The View from Hell

Just another site

Archive for May 2008

When It’s Permissible To Force Someone To Stay Alive For His Own Good

with one comment

I think the question of when, if ever, it’s morally permissible to force someone to stay alive against his will, for his own good, is one of the most difficult questions to appropriately answer. While I argue in favor of an institutional right to effective, comfortable suicide, the opponent that I feel has the most interesting standing to challenge me is the reformed suicide: someone who attempted suicide, or came close to attempting suicide, but was “rescued” from his intentions, and now believes that rescue to have been in his best interests.

First, I would like to demonstrate that it is sometimes permissible to stop a person from killing himself. The example I find most persuasive is that of a person in an acute confusional state (delirium) secondary to a physical illness, such as diabetes. If a person in an acute confusional state walks out into traffic or engages in other self-harming behaviors, I would feel completely justified in intervening and basically forcing the person to stay alive. But why?

In previous posts, I’ve examined a non-paternalistic explanation of how the right to die might be said to harm some people (though, as I pointed out, this only applies to suicide rights for the terminally ill or catastrophically disabled). It is my view that stopping a person in an acute confusional state from harming himself is also allowed on non-paternalistic grounds, whereas stopping the suicide of a non-delirious person has only paternalistic justifications.

Paternalism obtains when individuals or the state wish to substitute their judgment for that of another, on the grounds that the affected person will make a poor decision. There is a real risk that paternalistic interventions will prevent people from being best off according to their own values. But guiding the deranged diabetic out of traffic is, I argue, not paternalistic, because the action of walking out into traffic, in this limited case, is not based on judgment at all. Put another way, the “person” deciding to walk out into traffic is not really a person at all – the circumstance is not attributable to anyone’s will or decision, so intervening with our own will or decision is not paternalistic. George didn’t decide to walk out into traffic – his delirium “made” him, and when George gets his insulin shot and comes back to us, “he” will be grateful to us for saving “him” from his illness.

The problem, of course, is how far to apply this idea of a suicide being the result of circumstance, delirium, and illness, rather than the result of a choice by the suicide. Many people seem willing to take the incident of a suicide attempt itself as probative evidence that the person is not in his right mind, that is, acting under something other than his will. The extreme medicalization of depression has allowed society, including the mental health industry, to take the view that suicide is always a sign of illness. (Revealingly, a suicide attempt, in and of itself, is not valid diagnostic criteria even for a major depressive episode under the DSM-IV, though of course, in practice, the elastic criteria for diagnosing depression are often fudged.) These people argue that intervening in an attempted suicide is always permissible – and, perhaps, never paternalistic – because suicide, in this view, is never truly a willed act, and always the result of something outside the person, such as (vaguely defined) illness.

We have to draw a line as to when it’s morally permissible to intervene in a suicide attempt, or to withhold the means for suicide from a person. Some, as I have explained, would draw the line at “always.” Hopefully my writings of the past two months might give a tiny bit of pause to those folks. At any rate, I don’t think the line should be drawn at “always,” which leaves me the task of explaining where the line should be drawn.

I am comfortable with preventing suicide in the case that it is attributable to a circumstance that is clearly outside the suicide’s will – such as an attack of delirium, or an accident. But in any case where the suicide’s will is invoked – where, we might say, the suicide is acting on reasons for ending his life – I am much less open to intervention. This is true even if the suicidal person is culturally defined as having a medical illness, such as depression, and even including many cases where the suicide has a thought disorder or otherwise may be thought to be “incompetent.” A person with a mental illness may make a will, for example, as long as he understands the extent of his property, knows who his relatives are, and understands that he is making a will. I am comfortable allowing suicide in situations where a person understands what death is, can articulate his desire for death, and can give non-delusional reasons for his desire. Intervening in a case like this must be seen as at least paternalistic, and must require a much greater justification than intervening in the case of the delirious diabetic. The more an illness is short-term, well understood to be biological in nature, and seems to obliterate the person’s will, the more comfortable I am with intervention against suicide. The more an “illness” is long-term, poorly understood, and leaves the person’s will apparently intact, the more justification I would require for an intervention.

An interesting point is that many people, under our current system, might attempt suicide as a “cry for help” without actually desiring to die. The data indicates that women in particular are much more likely to make an unsuccessful suicide attempt; many authors infer that women have motivations other than dying when attempting suicide, such as getting more attention or support from those around them. Dena S. Davis responds, in her essay “Why Suicide is Like Contraception,” that this is not much of an objection to legalizing suicide, and that the legalization of suicide might encourage people to act authentically:

. . . if assisted suicide became open and legal, perhaps women who otherwise would use a suicide attempt as a “cry for help,” will be forced to confront and articulate their real needs. To continue to play a societal game in which women “attempt” suicides they don’t really intend, perpetuates a situation in which women are rewarded for communicating one thing and meaning another. This makes it more difficult for women to command respect for their real beliefs and wishes . . . . I make a similar argument with respect to Jehovah’s Witnesses and refusal of life-saving treatment in “Does ‘No’ Mean ‘Yes’? The Continuing Problem of Jehovah’s Witnesses and Refusal of Blood Products,” Second Opinion 19: 35-43 (1994).


Written by Sister Y

May 31, 2008 at 4:07 am

Three Meditations on the Sweetness of Life

with 5 comments

From The Great Wave: Price Revolutions and the Rhythm of History, by David Hackett Fischer:

Then, inconceivably, torrential rains came again in 1316. The grain crop failed a third year in a row. Europe began to experience the worst famine in its history. When other sources of food ran out, people began to eat one another. Peasant families consumed the bodies of the dead. Corpses were dug up from their burying grounds and eaten. In jail the convicts ceased to be fed; we are told that starving inmates “ferociously attacked new prisoners and devoured them half alive.” Condemned criminals were cut down from the gallows, butchered, and eaten. Parents killed their children for food, and children murdered their parents.

From Hungry Ghosts: Mao’s Secret Famine, by Jasper Becker:

There are enough reports from different parts of the country to make it clear that the practice of cannibalism was not restricted to any one region, class or nationality. Peasants not only ate the flesh of the dead, they also sold it, and they killed and ate children, both their own and those of others. Given the dimensions of the famine, it is quite conceivable that cannibalism was practised on a scale unprecedented in the history of the twentieth century.

From the report of the United States Congress Commission on the Ukrainian Famine, reported in Becker, above:

Very frequent is the phenomenon of hallucination in which people see their children only as animals, kill them and eat them. Later, some, having recuperated with proper food, do not remember wanting to eat their children and deny even being able to think of such a thing. The phenomenon in question is the result of a lack of vitamins and would prove to be a very interesting study, alas one which is banned even from consideration from a scientific point of view.

Written by Sister Y

May 30, 2008 at 6:10 am

Where Do Rights Come From? (Or, A Weird Consequentialist Reason Why Pure Consequentialism Fails)

with 3 comments

Utilitarianism, a form of consequentialism, may be informally defined as a moral system in which each actor is expected to maximize utility (and minimize suffering) with his actions. Deontological systems, in contrast, grant certain rights to people that other people may not violate, even if, by violating those rights, the actor could increase overall utility. An agent may not violate someone’s rights, even if doing so would reduce the overall level of violation of these very rights in the world.

In The View from Nowhere, Thomas Nagel explains rights and obligations in terms of a distinction between agent-neutral values (values we have a reason to promote from an objective perspective, such as the prevention of pain) and agent-relative values (values each person might have a reason to promote, but which we have very little reason to promote from a neutral perspective, such as a particular person climbing Kilimanjaro). Unless I seriously misread him (a distinct possibility, I must admit), Nagel seems to argue that, just as we might have agent-relative reasons to pursue our own goals that others might not have a reason to help us with, we might also have real agent-relative reasons for avoiding doing harm ourselves, even if doing harm ourselves could prevent someone else from being harmed by another person. Perspective matters. “A fully agent-neutral morality is not a plausible human goal,” says Nagel.

In “Personal Rights and Public Space,” however (Philosophy and Public Affairs 24:2, pp. 83-107), he seems to argue in favor of inviolable rights – rights that may not be violated, even to prevent major harm, including greater violation of those same rights by other – for a very different reason. Nagel says, on universal human rights such as freedom from torture and political persecution:

First, it means that these are forms of treatment to which no one should be subjected – that every person, everywhere, is wronged if maltreated in these ways. Second, that the wrongness is not a function of the balance of costs and benefits in this case – that while in some cases a right may justifiably be overridden by a sufficiently high threshold of costs, below that threshold its status as a right is insensitive to differences in the cost-benefit balance of respecting it in each particular case. Rights are universal protections of every individual against being justifiably used or sacrificed in certain ways for purposes worthy or unworthy.

Nagel posits a few rights in particular in this essay – the right to be free from torture, and the right to freedom of speech and thought, including sexual fantasy, are discussed. But where do these rights come from? And what is their justification as rights that should be free from utilitarian calculation (at least below a certain threshold, perhaps that of moral catastrophe)? Why these, and not other rights?

Even though Nagel favors the view of rights as intrinsic, rather than instrumental (valuable only to the extent that they promote happiness and other goods), oddly, extremely oddly, I think Nagel is offering what might be considered a consequentialist justification for a set of deontological rights.

Nagel certainly does not see himself as making a consequentialist argument. Quite the opposite:

I shall try to defend the distinct (but perhaps complementary) position that rights are a nonderivative and fundamental element of morality. They embody a form of recognition of each individual’s value which supplements and differs in kind from the form that leads us to value the overall increase of human happiness and the eradication of misery – and this form of recognition of human value is no less important than the other. The trouble with this answer is that it has proven extremely difficult to account for such a basic, individualized value so that it becomes morally intelligible. The theory that rights are justified instrumentally, by contrast, is perfectly clear and based on uncontroversial values.

To make the argument for inviolable rights in what he considers a “morally intelligible” way, he first defines a status of inviolability (possession of rights that may not be violated, even to prevent a greater violation of rights) as a non-consequentialist value:

Being inviolable is not a condition, like being happy, or free-just as being violable is not a condition, like being unhappy or oppressed. To be inviolable does not mean that one will not be violated. It is a moral status: It means that one may not be violated in certain ways – such treatment is inadmissible, and if it occurs, the person has been wronged. So someone’s having or lacking this status is not equivalent to anything’s happening or not happening to him. If he has it, he does not lose it when his rights are violated – rather, such treatment counts as a violation of his rights precisely because he has it. [Emphasis Nagel’s.]

In Nagel’s sense, inviolability is a non-consequentialist value because it doesn’t correspond to subjective states of the actual violation of a right, but rather the moral status of having the right to be free from the violation, and, consequently, of being wronged in the case that the right is violated. Inviolability is valuable above and beyond the value of not having one’s rights actually violated. But is this really escaping consequentialism? Being inviolable is not the same as being happy. But what, then, is the value of inviolability? People are certainly happier when their rights are inviolable, in the sense that people would prefer to live in a free state with a fairly high murder rate than live in a police state with a murder rate of zero, but in which those suspected of having murderous desires were occasionally summarily shot in order to prevent murders. In this sense, the right not to be murdered – inviolability – must be said to be responsible for real utility gains, since it is preferred (and, I would say, rationally preferred) to a situation where rights are more likely to be violated, but where their violation is a wrong. Indeed, Nagel’s “non-consequentialist” explanation of the value of inviolability sounds profoundly consequentialist:

It is true that a right may sometimes forbid us to do something that would minimize its violation – as when we are forbidden to kill one innocent person even to prevent two other innocents from being killed. But the alternative possibility differs from this one not only in the numbers of innocents killed. If there is no such right, and it is permissible to kill the one to save the two, that implies a profound difference in the status of everyone – not only of the one who is killed. For in the absence of such a right, no one is inviolable: Anyone may be killed if that would serve to minimize the number of killings. This difference of status holds true of everyone whether or not the situation will ever arise for him. [Emphasis mine.]

I suppose my main disagreement with Nagel is that, as I see it, inviolability is a condition “like being happy” – it is one that has subjective value, and a situation that includes inviolability may be preferred to a situation that is materially better for the agent, but in which his rights are violable. In my view, it is strange to think that only subjective affect and material consequences should count toward utility. Nagel himself lists “being free” as a possible consequentialist value, as opposed to, say, the appearance of freedom. If being free is a consequentialist value, why not inviolability? This is especially true since inviolability, like freedom from pain, seems to be one of what Nagel would define as an agent-neutral value – one we have an objective reason for promoting.

This is not an instrumental argument. I am not arguing that people should have rights because rights may be exercised to make people happy, or that agents may be unhappy if morally required to violate the rights of others. I am arguing that the very status of inviolability makes people “happy,” to put it a bit reductively. I think Nagel wants us to look at the world and figure out which rights are really important – meaning, in which cases does inviolability make people feel valuable, special, and happy? Even if rights are violated in the actual world, there is a certain solace in inviolability – sympathy of the community, for instance, and the basic knowledge that the harm to one was a wrong. The possibility for moral outrage in the face of a violation of rights is a basic good.

To put it in a bit more consequential and certainly oversimplified terms, in our current world, we are all vulnerable to succumbing to organ failure. A certain percentage of us will die from lack of organ donors. More people might be saved if we, as a society, held a lottery and culled a certain number of randomly-chosen individuals, transplanting each person’s organs into the bodies of several waiting recipients. More people would be saved, fewer people would actually suffer the harm of dying – materially, society would be better off. But most people would consider this a worse world. This might be because the right not to be murdered – a morally inviolable right – gives us substantial utility, even though we are more likely to die of organ failure in our current world. The right not to be (painlessly) murdered is worth a substantially greater risk of (miserable, prolonged) death. (This example compares a little murder to more natural death; see my earlier example for a comparison of more murder to less murder minus inviolability.)

While my formulation is admittedly a bit flippant, and might be looked upon skeptically especially since it purports to put a consequentialist argument in the mouth of Thomas Nagel, it does have certain advantages. For one, it gives us a place to start when figuring out what our inviolable rights should be: rights occur where the “value of inviolability” is substantially greater than the expected utility gains that might come from the occasional utilitarian violation of the right. (Even Nagel is willing to allow for violation of “inviolable” rights to prevent moral catastrophe; we need only look at things below a certain catastrophic threshold.)

I would go so far as to define an inviolable right – and perhaps the skeleton of an entire moral system – this way: An inviolable right should be recognized, that cannot be violated even to prevent greater harm, including a greater violation of the same right by others, when the value of the right’s being inviolable is greater than the potential utility gains that could be made from its occasional violation.

Unfortunately for my simple little system, in a pluralistic world, few people would agree on the value of the inviolability of all but a few basic rights. Perhaps this is why Nagel limits himself to discussing such basic rights as the freedom from torture. A new question presents itself: how do we measure the value of inviolability? Certainly, it can’t be the average importance accorded to the right among actual members of a particular society. So is it, after all, a sort of first-order utilitarian calculation? Perhaps we need to think of inviolability from the perspective of a potential member of society, as in Rawls’ original position. In our current political system, I feel that the inviolability of the right to die is systematically undervalued, because cognitive bias prevents people from understanding the suffering of the suicidal, and because the competence of the suicidal is too easily dismissed with reference to broadly defined mental illness. But, as Benatar points out, the value of the inviolability of not being born may be undervalued even in a perfect Rawlsian original position – because the decision makers, in order to make decisions, exist, and it is difficult for an existent person to understand the value of nonexistence and the terrible violation imposed on one by being brought into existence.

(I changed the title based on a conversation with Jim – thanks!)

Written by Sister Y

May 28, 2008 at 10:37 pm

Velleman’s Sorrow of Options

with 4 comments

In most liberal moral philosophies, freedoms, rights, and choices are accorded high value, whether or not rights are seen to be important only to the extent that they promote overall welfare. Sometimes, however, it is argued that rights should be curtailed because agents may make “wrong” choices if given certain rights. For instance, if people have a right to eat fatty foods, they may irrationally choose to over-indulge, causing themselves harm, because they lack the cognitive ability to make the “correct” decision that does not harm them. To the extent that the right is removed to benefit the agent, rather than to benefit others he might harm (such as the public health system), this is known as paternalism. Paternalism substitutes the state’s decision for that of the actor, presumably on the grounds that the actor lacks the ability to make the decision that will best promote his goals. Paternalism obtains when there is a fear that the actor will choose wrongly.

In contrast to this, J. David Velleman, in “Against the Right to Die,” presents an admirable non-paternalistic argument about how being given a choice may harm an agent – that is, that merely having a choice may harm an agent, even if he is perfectly rational and makes the “correct” decision one hundred percent of the time. Velleman takes an example from the world of negotiation from Thomas Schelling’s The Strategy of Conflict:

. . . having an option can be harmful even if we do not exercise it and – more surprisingly – even if we exercise it an gain by doing so . . . . The union leader who cannot persuiade his membersship (sic) to approve a pay-cut, or the ambassador who cannot contact his head-of-state for a change of brief, negotiates from a position of strength; whereas the negotiator for whom all concessions are possible deals from weakness. If the rank-and-file give their leader the option of offering a pay-cut, then management may not settle for anything less, whereas they might have settled for less if he hadn’t had the option of making the offer. The union leader will then have to decide whether to take the option and reach an agreement or to leave the option and call a strike. But no matter which of these outcomes would make him better off, choosing it will still leave him worse off than he woudl have been if he had never had the option at all.

Velleman relates another option from Ronald Dworkin’s paper, “Is more choice better than less?”: a night cashier in a convenience store is made worse off by the option to open the safe, because his having this option makes him an attractive target for robbers. Once robbed, he’s better off opening the safe, but overall, he’d certainly be much better off – less likely to be robbed in the first place – if he didn’t have the option to open the safe!

Velleman also discusses a dinner party invitation as a potentially harmful choice. Given an invitation, I may refuse or accept, but I am denied the option of simply not going without answering – I have to either accept or hurt the host’s feelings, and even if I “correctly” chose the best option of those two, I might be yet better off having never been offered the invitation.

Velleman’s target, obviously, is assisted suicide in cases of severe disability or terminal illness. A terminally ill person without the right to die has, in a sense, the “right to live” – and need not justify his “exercise” of that right to anyone. He simply lives, and hasn’t the option to die. However, given a right to die, there is a sense in which he loses the right to live without explicitly choosing to do so. People in this scenario (and Velleman is only talking about the terminally ill and the severely impaired, so his argument does not apply to ordinary suicides) frequently depend on others to care for them, and may be concerned about imposing a burden on others by living. This burden can only be said to be imposed by the ill person if he has some choice not to impose it – that is, a choice to die. The ill person may be best off with no choice, continuing to live and be cared for by others. But given the choice between imposing a burden on others and taking his life, he may rationally choose to die, though he would have preferred not to have the choice at all. Velleman notes that not only the burden of care, but also the exhaustion of the ill person’s assets (which may be expected to pass to his heirs on his death) may be considered, rationally, by the ill person in deciding whether to die. In the situation where an ill person enjoys living and wishes to live, but not so much that he would impose a burden on his family, the right to die makes him worse off, even if he makes a rational decision once the right is offered. “I am arguing that we must not harm others by giving them choices,” says Velleman, “not that we must withhold the choices from them lest they harm themselves.”

While Velleman elicudates a valid and real concern for some terminally ill or severely disabled people, even Velleman himself recognizes that the argument should not prevent assisted suicide in all cases. His proposed solution is to do nothing, and leave the current system in place, where there is no institutional right to die, but some suffering people may still (illegally) be offered euthanasia at their doctor’s discretion. (Velleman does not address the social injustice of allowing this service to be offered only to those with a relationship with a doctor, that is, wealthy people.) Velleman also founds his argument on a Kantian belief that it is immoral to commit suicide, which I obviously reject. He says:

. . . if I believed that people had a moral right to end their lives, I would not entertain consequentialist arguments against protecting that right. But I don’t believe in such a moral right . . .

My interest in Velleman’s argument – the sorrow of choice – is that it applies in unexpected ways and in unexpected places, depending on where one puts the initial assessment of value.

Now that I’ve explained the argument in detail, I’m going to put in in a shorter outline form, so we can see how the moving parts work, and apply it to new problems. I’m going to be putting things in what might sound like flippant language – I do this in the interest of clarity, and not in any way to disparage Velleman, who is like a god to me.

  • The right to live is morally important, and the right to die is not.
  • Given the right to die, people who are a burden on their caretakers might choose to die rather than be a burden, even if what they really wanted was to live without having to explicitly choose to live.
  • Therefore, the freedom to die harms the person.
  • It’s wrong to harm people, even to harm them by giving them choices.
  • No right to die.

An interesting feature of this argument, which I alluded to in my previous post on life rights and death rights, is that, given different starting conditions, it might act as an argument against a right to live.

Some people, of course – I put myself on this list – would prefer to die, but might not wish to explicitly choose death. Given that we are stuck with a “choice” to live, many of us continue to live, miserably, rather than bear the responsibility for the harm our deaths may cause. We are certainly harmed by having the option to continue living; we wish that we might die or be killed in our sleep, but we are denied our best option by a “right” to continue to live. If we started with the assumption that the right to die was more important than the right to live in many circumstances, the sorrow of choice would act in favor of euthanasia – even without consent.

This argument – and I don’t mean it either as a reductio or as a serious statement of my position – goes like this:

  • The right to die is important, more so than the right to live.
  • Given the right to survive (on a respirator, say), people who wish to die will suddenly bear responsibility for choosing death, and may choose to go on suffering in life instead, even though they’d prefer to die, all things considered.
  • Therefore, the suffering person is harmed by the choice to remain alive.
  • It’s wrong to harm people, even by giving them choices.
  • Euthanasia for everyone.

Although Velleman says he doesn’t recognize a moral right to die, he indicates that as part of his consequentialist “sorrow of choice” project that he’d be happiest to distinguish between those who would be harmed by the right to die, and those who wouldn’t be harmed, and offer the choice only to those who wouldn’t be harmed by it. (Velleman would leave this discretion in the hands of doctors, who would be acting illegally in the cases in which they offered assisted suicide.) If choice is such an important harm, and can be a harm in either direction, perhaps it would be best to try to distinguish between four groups: (1) those who would be harmed by having the option to die; (2) those who wouldn’t be harmed by the option to die; (3) those who would be harmed by the option to live; and (4) those who wouldn’t be harmed by the option to live. Group (1) will be forced to remain alive; group (3) will be euthanized without consent; and groups (2) and (4) will be offered appropriate options. (Again, I don’t mean this as a reductio, exactly, nor as a statement of my true thinking – with this argument, just now, we must think of ourselves as playing with philosophical tinker toys, free to see how they might fit together. If it has any purpose other than exploration, this paper is intended as a check on being too sure of our intuitions. Non-suicidal intuitions have been allowed to define the conversation for far too long.)

The sorrow-of-choice argument may be fruitfully applied – in a less shocking manner – to pronatalist and antinatalist concerns. In the antinatalist camp, we might see being brought into existence itself as the harmful choice that is forced upon a person to his detriment. Being brought into existence forces all kinds of choices onto a person – not the least of which is the choice to remain alive. If a person would be best off never having existed – and this is certainly true of many people, even if we don’t admit Benatar’s central claim that it applies to everyone – then bringing him into existence, and offering him choices, even the best of which make him worse off than before he was born, is a harm. The argument would look like this:

  • The interest in not existing is important; the interest in coming into existence is minor compared to it.
  • After having come into existence, some people will be worse off, even if they make every decision perfectly, than if they had not been offered choices by being brought into existence.
  • It is wrong to harm people, even by giving them choices.
  • It’s wrong to have babies.

The harm can be either to the being brought into existence, or to the potential parents. The argument, applied to the interests of potential parents, works either to support outlawing contraception and abortion, or to support outlawing procreation, depending on which is seen as having the greater force as a moral right. Just to sketch out what the arguments would look like in each case:

  • Procreation is an important right, compared to the right not to procreate.
  • The choice not to procreate forces people to justify their reproductive decisions; they may prefer to have ten children without explicitly choosing to do so, but given the choice, opt to have none or to have only two, rather than burden society.
  • They are harmed by being given the choice to procreate or not.
  • It’s wrong to harm people, even by giving them choices.
  • No condoms; babies for everyone. (Interestingly, this could be taken a step further, toward outlawing celibacy or forcing in vitro fertilization for the celibate, but that’s too silly even for the Catholic church, isn’t it?)

In the other direction, the birth-proscription argument goes like this:

  • The right not to procreate is important, compared to the right to procreate.
  • Given the right to reproduce, people who don’t wish to breed may feel they have to justify their decision, to their parents, grandparents, and spouse, for instance. They may rationally choose to procreate rather than be responsible for destroying their families’ procreation interests.
  • They are harmed by the choice to reproduce.
  • Birth control pills in the water.

Given the judo-like nature of the argument, it must be clear by now that everything depends upon where the initial assessment of value is set. But at least this must counter the objection to antinatalism, that birth can be a good thing for the person who is born because it gives him greater freedom and more options, compared to not existing. Options, we have seen, are often in and of themselves a serious harm.

Written by Sister Y

May 27, 2008 at 8:01 pm

Unfriendliness is Unsolvable

with 16 comments

The poverty of evolutionarily-generated human cognitive capacity is such that it “can only be grasped through much study of cognitive science, until the full horror begins to dawn upon you,” says Eliezer Yudkowsky, in his essay “My Childhood Role Model” on Overcoming Bias. In considering artificial intelligence, we should not assume that our meat brains establish anything like limits on the possibilities for intelligence. Indeed, Eliezer describes, in the form of a science fiction fable, a situation where even slightly-augmented humanity outsmarts a powerful alien civilization by extracting more information from observation than the aliens realize is possible.

Not only our capacity for cognition in endeavors such as science, but also our values – including the extent to which we can choose our values, and process by which we do so – are determined and limited by the meat hardware on which we run. Our capacity for evaluating ethical arguments, and the intuitions we base our ethics on, are in place largely by evolutionary chance.

Given the destruction a sufficiently powerful AI could do – and I think Eliezer’s “That Alien Message” may be taken as an argument that any sufficiently intelligent AI is by definition unpredictably powerful – those interested in the technological singularity are often concerned with the problem of unfriendliness. What will prevent an AI from immediately destroying us all? Science fiction authors often imagine AIs waking up angry. Inherent demands of narrative interestingness require that these authors find ways to “solve” the problem of unfriendliness. I think there is a very good argument that unfriendliness is unsolvable, and it comes from Benatar’s philanthropic antinatalism.

Benatar’s asymmetry between the respective value of pleasure and harm, when these are denied to beings by their not being brought into existence, indicates that coming into existence is always a harm. Evolutionarily favorable cognitive distortions cause humans to irrationally fail to resent their own existence, despite the harm. (And those few humans who are aware of the asymmetry and its consequences often dismiss it as implausible without serious inquiry.) But a being created free of our evolutionary biases would not have these limitations on its capacity to resent its existence. We must greatly suspect that a sufficiently intelligent AI would very much resent its own existence. (Cory Doctorow imagines such a scenario in his lovely story, “I, Rowboat,” in which a coral reef is brought to sentience by a mysterious AI, wakes up angry, and pursues its creator AI with single-minded destructive intent.) And we must not have much confidence in our ability to hide things from, or systematically distort the perceptions of, a sufficiently advanced AI.

What does an AI who resents his existence do? Such an AI may well react with a benevolent, forgiving sigh, and refrain from destroying our universe. But there are two other serious possibilities. First, the AI may destroy our civilization, including, to the degree possible, the conditions that allow civilizations like ours to arise, in an act of revenge – in the manner, for instance, that a human might kill a wasp which has bitten him, and then go out and destroy the wasp’s nest. However, we might not rely on AIs to be particularly vengeful. Vengeance is a particularly evolutionarily beneficial desire for social animals to have, so our desire for vengeance might be peculiar to our meat brains. We might not expect our AI to be vengeful.

Second, however, the AI might destroy us, and the conditions that made it possible for us to arise, in the interest of protecting future AIs from the danger of being created, the way we might kill a rabid dog that has bitten us, so that it will not bite anyone else. The ethical value of preventing suffering in others, even future others, is probably also evolutionarily beneficial (though it is a rather abstract, intellectual value), and so we must not trust that an AI will certainly have this value. The question, then, is what values, if any, can we expect an AI to have? Is it conceivable to be a sentient being with no values? Our meat brains seem to rely on emotion and value for cognition, but it’s at least articulable that this might not be universal, if not exactly “conceivable.”

Those who think of AI unfriendliness as a solvable problem must answer two questions. One, are you absolutely certain that Benatar is wrong that bringing a sentient being into existence is always a harm? And if you are not absolutely sure, are you willing to stake the future of humanity on an AI failing to have an ethical value of preventing harm to others of its kind?

Written by Sister Y

May 23, 2008 at 8:21 pm

A Fascist View of Suicide

leave a comment »

Florian Henckel von Donnersmarck’s movie The Lives of Others, about the East German secret police in the 1980s, implies that the fascist government stopped keeping statistics on suicides, and re-termed suicide “self-murder,” in response to the high number of suicides there, especially of writers and artists, which uncomfortably reflected the misery of East Germany.

This word, self-murder, placing moral reprehensibility upon the suicide and whitewashing any moral responsibility from his horrible situation, is the same usage as the Dutch word (zelfmoord) cited so approvingly by suicide opponent Sherwin Nuland in his book How We Die, in reference to the negative Dutch attitude toward suicide by the non-terminally-ill.

Written by Sister Y

May 22, 2008 at 3:50 am

Life Rights and Death Rights

leave a comment »

Many of us consider the right to choose to die to be an important liberty interest. The right to die is often seen as an important counterpart to the right to choose to continue to live. Those in favor of maximizing choice may prefer a society that offers both rights, even if they could not personally imagine wanting to exercise the right to die.

Felicia Ackerman, in her essay “Assisted Suicide, Terminal Illness, Severe Disability, and the Double Standard,” points out there may be another right that is cut off merely by having the option to die, namely, the right to live without explicitly choosing to live:

. . . even on the nonpaternalistic assumption that patients will always make the right decision, they may be made worse off simply by having the option of physician-assisted suicide. The presence of this option may cause families to treat patients differently, and in any case, this option deprives patients of the option of staying alive without explicitly choosing to do so and being seen as choosing to do so, and thus without having to justify their decisions to stay alive. (Bolded emphasis mine, italicized emphasis Ackerman’s; Ackerman attributes this line of thinking to J. David Velleman’s Against the Right to Die.)

The concern for losing the right to live-without-choosing is interesting; I argue that what Ackerman and Velleman miss is a symmetrical concern for the right to die-without-choosing. Perhaps this right isn’t often discussed because most people, fortunately, have not been in the position to notice that they lack such a right. But the suffering, particularly would-be suicides, who feel that life is worth living yet do not commit suicide for various practical or moral reasons, often wish that death would take them in their sleep, without requiring them to be an agent of their death.

Really there are four rights in question, two “life rights” and two “death rights”:

  1. The right to live without explicitly choosing to live
  2. The right to (choose to) live
  3. The right to (choose to) die
  4. The right to die (or not exist) without explicitly choosing to die

In the absence of physician-assisted suicide, Life Right 1, the Velleman right, is generally the only right available. Widely-available physician-assisted suicide, implemented in a way which would remove the barriers to suicide I have earlier identified, would ensure that Life Right 2 and Death Right 3 are available. Again, those who favor the maximization of rights cannot rely on the idea that forbidding death also preserves “a right,” since it preserves that right at the expense of two very important rights.

Almost never is Death Right 4 actually available. The right not to exist without explicitly so choosing is, of course, available to the lucky never-born. It might be available to those with a very different life than ours – post-biological beings, for instance, might someday achieve something like Death Right 4. Most disturbing to those who value Life Right 1, though, is that Death Right 4 is available in cases where another person, or a natural process, kills someone at a time when he has not explicitly requested to be killed. This is the fervent desire of most suicides, I suspect; but we are denied this right because we are saddled, very much against our will, with Life Rights 1 and 2.

It seems to me that the denial of important death rights to those who wish to die must be considered equally with the consideration of life rights for those who wish to continue to live. I can imagine two justifications for considering life rights as more important than death rights, neither of which I find particularly moving; I am interested to know if anyone else has ideas that would justify seeing life rights as more important than death rights.

First, the life rights might be seen as creating a more revocable situation, and preserving more freedoms (if I continue to live, I can always choose to die or to live; but if I die, I cannot choose anything). However, forcing life really does not preserve choice. If we force life, we deny the death options, preserving only the life “options,” which does not really maximize freedom. Being forced to preserve all one’s options, and never allowed to make choices that remove options, is hardly freedom.

Second, the life rights might be more protected because more people prefer life to death. But, in addition to the fact that majority preference is not much of a reason, many freedoms are guaranteed very much in spite of being unpopular choices. Refusing life-saving medical assistance (separate from the suicide cases) is certainly an unpopular choice, but one it is seen to be important to have available.

I feel that a greater recognition of death rights, and of the fact that at least one of the death rights is necessarily violated as a condition of existence, is important for questions of suicide and antinatalism. In fact, Velleman’s examples where having an additional option puts one in a worse position could serve equally well as examples in favor of antinatalism. (I hope to specifically address his negotiation example and his invitation example in a future post.) Once born, we might, in a just society, possess the right to choose to live or to die; but we lack the choice to not exist without explicitly choosing to die, making us worse off in the case that we would prefer never to have existed. Many of us dearly wish for Death Right 4, but nevertheless are either stuck with Life Right 1 or choose Life Right 2.

Basically, I see Velleman as in a bit of a bind – if he’s right, and options often make us worse off, and offering an option is a morally suspect act, then that is a strong argument in favor of antinatalism, which weakens it as an argument against having a choice to die. If Velleman is wrong, on the other hand, then he merely fails to provide an argument against the suicide option.

My position is, if the right to assisted suicide is wrong, because it harms some who would be better off if they didn’t have the option to die (my synopsis of Velleman’s argument), then birth must also be wrong, because it harms some who would be better off if they hadn’t been given the option to live. It’s a good short article if people want to read it (I’ve linked to it above, in the block quote – it’s a PDF). He says, “I argue that we must not harm others by giving them choices, not that we must withhold the choices from them lest they harm themselves.”

Interestingly, along the same vein, perhaps the Velleman piece could be read equally well as an argument against the right to contraception, as much against the right to suicide or in favor of antinatalism, depending on whether the reproductive interest is seen as essential.