Open threads at the Open Thread tab every Sunday and Wednesday

In Defense of Psych Treatment for Attempted Suicide

A lot of the comments in my recent post on the implicit association test asked for a defense of why society should be hospitalizing suicidal people in the first place. If people have, after much thought, decided they prefer death to life, isn’t that their right?

I am extraordinarily sympathetic to this position, which has been most eloquently defended by Sister Y of The View From Hell. Sister Y lists many harmful effects of suicide prohibition and many reasons why rational people might want to end their lives. She suggests a policy of legalizing fatal doses of barbituates for people who want them, allowing people tired of existence to leave the world without grisly suicide attempts that might leave them permanently injured or cause collateral damage to bystanders. I can’t find her opinion on whether these should be provided on demand or whether you should have to undergo a psychiatric assessment first.

If she in fact believes the latter, then I think that position is defensible, and for professional reasons I won’t publicly say anything further than that. But this post is to explain why it should require one hell of a psychiatric assessment and why the overwhelming majority of real-world suicide attempters would and should fail such an assessment.

Again, my point of disagreement is not on the ethics involved of letting some hypothetical perfect philosopher commit suicide – nor even on the fact that perhaps some cases genuinely are these perfect philosophers including Sister Y herself. I am trying to emphasize the practical point that in the real world, attempted suicides are rarely perfect philosophers and almost always people who have made sudden, impulsive, and very bad decisions.

The greatest burden of suicide is of course to the friends and family of the person involved. But you don’t have to be a Randian to think it’s morally abominable to require someone in pain to continue living solely to please other people, so this post will focus solely on the welfare of the person involved.

What Does Youth Suicide Tell Us About Adult Suicide?

Start with the clearest case. About 4% of teenagers attempt suicide at some point (there are some much higher values from the CDC, which is usually pretty trustworthy, but some of the comments point out reasons why their estimates here are pretty hard to believe.)

I’ve gotten to observe some teenagers admitted to hospitals for attempted suicide. Some have incipient mental disorders that no one has noticed or considered treating. Some have unbearable home lives. Others have the standard litany of teenage problems – broke up with their boyfriend/girlfriend, bullied by the popular kids at school, got into a fight with their parents. Many have a combination of all three.

Some were the classic “cries for help” that were never meant to actually end in death, but others were entirely serious. A tragic few intended to take enough of an overdose to make Mom scared but not enough to actually kill them, but muddled their pharmacology in the most permanent possible way.

And I think most people agree that teenage suicide is terrible and requires treatment. Heck, most people won’t even let teenagers make the decisions of whether or not to purchase alcohol, let along the decision to end their own lives. But I think aside from the inherent tragedy of teenage suicide it illuminates something about adult suicide as well.

These people have only the tiniest glimmer of knowledge about the likely happiness of their future lives. Most of their problems – the bullying by popular kids, the failed first relationship, even the awful families – are eminently wait-out-able. “It gets better” is not just for gay people (who, by the way, have a suicide rate up to 15x that of the straight population).

And yet teens attempt suicide at staggering rates.

There are certain depths of despair dark enough that the knowledge that the despair is completely temporary cannot penetrate them. It is this state, defined by the clouding of rationality by suffering, that I think most teenage suicides occur in.

And it would be very strange if this suddenly changed as soon as the victim hit eighteen.

Connection Between Suicides And Mental Disorder

It is generally reported that about 90% of suicides have some mental disorder. No, this isn’t an artifact of psychiatrists assuming anyone who commits suicide must have a mental disorder – various half-decent methodologies have all converged around the same number, including a multitude of controlled studies (where psychiatrists evaluate a subject’s mental status based on notes before knowing whether the person committed suicide) and prospective studies (where people only count as mentally disordered if they were diagnosed before the suicide occurred).

Sister Y has tried to poke holes in these statistics. First, she noted that the controlled studies showed 37% psych diseases even in the control population. But this number is probably correct – NIMH estimates that about 26% of people have mental disorders in a given year, and no doubt that number is significantly higher among people who make good controls (ie are matched on demographic factors) for suicides. Second, she pointed out that the number included what she considered relatively “minor” disorders like alcohol dependence.

So first of all, alcohol dependence probably septuples your chance of committing suicide and something like 25% of suicides include alcohol. So I don’t think it’s unfair to include that in the list of how suicide is influenced by mental disorder.

But second of all, let me give totally anecdotal and probably unrepresentative examples of some other ways mental disorder can affect suicide.

As stereotypical as it sounds, the voices in people’s heads do tell them to kill themselves a lot. Voices in people’s heads are huge jerks and occasionally people will do what they say just to make them shut up. The tragedy here is that antipsychotic drugs are pretty good at dealing with this if people can just get access to them. Among schizophrenia patients (the group most commonly identified with these sorts of symptoms), almost half attempt suicide and 10% complete it. Since schizophrenics make up 1% of the general population, that’s a non-negligible fraction of total suicides.

You know what’s an even less fun form of psychosis? Psychotic depression. This is where people get so depressed they start hallucinating about how horrible they are. I will never forget the patient who stopped eating because she believed her digestive system was rotting away and infested with maggots. And a lot of the time these people’s self-hatred reaches completely bizarre proportions in which they will confess to causing the Holocaust or the 9-11 attacks just because it seems like the sort of thing someone as horrible as them might do. If you believe you caused the Holocaust, this seems like a pretty good reason to kill yourself in the name of justice, and sadly this is what many of these people do. And again, this is tragic because psychiatry is actually not so bad at dealing with this kind of over-the-top depression (the rotting-intestines woman became much better after a short course of electroconvulsive therapy, but some people will get better just on medications).

Borderline Personality Disorder is another common cause of suicides. It intensifies emotions so that anyone so much as making a mildly critical remark makes you think everyone will hate you forever and you deserve to die. And then six hours later someone smiles at you and you feel like the world is perfect and beautiful. But if you commit suicide at one of the low points, then that’s it. And Borderline Personality Disorder, again, is sorta amenable to therapy, and even without therapy half the time it just goes away after a few years to a decade.

Alcohol and drug abuse is another big one. Some of it is that abusers have worse lives – poor health, financial issues, more likely to have trouble at work. But a big part of it is just lowered inhibition. If a sober person is walking on a bridge after some life crisis, they might have fleeting thoughts of jumping but suppress them after thinking of the future. If a drunk person is walking on a bridge after some life crisis, the frontal lobes that would normally suppress those urges are partly out of commission.

And then there’s depression. I’m trying not to make a big deal about it because everyone associates suicide and depression when in fact the correlation is no higher than many other mental illnesses (although the greater number of depressed people does make absolute numbers higher). I guess all I’ll say here beyond what everyone already knows is that Major Depressive Disorder (classic depression) is an intermittent disease. The average depressive episode lasts less than six months, and the average person with MDD has only four depressive episodes in their lifetime (these numbers are even better if you’re on medication, which many depressed people fail to be). There’s a thing called dysthymia, which is like having depression all the time, but it is thankfully less common and less severe and not where most suicides are coming from.

I am certain that six months feels like an eternity if you are depressed. And no doubt knowing that you’re going to have to deal with the same thing a few more times in your life (ALTHOUGH SERIOUSLY, MEDICATION DOES HELP WITH THIS) must also be, well, depressing. But the average depressive suicide is not a Perfect Philosopher who has calculated, while healthy, that the possibility of another six month depressive episode is too much to bear.

The average depressive suicide is someone in the middle of one of their episodes who, like the teenagers above, is in the place so dark that they’ve forgotten the existence of hope. They’re somewhere so dark that “this will probably go away in a couple of months” has no meaning. Somewhere so dark that one of the main side effects of effective antidepressant drugs is suicide, because a few weeks after starting the patient finally has enough energy to go kill themselves, but doesn’t consider waiting a month or so for the drug to take full effect.

I want to end this section with a study – small, but encouraging – that cognitive-behavioral therapy (aka That One Type Of Psychotherapy That Sometimes Works) reduces suicide 50% in at-risk populations. Think about that. What percent of suicides do you think haven’t had cognitive-behavioral therapy? 80%? 90%? Whatever that percent is, half of them would have been fine if they had just had access to a good psychologist.

Empirically, Suicides Regret It

People who commit suicide can’t change their minds. But attempted suicides can and do, and we can analyze these changes both in their actions and in their words.

In terms of revealed preferences, most people who are prevented from completing their suicide do not go on to kill themselves. Sister Y critiques a study saying only 4% later go on to kill themselves, and offers as counterpoint a study she prefers claiming 13% do (she finds a way to round up to 19%). I have also heard 10%, although I can’t remember where. Do you know what the numbers 4%, 10%, 13%, and 19% all have in common? Yes. They are all significantly less than 50%.

It is somewhat harder to find good studies on what percent attempt suicide again. By eyeballing some other statistics and trying to fit them together, I believe it is greater than 25% but less than 50%. One textbook whose studies I have not been able to verify says that 30% of untreated and 15% of treated suicide attempters try again. 15% and 30% are also among the many numbers that are less than 50%.

And keep in mind what these data don’t show. They don’t show that the 25-50% who try again have lives so constantly miserable that they continue wanting to die. Remember that intermittent depression from before? Imagine a world in which depressive episodes last one day each, and people only have two of them in their lives. Other than those two days, they live happy lives and are grateful to be alive. Doesn’t matter. This pattern would still be consistent with 25-50% of attempted suicides making repeated attempts, if that second day of depression was bad enough

(out of fairness I should mention this data also doesn’t show that 50-75% of people get over their suicidality; it’s consistent with them just being tired of suicide attempts not working and settling for continued existence. I guess what I’m saying is that the data don’t prove very much)

So moving from boring data to the much-more-fun domain of anecdote, a surprising number of suicide attempters change their mind during the suicide attempt. One particularly famous case is that of Kevin Baldwin, who survived jumping off of America’s favorite suicide spot. He says that while still in the air “I instantly realized that everything in my life that I’d thought was unfixable was totally fixable—except for having just jumped.”

Most realizations are slightly less dramatic, but my work in a psych ER taught me that many 9-1-1 calls about suicides are from the victims themselves. I remember one patient, a typical case, who overdosed on pills. As she lay on the ground starting to feel sick, she thought about her problems a little more deeply, thought about how her family would feel, and decided she preferred to live. She called 9-1-1, they sent an ambulance over, and the hospital managed to keep her alive until the drugs passed out of her body. This is quite common. It also contradicts one of Sister Y’s strongest arguments – that the reason many people avoid suicide is out of fear of making the attempt. A non-negligible number of people who have already made the attempt and just have to sit back and day find themselves changing their minds and actively working to save their own lives.

But most of the stories I can generate from my personal experience are nothing more dramatic. It’s people who were found by their parents or partners or friends, dragged kicking and screaming to the hospital, treated for a couple of days, and by Day 3 they’re saying oh my god I made a horrible mistake I can’t believe what almost happened.

And I know what the response will be – that of course they’d say that to their psychiatrists, they’re trying to get judged Officially Sane so they can get discharged and maybe try again. I accept that as a possibility, but since this whole section is about totally useless anecdotal data, let me just say I don’t feel like that was what was happening. I met people who were going out of their way to look for and thank their psychiatrist when he was busy in his office after the discharge papers had already been signed and they were on their way out. One time I met a patient at the bus stop a few days after she had been discharged, and she asked me to thank my boss and the rest of the team for what must have been the umpteenth time.

Finally, I have some personal friends who have attempted suicide. In every case I am incredibly glad they remain alive, and more importantly, usually they are as well. And I know there’s social pressure here – that psychiatrists aren’t the only ones you have a vested interest in appearing cheerful to – but some are very close to me indeed and I do not believe they would lie about something this important.

Psychiatric Care Probably Helps

One of the most common objections to sending people who attempt suicide to psychiatric hospitals is that it is a terrible punishment, that we are essentially locking up and drugging and torturing people whose lives are already apparently pretty bad.

But mental hospitals for people who attempt suicide (actually almost always just the psychiatry floor of a regular hospital) are not like that one in One Flew Over The Cuckoo’s Nest. I can’t repeat that enough. I know that as a psychiatrist-in-training I have no credibility on this issue, so take it from a former psychiatric patient. Your problems are much more likely to be along the lines of a terrible selection of books in the ward library than torture by sadistic nurses (I do not deny the latter occurs, just as some schools have torture by sadistic teachers, but it is extremely rare and nowhere near for example what goes on in nursing homes).

According to the CDC, the average length of stay in a mental health ward is one week (this brief by an organization I’ve never heard of says 8 days). That includes catatonic people and people who have long animated conversations with the Devil, so the average suicidal person isn’t going to be the one bringing up that average.

In practice I have a pretty good guess for the exact length of stay the average suicidal person without associated mental disorders will experience, and that is 72 hours. That’s the maximum amount of time a hospital can legally commit someone against their will. After that they have to get a court order allowing them to hold the patient longer, and this requires swearing that the patient is mentally incompetent to make their own decisions, and most doctors will not do this without reason.

But if you don’t trust doctors’ benevolence, at least trust their self-interest: it takes a lot of paperwork, it requires them to go all the way to a courthouse, and the hospital management is going to be breathing down their back the whole time about how they could really use an extra bed on Ward 4 and of course we would never pressure you to discharge any patients before they’re better, but seriously, have a bed open on Ward 4 by tomorrow. Trust me, doctors are not plotting to keep people in the hospital longer than necessary. If you like conspiracy theories, the opposite conspiracy is a much bigger cause for concern.

That’s usually just enough time to evaluate the patient for mental disease, start them on some medication, and refer them to an outpatient psychologist and/or psychiatrist. One hospital I worked at kept (mostly willing) people in a little longer to see if the drugs actually took effect, but that was a luxury they could only afford because they were a rich academic institution.

But the thing is, this really helps. If 90% of people committing suicide have some associated mental disease, and mental diseases can dectuple your risk of committing suicide, then connecting these people – many of whom have never interacted with the mental health system before – with someone who can help them (or even with a Prozac prescription) can be a really, really big deal.

I mentioned before that one specific form of therapy can decrease future suicide rates 50%. That was in a study where both groups were getting the recommended psychiatric drugs. Another study I cited above said that “psychiatric treatment” (whatever that means; I bet it didn’t include the CBT from the last study and so they’re cumulative) can also cut future suicide rates in half. There are more specific studies on the anti-suicide effect of each individual drug – lithium is an example of a particularly good one.

(fun fact which there is a small chance I will devote my life to studying: even areas with slightly higher trace amounts of lithium in the water supply have lower suicide risk.)

And even if you’re one of the depressingly high number of people who throw away their prescription and never show up to their psychiatrist, you know what? You’ve been stuck in a big building with lots of people watching you for the three days or so immediately after whatever horrible event made you become suicidal in the first place. Drugs in your system? Now you’re clean. Angry at a family member? Maybe you’re less angry now. Upset over a breakup? Maybe you’ve had a chance to think about it a little more.

I am very reluctant to get into in what situations I believe suicide is acceptable. I am scared that one day my future employers will read this post. Or worse, a future patient will read it and start arguing “You said suicide was acceptable if A or B, so I did those things”. So all I will say is that I wish Sister Y and those like her maximum utility however they define their utility function. But anyone considering suicide who has thought about it less than she has or lacks her philosophical acumen should consider getting professional help (or even non-professional help) or at least meditate long and hard on that cliche about “a permanent solution to a temporary problem”.

EDIT: Since people are missing something I said like a thousand times in the post itself, I’ll put it down here in bold. I am not claiming that suicide is never rational and that all suicides are stupid and impulsive, or that no one can ever legitimately want to die. I am saying those people make up a very small portion of suicides, and that the typical case is people who do it impulsively or in a state where they lack full decision-making capacity. And that the psychiatric system can be of huge help to this latter group, and that helping the former group is a different question which I do not want to talk about publicly for professional reasons.

This entry was posted in Uncategorized and tagged , . Bookmark the permalink.

96 Responses to In Defense of Psych Treatment for Attempted Suicide

  1. anon1 says:

    It seems to me that most of these problems could be solved by some sort of waiting period.

  2. Jackson says:

    “If a drunk person is walking on a bridge after some life crisis, the frontal lobes that would normally suppress those urges are partly out of commission.”

    Ah! That’s what I’ve been doing wrong. Keep forgetting the damn alcohol. Thanks!

  3. Joe from London says:

    I don’t know if this has been suggested before, but a time-delay suicide clinic would solve a lot of problems. You go there one day and inform them that you wish to die. One month later, you may turn up and have your consciousness terminated.

    • Scott Alexander says:

      That would certainly be a part of the suggestions I don’t want to post. But I would suggest longer than one month. The average depressive episode being six months long seems pretty relevant here.

      • arrowpker20 says:

        Can you get in trouble for even implying that you think this is a decent idea?

        • Scott Alexander says:

          See the last paragraph of the post. I don’t think it would be a huge problem, but why tempt fate?

      • Ted Seeber says:

        I’m from Oregon. I disagree with you on euthanasia- but if there was ONE IMPROVEMENT I’d like to see in our euthanasia law, it would be this.

        Though I’d set it at 4 months (since to even get evaluated for the euthanasia law in Oregon, you need to have a terminal diagnosis with an expected life span less than six months).

        Or maybe an option for at 9 months *with hospice care and pain medication* as well as *anti-psychotic medication if required*.

        By 9 months you’d be over any potential clinical depression AND will have a 2nd opinion from nature on your terminal illness.

        • Scott Alexander says:

          I like the symmetry. You have to wait nine months to create a life, and you have to wait nine months to end one.

  4. cool rich guy says:

    Fantastic post as always. I think suicide is rational in cases where the rest of one’s life is likely to be one of constant pain with very little pleasure. E.g. someone with a terrible illness, someone being tortured to death in a POW camp, a hypothetical person who may or may not exist whose brain is wired so they are incapable of ever not feeling sad. It seems to me that, as you have shown, the vast majority of suicides do not fall into this category.

    “Start with the clearest case. According to the CDC, every year about 8% of teenagers attempt suicide. Every single reference to this study gets it wrong saying that the total number of teenagers who attempt suicide is 8%, but if it’s 8% per year then barring some unrealistic assumptions the total must be much larger (but is surprisingly unavailable).”

    I notice that I am confused. My high school had 400 people. So that would mean 32 attempted suicides per year. I imagine that at least 1 in 20 teen suicide attempts are successful, so you would think that on average, someone would die every year. However, I know of no one from my high school that has successfully committed suicide. I went to private school, so it’s possible that my school is atypical and that poorer, public schools have a much higher suicide rate. Even so, public schools are often enormous, and it seems hard for me to imagine that there could ever exist a school where at least one person kills themselves regularly every year. But now that I think about it, maybe this is naive of me, and maybe in some high schools, suicide is a regular occurrence?

    • Eric Rall says:

      That struck me, too. If the statistics cited are at all accurate, that implies that most people who attempt suicide are astoundingly bad at it.

      Which in turn tells me one of three things:
      1. There’s something wrong with the statistics,
      2. Suicide is much, much harder than my intuitions suggest, or
      3. Wanting to attempt suicide is a strong predictor of not being able to do so effectively.

      #3 is consistant with the thesis of the original post, that a large portion of suicide attempts are linked to various forms of severely diminished judgement, but for which suicide would not have been attempted.

      • cool rich guy says:

        My guess is that the vast majority of people, especially teenagers, who attempt suicide do not really want to die, i.e. of all the conscious, subconscious, unconscious, etc. decision modules in their brain, many are deciding to stay alive. Please do not pattern match this to “suicidal people are whiny emo faggots who have no problems and just want attention”.

      • Brian Delwiche says:

        Seems to depend mostly on method. This paper breaks it down; notice that suicide by poisoning is by far the most common method and also one of the least effective, with a success rate of about six and a half percent.

        • Mary says:

          To be sure, choice of method probably has something to do with intensity of the desire to die. Especially if you are “committing suicide” in order to emotionally blackmail the woman you are living with and have just battered, or to both evade a court appearance and make yourself sympathetic when you must appear.

          (More here.)

    • Douglas Knight says:

      I don’t believe the CDC study. Every other source says that women have much higher rates (4x) of attempts than men. Measuring attempts is difficult, but the CDC study also says that about 2% of both male and female high school students claimed to have made an attempt requiring medical attention in the past year. This is something that can be checked on the hospital side. Presumably such checking is what leads to the claim that women attempt more than men.

      In any event, the rate of completed suicides by high school males is 1 per 10k years and for females about a quarter of that.

    • Scott Alexander says:

      According to NIMH, about 8/100,000 teenagers (~0.01%) commit suicide every year. That suggests your school should have about one suicide every 25 years.

      If 8% attempt per year, that suggests 1/800 attempts are successful. I’ve heard 1/25 before as the actual statistic, so clearly something’s up. In fact, if we take CDC’s 2.5% in hospital number, that means only 1/250 attempts that land you in the hospital will kill you, which seems clearly wrong.

      EDIT: These guys say 4% lifetime attempted suicide risk among teens, which is significantly below the CDC figures. But it would still mean that…okay, let’s say there are ten years in “teenagerhood”. If 0.01% kill themselves each year, then 0.1% kill themselves total. That means the number of teens who complete suicide is about 1/40 the number who attempt it, which is different than saying 1/40 suicide attempts fail (since probably many people make multiple attempts) but still close enough to the 1/25 statistic that it’s not completely bizarre. I’m going to replace the 8% statistic above with the 4% statistic here.

    • Vicki says:

      Four percent of teenagers doesn’t (neessarily) mean 4 percent each year: if it means that 4 percent of all people (who don’t die of other causes) attempt suicide somewhere between their 13th and 20th birthdays, that’s about half a percent per year. Also, suicide isn’t just someone leaving a note and then (trying to) kill themselves: it includes people deliberately driving their cars into trees, which is almost never written down as suicide by coroners. It includes those attempted self-poisonings, and I would guess that most of the people who do that and live aren’t going to come back to school a week later and say “I can’t even kill myself properly,” they’re going to tell their friends they had the flu, or had to visit grandparents, or something.

  5. anon259 says:

    Less Wrong discussion. The following point in particular is worth noting:

    I passionately hate that all of the mental health people are obligated by law to commit me to an asylum if they think I’m about to kill myself. They can’t be objective. You know, if they could talk to me without such stupid constraints, they might have prevented this very suicide

  6. Gabriel says:

    “As she lay on the ground starting to feel sick, she thought about her problems a little more deeply, thought about how her family would feel, and decided she preferred to live.”

    Hm, I wonder about the market for guided suicide attempts — it’s just a bit further along the same line of reasoning that makes skydiving and boot camp good ideas.

    • Mary says:

      There was a M*A*S*H where one soldier was trying to kill himself.

      This was finally fixed when Col. Potter caught him trying to give himself an anesthesia overdose, shouted he was doing it all wrong, and went to force him to wear the mask properly. He freaked out and stopped trying.

  7. arrowpker20 says:

    So, your answer is basically that suicidal people are suicidal for a reason, and those reasons usually can, and probably will, change? So what?

    What if a man in the grip of despair backs up for a minute and looks at things from a viewpoint as objective as he can manage. What if he sees that his death wish was triggered by something trivial, and that it won’t matter to him tomorrow. What if he then notices that this does not actually effect the fact that he desparately wants not to exist? The fact that he will soon stop wanting to die will seem like a threat to his goals, and will probably intensify his desire to die then and now.

    So if a person is entirely lucid about his or her predicament and decides rationally that the viewpoint of his or her future self is of no concern, you take offense at the idea of that person commiting suicide? If so, then I’d suggest you reconsider your conclusion that your outlook are motivated by the values of the person in question. I think you intuitively find the idea of a human commiting suicide categorically sad and wasteful, and so want to minimize it. I probably don’t know your mind better than you do, but that’s the only way I can figure.

    • Ben L says:

      His point was that almost nobody does this – they almost all reconsider on their own.
      “The fact that he will soon stop wanting to die will seem like a threat to his goals, and will probably intensify his desire to die then and now.”
      E.g. almost everybody takes hope from the idea that they will soon stop wanting to die, not fear that their desires will be thwarted.

      • arrowpker20 says:

        They don’t all reconsider on their own, and even if they did it wouldn’t prove anything. It’s pretty condescending to all but say that people who say they want to die don’t actually want to die, or if they do, it doesn’t count. My moral intuitions tell me that you can’t possibly know that and the authority to decide should go to the individual.

    • Patrick says:

      You’re treating “desire to die” as an end-state goal in and of itself rather than a means to a separate end like “cessation of suffering.” Stop that.

      • arrowpker20 says:

        Are you saying that my analysis of dying as end-state is a flawed observation, or that I shouldn’t have a desire to die as a goal in itself?

        I know the experience, and it’s pretty clear that I sometimes feel like dying is something I want in and of itself, and a person in the midst of a suicidal episode can’t always choose what it is they want, otherwise he or she probably wouldn’t be having a suicidal episode. They kinda hurt.

    • Scott Alexander says:

      I guess I’m challenging the “lucid about predicament” assumption. I guess there could be two ways to do this:

      1. Consider the fact that things will be better tomorrow, but decide that current suffering is so unbearable that this doesn’t matter.

      I have a lot of uncertainty about what the ethics around (1) are, but I think we’re mostly dealing with (2), which I don’t describe as being lucid about the predicament.

      • arrowpker20 says:

        What if someone doesn’t wants to die as a reflex action, but is rational about the fact that they truly want to follow through on that reflex? The suffering in such a case comes from the dissonance between the fact that the person wants to die and the fact that he or she probably won’t until long after the reflex passes.

        I guess what I’m asking is, do you think a person can ever rationally pick the reflex to die? What if that reflex comes up nearly every day for the better part of a decade? You don’t have to say whether or not you’d support the person’s suicide, but do you at least think the person is nearing the border to acceptable?

        • MugaSofer says:

          Scott clearly stated that he wasn’t going to talk about the rare situations in which suicide is permissible in case it gets used against him.

    • Damien says:

      So I’m guessing that most people, if asked about it, would prefer not to kill themselves at a future point just because of temporary depression or loss of inhibition, and thus would sign up for not being allowed to do so. We’re not big these days on irrevocably committing one’s future self. If you can’t sell yourself into slavery, why should you be able to kill yourself, which is even more binding?

      A tiny minority of people who might fully rationally decide to off themselves in the middle of a physically healthy life seems to in fact be a tiny minority, so they can suck up the inconvenience of having to go through CBT (man that acronym is loaded) and other psych treatment.

      End of life conditions, with painful cancer or deteriorating brains, are another matter.

      • MugaSofer says:

        “If you can’t sell yourself into slavery, why should you be able to kill yourself, which is even more binding?”

        Ooh, nice. Very quotable. I may have to steal that.

      • Anonymous says:

        “If you can’t sell yourself into slavery, why should you be able to kill yourself, which is even more binding?”

        The difference is that slaves can suffer against their will, while dead people are just dead.

        Alternative solution: You can sell yourself into slavery, but even as a slave, you have the right to end your own life at any time.

  8. Anonymous says:

    You seem to make the implicit background assumption that the average life is REALLY worth living (why else would suicide be such a bad thing?)
    I doubt this.

    My impression is that most people are constantly struggling to keep their life above the worth-living- waterline.
    This link gives a good summary of my impressions:

    Of course, there are exceptions , but those are exceptions (maybe 10% of the population).

    If the best thing that a therapist can promise to his client is, that after years of therapy, the client will be just as miserable as the average, that is not a good motivation to keep on living. Enduring years of a life that is clearly not worth living, just for the faint hope of accomplishing a state of existence that is just barely worth living is not a very enjoyable prospect.

    • Eric Rall says:

      That may prove too much. As I read it, this would also support the claim that murder should be considered a much less serious crime, since by your argument ~90% of lives are either not worth living or only just barely worth living, and thus killing someone not in the remaining 10% isn’t depriving them of anything much worth having.

      Would you agree with this claim, or is there a distinction you’d make between the two claims that I’m missing?

      • arrowpker20 says:

        I think people oppose murder more because it offends them as because it’s a threat to society. It is a threat to society, but it’s more of an offense to average human moral intuitions.

      • naath says:

        The distinction is consent.

        If I believe my life not to be worth living and choose to end it that is me, making a decision about me.

        If you decide my life is not worth living and end it for me then that is different.

        Murder is not just depriving me of my life (which has value to me, and you don’t know how much) it is depriving me of the opportunity to make my own decision about my life.

        • MugaSofer says:

          So “consent” trumps utility calculations? Interesting.

          Tell me, do you actually identify as a deontologist?

        • Mary says:

          Well, obliviously. Since if the trump goes the other way, it also undermines the whole argument for suicide. If our utility is maximized by forcing people to stay alive, the only conceivable argument against it is that consent trumps utility.

    • Scott Alexander says:

      You’re right. Much of this argument assumes that people’s decision not to attempt suicide (or even want to attempt suicide but not do so because it’s too hard) implies that their life is positive-value from their perspective.

      I respect the opposite assumption but it’s not really what I’m arguing against here.

    • cool rich guy says:

      The link you posted cites statistics that show that people in first world countries are incredibly happy, then goes on to say that people in first world countries are actually miserable because

      1. Most people don’t really want to live that much longer than their current lifespan
      2. (a very negative picture of middle class life as a hellhole where everyone is trapped in a job they hate and blah blah blah)

      In response to number one, there are tons of reasons why people might not be interested in cryonics besides a secret hatred of life. (Belief in an afterlife, years of rationalizing death, attempts to be deeply wise, a vague distaste for the idea of cryonics that they can’t explain, etc.)

      In response to number two, it honestly would be just as easy to create a picture of modern life where it is portrayed as amazing. I don’t feel like writing out that entire picture right now in lurid emotional detail, and I’m not saying it would be any more accurate than the one in the article you linked to, but use your imagination a little – most middle class Americans experience exciting first kisses, wild nights of drinking and sex, hobbies that they are passionate about, unforgettable travel experiences, deep friendships, true love, the pride that comes from raising children, etc. And not everyone is stuck in Dilbertland where they despise their meaningless job – there are plenty of elementary school teachers who care about each and every one of their kids, tattoo artists who are passionate about their craft, therapists who change the lives of their clients, etc. Many people are even incredibly devoted to their “boring” corporate/office/business/financial/whatever job. And then think about how much material wealth middle class Americans have: we eat like kings breakfast, lunch, and dinner, we have basically an infinite amount of entertainment available on the internet at our disposal, we can go to other side of the world in a day if we want to, etc.

      I think the statistics are basically right, and that life is hard but pretty good. I have a 90% probability that either your social needs as a primate aren’t being fulfilled, or you’ve gone through some drastic unpleasant life experience in the past year or so.

      Also, this isn’t directed at you specifically, but I don’t understand the vocal subset of LW-sphere commentors that seem to think that life is inescapably awful, a far-right political ideology is basically accurate, and that almost any seemingly noble human behavior is a self-serving signaling mechanism. The amount of people on this post arguing for suicide is exasperating. What makes some people gravitate to this memeplex but not others? Is this a valid distinction that I am making? How can we determine if pessimism is accurate or not?

      • suntzuanime says:

        “Also, this isn’t directed at you specifically, but I don’t understand the vocal subset of LW-sphere commentors that seem to think that life is inescapably awful, a far-right political ideology is basically accurate, and that almost any seemingly noble human behavior is a self-serving signaling mechanism.”

        I think the last is much more common in the LW-sphere than the first two, I would almost say it’s conventional wisdom in this group. For instance, I’m basically okay with life, identify as a progressive if a heterodox one, but I completely accept the signaling theory of noble human behavior.

        I dunno where the suicide advocates and the reactionaries come from except that the LW-sphere selects for openness to weird ideas, but LW used to be Overcoming Bias which was partially Robin Hanson’s blog, and Robin Hanson writes a *lot* about signaling explanations for human behavior. So that’s a good guess as to why the LW-sphere is unusually likely to reach for signaling explanations.

        • cool rich guy says:

          I agree that signaling causes more of our actions than the average person would admit to, but there are people who I think take it too far and in a misanthropic direction. :\

          For example, I think that most people choose their political opinions for signaling reasons, but not, say, everyone who disagrees with me.

          An egregious example not of signaling specifically but of the larger meme of “people never do things for the reasons they say they do” is the people in the recent polygamy post on this site who argued that the Berkeley rationalists were polygamists because they’re romantically undesirable and undesirable people don’t want to commit to just one partner because that means they’ll have to spend their life caring passionately about an undesirable person. (long ass sentence)

      • arrowpker20 says:

        How can you determine if optimism is accurate or not?

        I’d imagine people gravitate towards right to die if they either want to die or deeply sympathize with those who do. People ordinarily oppose suicide because natural selection programs them to maximize their existence. I see no reason for that to correlate with actually enjoying life or wanting to experience it, and, given what I’ve seen of natural selection, it seems more likely to me that the brain would trick people into pretending that they don’t ever want to die when they actually do, on occasion.

        I’m not saying that pessimism is a more accurate model of the world than optimism, or vice versa (really, I think they’re both incorrect. Rationality can’t decide preference). I’m just saying that you have to admit to being extremely biased in favor of optimism if you’re human and don’t want to die.

      • Nick T says:

        Also, this isn’t directed at you specifically, but I don’t understand the vocal subset of LW-sphere commentors that seem to think that life is inescapably awful, a far-right political ideology is basically accurate, and that almost any seemingly noble human behavior is a self-serving signaling mechanism.

        I don’t think this is anything like a complete explanation, but all of those seem to me like natural consequences of frustrated idealism. (A lot of the kind of far-right politics you’re talking about comes down to “people are vicious and will kill each other and not be able to live in civilization if not ruled harshly.”)

      • Randy M says:

        Is this really a memeplex? I haven’t noticed much association with life being worthless and right politics, nor either with signaling being all-explaining.
        Although rightists are much more likely to believe that the awefulness of human nature is largely inescapable, I don’t think this tends to overshadow all the joy or meaning that can come from life as well.
        Could be projection, though. Are there studies on suicide and political ideologies? I’d wager it’s slightly correllated with more extreme positions in either direction, but also more with left than right, unless possibly controlling fo religion.

        • Mary says:

          Indeed, we will assure you that being rightist makes your life much happier!

          It’s that we get so many fewer unpleasant surprises, you see.

      • Fnord says:

        One of the effects of exposure to Less Wrong and/or an existing condition that attracts people to the LW-sphere is an erosion of the absurdity heuristic and similar mechanisms that keep beliefs in line with the median beliefs of society. Hence, you see greater adherence to all sorts of absurd (absurd, in this context, meaning vulnerable to the absurdity heuristic) beliefs.

  9. Kevin says:

    More evidence that suicide is often an impulsive action: Decrease in suicide rates after a change of policy reducing access to firearms in adolescents: a naturalistic epidemiological study.

    The use of firearms is a common means of suicide. We examined the effect of a policy change in the Israeli Defense Forces reducing adolescents’ access to firearms on rates of suicide. Following the policy change, suicide rates decreased significantly by 40%. Most of this decrease was due to decrease in suicide using firearms over the weekend. There were no significant changes in rates of suicide during weekdays. Decreasing access to firearms significantly decreases rates of suicide among adolescents. The results of this study illustrate the ability of a relatively simple change in policy to have a major impact on suicide rates.

    • Douglas Knight says:

      It decreased from 28/year over 3 years to 16/year over 2 years. How much variance is there in this time series? Why don’t they at least give us the 5 numbers? Pure datamining. the paper

    • Scott Alexander says:

      I’m aware of that research (there’s another good one about how suicide rates in Britain decreased by a third after carbon monoxide stoves were banned, apparently because the third of suicides caused by carbon monoxide stoves just disappeared without being replaced by anything else).

      I didn’t include it because it seems equally consistent with my theory that suicide is impulsive and irrational, and Sister Y’s theory that suicide is premeditated and rational but most people aren’t doing it because they don’t have trustworthy painless methods.

  10. Jack says:

    “the rotting-intestines woman became much better after a short course of electroconvulsive therapy, but some people will get better just on medications”

    Wait, does electroconvulsive therapy actually work sometimes? I didn’t know that.

    • Mary says:

      Oh, yes, it works. Kitty Dukakis talked about the wonders it had done for her to encourage more people to use it.

    • Misha says:

      it actually has a bunch of well-known applications and isn’t just forced torture instituted on patients for the fun of sadistic psychiatrists like it’s portrayed

    • Scott Alexander says:

      Electroconvulsive therapy is extremely effective for its proper uses (usually severe depression).

      It definitely causes temporary amnesia, and there’s a debate going on over whether it can cause longer-term memory loss and cognitive disability (I think the consensus is that it doesn’t, but I’ve heard very strong lingering doubt from some corners and don’t know much about it).

      Barring that turning out to be a real and serious side effect, I think it’s right up there with nuclear power in the category of “extremely effective things that are just really really hard to sell”

  11. I’m always slightly puzzled by the voices thing. If I picture a voice telling me to kill the President/myself/whatever, it doesn’t seem very persuasive.

    Reports of people with actual voices suggest they are, in fact, persuasive.

    Clearly something non-intuitive is going on, but what?

    • MugaSofer says:

      Maybe you’re not trying hard enough to come up with convincing arguments?

    • Mary says:

      For how long do you imagine it? Imagine morning noon and night whenever you are awake, perhaps.

    • Randy M says:

      It’s not unlikely that along with the voices are less consciously felt urges in the same direction; that is, their minds interpret irrational urges to act as actual voices telling them to do so. I have no idea if this is so or not, but it seems like the kind of thing a misfiring rationalization circuit might feel like.

    • Julia says:

      Brainwashing works, at least sometimes. Hear a message long enough (especially if it’s internally produced) and it sounds more and more viable.

  12. Fnord says:

    Of course, beyond the issue of “to what extent should we prevent suicide”, there’s also the issue of “how do we go about preventing suicide”. Because to the extent that the specter of involuntary hospitalization interferes with people’s ability to get effective mental healthcare, that’s a problem, too.

    If people aren’t honest with their psychiatrist about suicidal thoughts because they fear involuntary hospitalization, that’s impairing the ability of the psychiatrist to provide effective treatment. If a test like you discussed in the last post becomes widespread, to the extent that it discourages people from seeking mental healthcare at all out of fear of being hospitalized based on a test they have no conscious control over, obviously those people are not going to get effective treatment.

    So it seems entirely consistent to agree with everything you said in this post, and agree that psychiatric treatment up to and including hospitalization may be something we should provide for suicidal patients, but still think that involuntary hospitalization for suicidality should be curtailed or eliminated.

  13. Michael, it might make sense to look at a conspicuous problem first.

    People (and there’s quite a few of them, though not nearly enough) who realize that focusing on the news is making their lives worse rather than better, and then spend less time on the news, is some evidence of the rationality level going up. I don’t know if it’s just people getting older and more experienced (and less excitable?), or if the idea of resisting hyper-excited news is moving down the age cohort.

  14. Grognor says:

    My comment was originally going to be two words long. Three guesses which two words they were going to be.
    Your whole post operates on the premise that suicide is bad because the person will enjoy life in the future assuming ¬suicide. But this future person is a fictional character. I don’t care about this person. I am not a fictional character. I am here, in real life, and I am suffering. Without this premise, your whole post falls apart. Your post also assumes that, assuming no intense suffering, life is inherently worth living. I disagree with this assumption, but as everyone in the entire world takes it as an inviolable axiom, I won’t even try to shake you of it. So right away your post had been doomed. It is a common position for disillusioned transhumanists to appreciate that life is not inherently worth living, but might be some day, and for that reason they choose not to die. But that is their choice to make, not yours. Making it for them is violating their autonomy.
    “And I think most people agree that teenage suicide is terrible and requires treatment.” Why do you care what “most people” think? “Most of their problems […] are eminently wait-out-able.” Again, it is not good to make people live for the benefit of their future selves, who are different people. You might as well be saying that if a person is suffering, instead of ending the suffering we should put them through enough Hells that they come out of it a different person who is not suffering anymore. “There are certain depths of despair dark enough that the knowledge that the despair is completely temporary cannot penetrate them.” No. I’ve been there, amigo. I knew just fine that it was temporary. And I didn’t care. And I didn’t care that I used to care and probably would in the future. You know what I did care about? That people like you made it impossible for me to get rid of the pain. “It is this state, defined by the clouding of rationality by suffering, that I think most teenage suicides occur in.” Bullshit. That’s not your true rejection at all. Suppose a happy person decided to commit suicide. “I love life and all, but I’ve seen enough. I’m gonna end it tonight. See you in Hell, boys.” You wouldn’t say his judgement was clouded by suffering. You’d say his judgement was clouded by something, anything else. Despite your sleight-of-hand at the beginning, you treat suicide as a plan that is not allowed to be rational. Not everyone has the time or clarity of cognition that Sister Y does. You shouldn’t have to spend a couple decades to prove your sanity in order to be allowed to end the pain. But you know, even if everyone were really capable of doing that, I have a funny feeling that, if it were up to you, the bar for being allowed to commit suicide would be raised arbitrarily high so that only a tiny minority of people are allowed to go through with it.
    Having a mental disorder doesn’t suffice to prevent people from making decisions in any other domain. It’s only when people want to die that all of a sudden their sanity must be established. But having children or driving or any equally foolish decision has no barriers whatever.
    I guess experiencing emotions more intensely than people normally do makes it more obvious that, being an evolved creature, bad emotions are badder than good emotions are good. Sister Y has written about this extensively, but really, there’s no need to read anybody on this. You can just observe it for yourself if you introspect honestly.
    “But the average depressive suicide is not a Perfect Philosopher who has calculated, while healthy, that the possibility of another six month depressive episode is too much to bear.” AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
    “Somewhere so dark that one of the main side effects of effective antidepressant drugs is suicide” how are you not taking the modus ponens on this!?
    “Do you know what the numbers 4%, 10%, 13%, and 19% all have in common? Yes. They are all significantly less than 50%.” You jerk. We are evolved creatures. Every human comes from a tremendous, unbroken chain of millions of creatures that lived long enough to have descendents. It takes an incredible amount of pain to overcome that kind of force, not to mention the incredible societal force stigmatizing and shaming suicidality. Consider again Sister Y’s thought experiment of the golem that suffers horribly every day but continues to profess to live. It would not be good to create this creature. Consider another creature, a creature that suffers every day and professes to want to not suffer but is prevented from doing so by people such as yourself making clever arguments. This is unambiguously the lot of actual humans, in real life. (Also, why should anything magically change when a proportion goes above or below .5? This number is only the reasonable-sounding Schelling point for making political changes seem legitimate.)
    As regards the people who regret attempting suicide after having made the attempt, I’ll just point you toward thinking about game theory and hope that you’ll figure out why this happens on your own. You’ll be more likely to realize the implications that way, than if I just tell you.
    “Psychiatric Care Probably Helps” My response to this section is simple. I was a victim of forced hospitalization, not for attempting suicide, but just for saying I was going to attempt suicide. To people who had promised not to repeat it to anyone (they did, obviously). I was out of work for two days. If it had been three days, I would have lost my job right there. (Wal-Mart doesn’t fuck around.) The staff at the hospital was ordinary, nice people who were only a bit dim. The patients were creepy and I couldn’t sleep because the one who was assigned as my roommate snored. I spent three days incredibly bored. So, not the worst thing that ever happened to me, but as a result I had missed two days of work (three days of my life total), been charged $1800 that I obviously couldn’t pay, lost my ability to ever join the military (I know because I tried), and I’ll never be able to get a loan as long as I live. Oh, I also had to take some pills. None of this is surprising to you. The point is that it made my life worse, not better. I came out of the experience more suicidal than before. And when I was released, it turned out that it ruined my ability to talk to my favorite person at the time! Which I have only gotten over in the past month.
    It makes me wonder at the culture-as-a-whole’s incredible desire to reiterate and reinforce the cheery anti-suicide-choice position. The only explanation I can come up with is that people who take the opposite position select themselves out of the population. That, and stuff that you said in your “Why I Defend Scoundrels” post.
    For obvious reasons, I will not be responding to, or indeed even, reading responses to this comment or any other comments in this thread. For less obvious reasons, I won’t be proofreading this comment. Apologies for fixable mistakes or any redundancies with other people’s comments.

    • Federico says:

      I was a victim of forced hospitalization, not for attempting suicide, but just for saying I was going to attempt suicide.

      By telling someone, you obviously decreased the probability of a successful suicide. This suggests that part of your brain did not really want to commit suicide, and foiled the suicidal sub-agent(s) by implicitly soliciting help in this way.

      People might also consider your friends and family in their ethical calculation, since (any temporal snapshot of) you would not be the only entity harmed by your suicide. Suicide ruins the lives of people close to the victim, including anyone whom you told about your plans.

      Also, my experience is that humans introspect poorly about their preferences and even their beliefs when clouded by emotion.

      • naath says:

        Sure, if I kill myself some people will be very upset. Some of those people are people I hate, and wish to suffer. Maybe their suffering is one of the objectives of killing myself?

        I guess there are people who hate their lives and yet still feel that there are people worth living for; but personally whenever I find myself hating my life the people in it are one of the worst features.

        Also it should be remembered that people-in-general will benefit from an individual’s death – so long as they aren’t close enough to be upset by that death; when I die there are charities that will get my money, someone else will get my job, I won’t be taking up space, eating food, drinking water… the population of humans is large enough that one fewer means more resources for the remaining.

      • anon1 says:

        Discussing suicide with a mental health professional is pointless because you know what they are required to think. But if my friend sees reasons for me to live, or points out something I should try first, that actually means something. I am very lucky to have at least two friends whom I know I could trust to discuss this sanely and honestly, with no risk of reporting me to the authorities.

        (The result of these discussions was effectively a self-imposed waiting period. After two months of consistently wanting to die and having no low-hanging fruit left like “drop out of grad school,” I would seek out personality-altering treatments. (I have never *not* been depressed so I can’t help seeing eg antidepressants as a form of suicide, albeit one that wouldn’t affect others in the same way. I want a future-me who goes on being gloomily amused by life, not someone so happy I can’t identify with them at all.) After some number of failed treatment attempts and repeats of the whole cycle, I would consider physical suicide. I have never yet run through the initial two-month waiting period, nor do I expect to.)

        Suicide is a serious decision and it’s very important to be able to get an outside perspective on it. Talking to a friend shouldn’t carry the same risk as talking to a psychiatrist, and shouldn’t be treated as the same kind of “call for help.”

        • Scott Alexander says:

          If you expect antidepressants to make you annoyingly cheerful, you are woefully overestimating the power of modern psychopharmacology. Remember, these are the things that keep spawning rumors that they’re actually placebo.

          I’ve been on and off antidepressants a few times, and I mostly just noticed some very subtle blunting of emotions, both positive and negative. I had the same concerns about it being a form of changing into a different person, but they seem kind of silly now. It’s like saying you’re a different person in sunlight and indoors, since sunlight has a mild (or sometimes not so mild) positive effect on mood.

          • anon1 says:

            Late update: began taking bupropion a few months ago (having gotten only 2/3 of the way through my waiting period because it turns out I’m embarrassingly weak-minded) and by my previous standards yes I am annoyingly cheerful. I’m happy as a default state much of the time, even when there is no reason for it. And the vague sense of doom at the back of my mind is consistently just gone. This effect is not at all the kind of thing I’d have gotten from spending more time in the sun and it is Not Normal. (Unless I *really believed* that spending more time in the sun would do this? I suppose I was primed for a massive placebo response. Not that there’s any point in trying to sort out “real” v. placebo effects on an individual basis.)

            A lot of my past thoughts and actions look like deluded nonsense now, and it’s hard to sympathize with that person. Whether that reaches the threshold of being a different person is hand-wavy and doesn’t matter now, since past-me didn’t actually want to exist, even in a hypothetical less stressful future where existence would be tolerable. So either way I’d have no obligation to ever bring back that version of myself.

        • anon1 says:

          Not “annoyingly cheerful,” just different enough that I’d be unable to identify with that particular future-me. I’ve read anecdotal reports of effects strong enough to scare me, even if this isn’t typical.

          I’ve also experienced major short-term (four days) personality change following a moderate dose of psilocybin, which suggests to me that “annoyingly cheerful” may be a near-stable state that my brain can be kicked into by outside forces.

      • Idealect says:

        Or maybe he just trusted his friend not to betray him.

    • Federico says:

      I would only respect suicidal “autonomy” when the person suffers an incurable and intensely painful or debilitating disease, i.e. merciful euthanasia.

      Take it from me, you do not experience emotions more powerfully than anyone else; suicidal depression is just an exceptionally intense feeling. It doesn’t last forever.

      • Scott Alexander says:

        I think I do know people who experience emotions more powerfully than other people. Why do you think Grognor isn’t one of them?

        • Federico says:

          Have you ever had a severe headache, and felt like “This *pain*, it is hell, no-one can understood what I’m going through! I want to decapitate myself!”

          And then when the head-ache is over, it’s more like “Ahh…I really don’t like head-aches”.

          Depression is like that. The word doesn’t evoke the feeling very feel—it’s basically “vomiting emotional pain”, with quiet sickness interludes. Not everyone experiences it, but I think almost anyone is capable of doing so, and as far as I’ve seen the intensity of depression is always extreme, seemingly exceptional and unbearable.

          Once the depression clears up, in hindsight those snapshots of personhood are pitiable and, like one in the throes of a migraine, a biased and selfish minority view within the temporal parliament of sub-agents.

          I wouldn’t want them to affect the rest of a person’s life, let alone his family, friends and colleagues.

        • Scott Alexander says:

          Sorry, are you making this claim about clinical depression? Are you saying that everybody goes through periods that are just as bad as a person with severe clinical depression does?

          Are you aware that there are levels of clinical depression where a person literally doesn’t move for days, starving and soiling themselves, because they believe any action they take would be hopeless? Isn’t that pretty far beyond what the average person deals with?

          I mean, everyone gets bad headaches sometimes, but only some people get the particular pleasure of dealing with cluster headaches, and it would be a big mistake to think your pain even comes close to approximating theirs.

        • Federico says:

          I don’t doubt that some people are more often and severely depressed, or have worse symptoms than others.

          I question the idea that any of them have, cet par, more intense qualia than others. Rather, I expect they are less tolerant of pain, more highly strung, have worse coping mechanisms or social support, etc., than others.

          I base this on my own, similar failures of introspection when depressed, and the way other depressed people have spoken and behaved. Depression, like a migraine, feels more unique and incredible than it is—even given the existence of genuinely exceptional cases.

          This doesn’t mean I don’t sympathise with depressed and suicidal people—far from it. I think the most effective cure is having affectionate and close (probably female) friends to talk to.

        • ozymandias42 says:

          I don’t know how you can tell apart “more intense qualia” from “less tolerant of pain.” I mean, I can’t experience other people’s feelings from their point of view; all I have to go by is their reactions. I react to stimuli much more strongly than other people do; therefore, I assume that my feelings are more intense than theirs.

    • Alex says:

      “Your whole post operates on the premise that suicide is bad because the person will enjoy life in the future assuming ¬suicide. But this future person is a fictional character. I don’t care about this person. I am not a fictional character. I am here, in real life, and I am suffering”

      This proves too much – it proves that there is no such thing as an irrational action no matter what the consequences are, because the future person who has to deal with the consequences isn’t you.

      It also proves that there are no rational actions either. Collecting lottery money isn’t rational and it won’t benefit you, because the future person who will enjoy the wealth isn’t you at all.

  15. sixes_and_sevens says:

    I’d just like to say I really enjoy your posts that deal with mental illness. Please make more of them.

    (Posts, that is, not mental illnesses).

  16. Pingback: In Defense of the Suicidal | Boredom Breeds Contempt

  17. Philo says:

    I really appreciate rational discussions about important and controversial topics, so thanks for this.

    My sympathies lie more with Grognor’s position for personal reasons and also due to the point made in the post about “future selves” being hypothetical entities. But reading both sides, and perhaps engaging in a little more debate about it, is helping me clarify my views.

  18. Teresa says:

    This is a great column– I speak as a senior citizen, teen-age survivor of psychotic depression (the description/explanaiton here is right on target!!) and suicidal, ECT treatment. Also, the intermittent episodes. However, I always see/feel it comming on and seek help because I know that I do not want to die, just because AT THE MOMENT, I do not want to live.

  19. I am very reluctant to get into in what situations I believe suicide is acceptable. I am scared that one day my future employers will read this post.

    I thought the main point of moving to this site from your old blog a few months ago was so you could talk about situations like these anonymously. As I remember it, the last post on your old blog (now deleted), which talked about the move, even named discussion of controversial work-related opinions as a reason to switch. Am I remembering wrong, or did you change your estimation of how safe blogging here is?

  20. Sister Y says:

    A few questions we should be asking:

    1. Since the time-delay comfortable suicide clinic option (with organ donation, which btw would 100% solve our organ shortage) is so obviously awesome and yet doesn’t exist, what does that say about the nature of our current suicide prohibition?

    2. In a slightly different context – countries where abortion is illegal – how would we interpret the choice of someone who got “rescued” from trying to “commit” abortion to later not do it? How fair and realistic is it to analyze the later actions of would-be suicides in the context of a suicide prohibition?

    • Sister Y says:

      One other thing: the prohibition itself tempts people (like probably some of those high school kids) who don’t really want to die to take deadly action in hopes of being “rescued.” If we had a non-prohibition culture, people attempting suicide would know they wouldn’t be rescued; would so many people hurt themselves? Also, if the time-delay clinic option existed, their dangerous-but-insincere attempts would have negligible signaling value; would they be as tempted to do it?

      • Scott Alexander says:

        I actually take this from a different angle than you. Yes, a lot of suicide is a “cry for help” not intended to cause actual harm, and yes, that is a function of our social prohibition on suicide. But a lot of the time this works pretty well.

        It’s a pretty good way to force a school to pay attention to bullying, it’s a pretty good way to force parents to man up and accept that their kid has a psychiatric disease, and it’s a pretty good way to jump the queue to get psychiatric treatment (which, let’s face it, if you’re upset enough to make a convincing suicide attempt you probably deserve).

        See also this study on income increasing after suicide attempts.

  21. Sister Y says:

    “the voices in people’s heads do tell them to kill themselves a lot”

    You are speaking of command hallucinations, an important symptom of the most severe thought disorders such as schizophrenia. There are really two separate question: “is this person making a deliberate decision to die out of a coherent desire to do so?” and “is this person’s life in the future going to be so worth living as to justify the pain, humiliation, and demoralization of forced hospitalization?”

    You seem to be focusing on the first question – that if a command hallucination tells you to do something, it’s hard to say it’s truly your will. However, focus instead on the implications of the action of forcible hospitalization and whether that’s justified: this requires extreme “livability” of the future life, I think.

    Depression carries relatively little increased risk of suicide (as you note, a huge percentage of people technically have a mental disorder, mostly depression and anxiety, because the DSM/APA rejection of etiology and scientific techniques in diagnosis allows vague definitions of disorders). However, a few disorders carry a genuinely high risk of suicide – severe thought disorders, bipolar disorder, and borderline personality disorder off the top of my head.

    I think it is very hard to say that a person with severe command hallucinations has a life very likely to be worth living enough to justify forced hospitalization. What does this kind of thought disorder most interfere with? It is primarily the ability to form social bonds and belong in a social community, as well as to be productive to others and not a burden on them, which are two of the three primary determinants of suicide noted by Thomas Joiner in his book “Why People Die By Suicide” (social belonging & burdensomeness). People with command hallucinations or borderlines (real ones) are extremely hard to be around, seem to sabotage social bonds, and act out unpredictably. There is obviously something wrong with them, and we are fooling ourselves if we think their fellow humans will not react badly to this. They are unlikely to be able to achieve the most basic human need (more basic than food): the need for social belonging.

    There are meds, but they are generally so awful that people frequently voluntarily stop taking them. Massive weight gain, tardive dyskinesia, or worse – many rational people would choose illness over these side effects. The side effects may, in fact, do almost as much as the underlying disorder in making social belonging impossible.

    Life is not always worth living, and while many people with severe mental illnesses might not be fully rational when they try to commit suicide, it is hard to say that there is a clear benefit in putting them through forced hospitalization when the only prize at the end is they “get” to live out the rest of their lives.

    • Scott Alexander says:

      So this is going to be ridiculous and hypocritical of me, but let me try it anyway and accuse you of paternalism here.

      You seem to admit at the end what was going to be my main objection to your comment: that the fact that people obey the command hallucinations and commit suicide is not causally linked to the fact that their life isn’t going to be worth living, even if that’s true.

      That means their decision to die isn’t made rationally, and to allow them to take it because you (who are rational) think it might be correct is in fact kind of paternalistic. I would rather they be treated, and then if in fact they find it’s true that the drugs have too many side effects and they can’t fit in anywhere, they can choose whether to live or die on a much more justifiable basis (and in my ideal world, which unfortunately is not the world I am enforcing the rules in, that decision would be respected).

      Less importantly, I think your picture of life with schizophrenia is overly pessimistic (you? pessimistic? I know, weird, right?) Tardive dyskinesia is now pretty rare (I’ve seen it mostly in elderly people who were treated with older-generation antipsychotics when they were younger). Metabolic effects are also not uniformly awful: olanzapine, the most infamous antipsychotic for weight gain, causes an average increase of about 15 pounds, and more typical numbers for other drugs in the class are <10 pounds, hardly enough to make life no longer worth living. A common practice among good psychiatrists (whom I admit are too rare) is to ask the patient their needs and preferences, and if someone is unusually opposed to weight gain their psychiatrist can usually give them a drug that does not have that side effect, by trading off either other side effects or time/energy finding the exact right drug and dose.

      Regarding social functioning numbers, it looks like about 15% of schizophrenic people in the community are married at any given time (to which we add people who married then divorced, and people who are young but will be married later) and just from anecdotal experience many more have relationships of various long-term-ness and stability. I also am (somewhat) hopeful these numbers will continue to increase as medical tech advances.

      So yeah, I agree it's not a great prognosis, but it's also not such a bad prognosis that we should assume the voices telling someone to kill themselves are correct.

    • Alex says:

      “You seem to be focusing on the first question – that if a command hallucination tells you to do something, it’s hard to say it’s truly your will. However, focus instead on the implications of the action of forcible hospitalization and whether that’s justified: this requires extreme “livability” of the future life, I think.”

      I thought that was the entire point of the section which showed that people who attempted suicide were overwhelmingly glad to be alive later on. If you think that their lives aren’t really worth living despite what they think, I’d be curious to know why.

  22. Pingback: Utility Weight Results | Slate Star Codex

  23. Pingback: Raikoth: Laws, Language, and Society « Random Ramblings of Rude Reality

  24. Context: My husband killed himself about a year and a half ago.

    He had been entertaining thoughts of death for a long time. Thanks to a combination of physical problems that we’d mostly gotten under control (but were still slowly deteriorating), physical problems that we were still trying to figure out, and drug-resistant major depression, I believe he was anticipating a life of constant pain with very little pleasure (at least, he said as much). He was also a PhD student working on his thesis, acutely aware of how badly his condition was affecting his research output, but determined to finish.

    So I believe there was a strong rational component to his decision (though I also think the stress of writing up was a less-rational component). Given certain aspects of the circumstances, I am also pretty certain that he had been making plans for a while (including setting the dominoes up so that it would be impossible for me to be the one to find him — I was overseas at the time). However, I am also pretty certain that there was a strong impulsive component to the precise timing of his suicide.

    Maybe this is just a long-winded way of saying “people who are under sufficient stress lack full decision-making capacity, here’s an example,” in which case it’s a crying shame that a horrifying teenage experience in a corrupt state psychiatric system led him to refuse any psychiatric intervention other than drugs thereafter. But we already knew that.

  25. Pingback: [Turing 2013] Atheist Entry #2

  26. Anonymous says:

    Jesus Christ. What an arrogant, insulting mess.

    You don’t own other people. Get the fuck over it.

  27. Idealect says:

    Where’s the proof that the people who kill themselves impulsively are wrong? It’s unprovable either way, but you’re taking it as a given that they’re wrong, and clearly so.

    When you’re presented with unacceptable conditions, backed up by force, often the choice is to fight those conditions, or become something to which those conditions are acceptable.

    System 2 can be as corrupt as system 1.


    A suicide isn’t the tragedy, it’s the tip of an iceberg of tragedy. When you say the suicide is the tragedy all you’re doing is formally withdrawing your allegiance and sympathy from the person that killed themselves: It doesn’t matter that they were suffering, it doesn’t matter that they wanted to die, it doesn’t matter the position they were put in, or the conditioning and traps and bullshit that society used to disrupt their mind, what matters is they didn’t go along to get along, didn’t capitulate, didn’t die slowly and quietly: breached the peace.

    I don’t often hear the feeling that they made a mistake. It’s how dare they, they didn’t have the right, or that they were just an idiot who refused to see how easy life is. And to be honest I get the impression most of the people saying it was a tragic mistake are posturing as wise, while the others are being more honest (having at least deluded themselves into these views, via just-worldism, before regurgitating them faithfully.)


    Revealed preferences are such a sick thing to refer to in suicidal people. There’s no such thing as revealed preferences. If a woman orgasms during a rape it doesn’t mean she wanted to get raped. A lot of people can’t help adapting to make horrific circumstances easier to bear. In fact, it’s a popular life lesson, uttered with grins rather than horror: “if you can’t beat them, join them.” Not to mention the sunk cost fallacy, and the temptation to have a way to say you knew it all along.

    The idea that suicidal or depressed people “know” that things will magically get better in the future, but just irrationally don’t integrate this knowledge into their thinking is equally sick.


    Conditioning can be broken on an impulse. When your mind is tied in knots most directions are rabbit holes to nothing, but if your neurons and the wind align they can overturn things out of nowhere. But it can be just a chance at a path, a fleeting tunnel. Inspiration, insight, spite, “entitlement”, whatever it is, a flash of it might be followed somewhere better.

    One of those insights is: if it doesn’t work out I can kill myself anyway. One of those adjustments is increased risk tolerance. One of them is caring less what other people think.

    Paternalistic “you don’t get to kill yourself and anyone who does was irrational” nonsense is a fairly neat piece of insiduous poison that almost seems designed to prevent that kind of loosening up.

    When you’re trying to fix a broken mind, you don’t have to hold anything in reserve to be able to gracefully handle it if it doesn’t work*. Your life and death are yours. You are a sovereign person. A lord of your own tiny skull sized kingdom, even if it’s been conditioned and tricked and trapped its whole existence into a state of horrific disrepair. It’s yours. You’re yours.

    *nor, of course, is the opposite an obligation.

    That’s what the get help crowd don’t want you to think. They want you to give up your judgement, and let the men in white coats guide you into the rigid embrace of the hivemind, which solves all things.


    (Good luck to you if you choose guidance, a little bit of standardisation, and what might be cynically called placebo, as if to better yourself through an anchor of thought is misguided. I’m hypercritical of it because it’s something whose elevation to dogma is harmful to me and people like me. But if its right for you, it’s right for you, and you’d be a fool to see it the way I do. I don’t live in your mind, and I haven’t been accounting for it, or most minds, here.)


    But aiming higher and ending in the gutter doesn’t mean you were wrong. Sometimes it’s right to go all in on 70%, sometimes 55%, sometimes 50%, sometimes less. To decide I’ll be strong and good and smart and me, or whatever it is you don’t want to compromise.

    Or I’ll die trying.

    And die trying.

    It’s usually correct to overshoot your commitments. They won’t go unharrased, unattacked. Most of the time you won’t have to die trying, won’t even come close. You can bide your time, come back from another angle, try again, and again, and again. Or come across a smooth, decent compromise. Or a lot of things. But if you decide in advance that total defeat from your own approach is an unnaceptable risk (maybe I should buy IBM?), and let people convince you your life is not yours, it won’t help you to keep your aim or your head high.