agrees this was lawful and ethical.
The next patient, with the same disease but not yet ventilator-dependent, asks the physician for a lethal medication to take at home.
This is not hypothetical. After Brittany Maynard’s death, California passed the End of Life Option Act (2015), joining Oregon (1997), Washington, Colorado, and others. By 2026, assisted death was legal in fourteen US jurisdictions — Delaware authorized it in 2024, and Illinois’s law takes effect in September 2026, at which point roughly a third of Americans will live where the option exists (Compassion and Choices 2026). The map is still moving.
Every US law allows only the prescribing model, requires a terminal diagnosis (usually six months), and demands the patient be competent and act voluntarily.
Discuss: Are those three safeguards — terminal, competent, voluntary — the right lines? Which would you tighten or loosen?
Note
A distinction the safeguards quietly assume:
US law has chosen the first model precisely because the patient’s last act is itself a safeguard. Canada and the Netherlands accept the second — and inherit a harder problem of verifying consent at the very end.
J. David Velleman offers the most interesting secular objection — one that grants almost everything to the other side. (Velleman 1992)
The claim is not that euthanasia is intrinsically wrong, or that no one should ever have it. His worry is what offering the option to everyone does to everyone — including those who never wanted it.
Why an option can harm you
Right now, a gravely ill person who keeps living needs no justification — there was simply no choice. But once death becomes an accepted option, the situation changes for everyone, including those who never wanted it.
Now a patient must, in effect, justify continuing to live — to family, to caregivers, even to herself. The frail and dependent may feel they ought to “stop being a burden.” Merely making the choice available can strip away the protection of having no choice at all. (Velleman 1992)
For disability-rights groups such as Not Dead Yet, Velleman’s abstract worry is daily reality. Their objection is not religious; it is about who is actually at risk.
Harriet McBryde Johnson was a lawyer and disability-rights activist who had a progressive neuromuscular condition and used a wheelchair. In 2002 she debated Peter Singer at Princeton and wrote about it in “Unspeakable Conversations.” (Johnson 2003)
Her core reply to Singer: his confidence that disability makes a life worse is an outsider’s assumption. Survey research finds that disabled people often rate their own quality of life far higher than the non-disabled predict they would — the so-called disability paradox. The premise of the quality-of-life argument, she said, is at least far shakier than Singer assumes.
Why she found it hard to answer
Johnson granted that Singer was personable and his logic internally consistent. That was her difficulty: she could not locate an obvious flaw to point to. Her objection was therefore aimed at his premises — the quality-of-life judgments the argument starts from — rather than at any slip in his reasoning.
The critics converge on three concrete risks that safeguards must answer.
These are empirical questions, which is why both sides now argue about data — taken up after one more conceptual tool.
The doctrine of double effect tries to explain how an act with a foreseen bad effect can still be permissible — if the bad effect is not intended. (McIntyre 2023)
A classic four-part test: the act is permissible only if
The standard example: a dying patient in severe pain is given escalating morphine. The dose needed for comfort may also depress breathing — permissible under double effect, because the doctor intends relief and merely foresees the risk. The same act would be wrong if death were the intention and pain relief the cover.
Palliative sedation extends the logic: sedating a dying patient — sometimes to unconsciousness — to relieve symptoms nothing else can touch.
The defense
The intention is to relieve refractory suffering, not to kill. Studies suggest properly titrated sedation does not clearly shorten life. By double effect, it stays on the right side of the line.
The worry
If you sedate someone to permanent unconsciousness and stop feeding them, the outcome is certain death. Critics call this “slow euthanasia” — euthanasia that keeps the form of double effect while abandoning its substance.
This is the same fault line as Rachels vs. Foot, in a new place: does the intention behind an act change what the act is?
| Jurisdiction | Since | Model | Who is eligible |
|---|---|---|---|
| Oregon (US) | 1997 | Assisted death (prescribe) | Terminal, ~6 months, competent |
| Other US states | 1997– | Assisted death (prescribe) | Same; 14 jurisdictions by 2026 |
| Netherlands | 2002 | Euthanasia + assisted | “Unbearable suffering,” incl. some psychiatric |
| Belgium | 2002 | Euthanasia + assisted | Incl. minors (since 2014) |
| Switzerland | (long-standing) | Assisted suicide only | Non-residents allowed; no terminal requirement |
| Canada | 2016 | Euthanasia + assisted (MAID) | Expanded beyond terminal in 2021 |
Oregon’s law is the most studied in the world, because the state publishes annual data. The findings are consistent year after year. (Oregon Health Authority, Public Health Division 2024)
The data complicate both sides: little evidence of the predicted “carnage,” but confirmation that the drivers are existential, not just physical pain.
Canada moved further and faster than anywhere else.
By 2024, MAID accounted for 5.1% of all deaths in Canada — about 1 in 20, or 16,499 cases in a single year — among the highest rates in the world, and the speed of expansion is itself now part of the debate. (Roughly 96% of those cases involved patients whose natural death was already reasonably foreseeable.) (Health Canada 2025)
The hardest frontier is assisted death for psychiatric suffering — where the illness is mental, not terminal.
In 2022, Noelia Castillo, a young Spanish woman with OCD and borderline personality disorder, attempted suicide and was left paraplegic and in constant pain. In 2024 she asked to die under Spain’s euthanasia law.
Her father fought her in court for 601 days, arguing her mental illness made her wish to die a symptom, not a choice. Every court — up to the European Court of Human Rights — found she had capacity. She died in March 2026, age 25.
Important
Why it’s the hard case
Her death was approved for physical pain, but fought over psychiatric capacity. She was not dying — she was disabled. So the case fuses both critiques: Is this autonomy honored, or a vulnerable woman society failed first and offered death second?
Autonomy honored
Refusing a capable woman’s request only because her suffering is psychiatric applies a double standard we never use for physical illness. Every court agreed she could choose.
Vulnerability abandoned
Safeguards checked whether she qualified, not why she got there: lost home, institutional care, untreated trauma, assaults never investigated. A pianist offered to fund her care. Was death simply more available than a livable life?
Paired case: Noelia Castillo: Psychiatric Suffering, Decisional Capacity, and the Limits of Euthanasia Law
Canada’s pause
Canada planned to extend MAID to people whose sole underlying condition is a mental illness. The start date was set for March 2023 — then delayed to 2024 — then delayed again to March 2027.
The reasons cited were exactly the critics’ worries: clinicians were not confident they could reliably tell an irremediable mental illness from a treatable one, or a settled wish from a symptom. Even a system committed to expansion found a line it was not ready to cross.
The pause may become permanent. In June 2026, the Special Joint Committee on Medical Assistance in Dying recommended an indefinite exclusion of people whose sole underlying condition is a mental illness — not another delay, but a recommendation to stop planning for the expansion at all. A recommendation is not yet law, and Parliament may or may not follow it. But it marks the first time Canada’s expansion has been proposed to run in reverse.
Why this matters for the slippery-slope argument: critics predicted that each extension would make the next one easier, with no natural stopping point. Canada extended MAID three times — and then, at psychiatric illness, stopped, delayed three times, and moved toward closing the door. That is evidence against the strong version of the slope claim. Defenders of the argument can reply that the door is not shut, that Spain and the Benelux countries did cross this line, and that a decade of pressure is not a long test. Hold both thoughts.
Note
“Slippery slope” is often dismissed as a fallacy, but separate two versions:
The honest question is not “is there a slope?” but “which slope, and is this society good at stopping on it?”
We began with one disputed line — killing vs. letting die — and watched Rachels argue it carries no weight in itself while Foot rebuilt it from duties. Brock and Singer pressed the case for assisted death from autonomy and quality of life; Velleman and the disability-rights critics, led here by Harriet McBryde Johnson, pressed back — not from religion, but from concern for the vulnerable and for what a mere option does to them.
Palliative sedation raised the same question in a new form: does intention change what an act is? And the law on the ground — Oregon’s steady data, Canada’s rapid expansion and its sudden pause on psychiatric cases — shows a society still arguing over where the brakes belong. Where the line between mercy and danger falls is not yet settled.
Recall
Explain the Smith and Jones case and state precisely what Rachels concludes from it — and what he does not conclude.
Apply
A hospice patient with no pain asks for assisted death, citing loss of dignity and fear of being a burden. Using Velleman and the disability critique, what would worry you about granting it?
Debate
Should assisted death ever be available for psychiatric suffering alone, as in Belgium and the Netherlands — or was Canada right to pause?
Bioethics OER · CC BY-NC 4.0 · Brendan Shea