Monday, September 13, 2010

A Duty to Rape?

Robin Hanson gamely bites the bullet I offer up in my Rape Doctor Hypothetical, a thought experiment I devised to test intuitions about risks of inflicting harm and benefit on others, in cases where a proxy for consent must be used because actual consent is impossible.

Consent is nice, when you can get it. When consent is impossible, as it often is - when providing medical care for unconscious patients, for example, or when parents make decisions for their children (at least preverbal children), or when we bring a new being into life - we must decide whether to use a proxy for consent. These might include:
  • Ex-post ratification (examined in my piece The Moral Effect of "Being Glad It Happened")

  • Predictions based on the ex-post ratifications of similarly-situated others (as I think Robin Hanson would have us use in the procreation case)

  • Predictions based on a mental model of the nonconsenting being, including perhaps its likely utility function and the costs and benefits of the action.

All of these, of course, involve probabilities; they are unlikely to be perfect, and are in fact virtually guaranteed to result in some margin of error. How good should we require the predictions to be before using them? How much risk is too much for the nonconsenting beings we are acting on behalf of?

Many accepted proxies for consent are used to avoid harm (e.g., treating an unconscious patient to save his life - since most people wish to remain alive). But what about using proxies for consent to provide a pure benefit - with some risk of harm?

Please read my whole hypothetical for details, but in short, I posit a situation in which a doctor has identified a class of patients with Forced Sexual Contact Arousal Syndrome, who are only capable of sexual arousal through rape and will be benefited, not harmed, by being raped:

Based on his research, Dr. A has statistical grounds to believe that, of FSAD patients who meet Criteria A, B, C, and D, 99.9% will experience sexual enjoyment exclusively from forced sexual contact. Beyond that, Dr. A notices that his FSCAS patients who have been raped are much more socially and emotionally well-adjusted than those who have not. It is statistically reasonable for him to believe that, out of 1000 patients with FSCAS who have not been raped, 999 will experience a great deal of sexual enjoyment and a much better quality of life if raped; one will experience the usual extreme distress that rape would cause a normal woman.

So should Dr. A rape his patients? Robin Hanson says: "I'll bite the bullet and say that the rape has expected good consequences in this case." I take this to mean that the special rape under these circumstances is at least permissible, and perhaps that Dr. A even has a duty to rape his FSCAS patients.

Intuitions are the stuff of ethics. Here, Robin Hanson is taking (I think) a position I describe in my article as an extreme form of consequentialism - the idea that the suffering of a few is offset by the pleasure of others. It is the move from humane Pareto efficiency to ugly, realist Kaldor-Hicks efficiency - that the suffering of a few is a fair price for the benefit of the many, even if that suffering is not consented to.

Hanson and I disagree as to whether a 99.9% chance of pleasure and life benefit is worth a 0.1% chance of the ordinary harm of rape. A more general phrasing of the question is this:

The Dilemma of Impossible Consent: In cases where consent is impossible and a proxy for consent must be used, how risk-averse should we be on behalf of those our decisions will affect?

My answer to this, supported by my own intuition and what I see as commonly-held intuition across a variety of situations, is: extremely risk-averse. In addition to the thought experiment above, I examine this notion in my post on dosing strangers with ecstasy. Seana Shiffrin examines this position in her paper "Wrongful Life, Procreative Responsibility, and the Significance of Harm" (Legal Theory 5:117-148, 1999), which I summarize here. It is a notion that is usually uncontroversial - except when it is brought to someone's attention that antinatalism is among its ethical conclusions.

How risk-averse should we be when potentially dealing out unconsented harm to others? I think the position Robin Hanson is articulating is: not that risk-averse. How risk-averse, then? As I mention in the comments, how far would we have to skew the probability in the Rape Doctor Hypothetical to make the rape impermissible (or, if there is a duty to rape under my facts, to make it permissible to refuse)?

There is a related question which I think is separate from the first, and that is:

The Dilemma of Uncompensated Suffering: To what extent may a few be made to suffer greatly, without their consent, so that many people will be benefited?

This is a separate question from the first, although both are appropriate perspectives to consider in the case of creating or refusing to create a person (and raping or refusing to rape a likely rape-beneficiary). The first question inquires how we should treat risk in a decision affecting a non-consenting other; the second inquires how we should balance and compare interpersonal utility functions.

I am interested in (but have not encountered) a strong defense of the position that some may (or must) be sacrificed for the benefit of many. John Leslie carefully considers the issues in his book The End of the World: the Science and Ethics of Human Extinction (he's anti-extinction, by the way), but acknowledges that he fails to provide anything like a proof of the position. (Note that this was written before Benatar's Better Never to have Been was published, and Leslie does not consider Benatar's arguments.)

Again, ethics must be based on intuitions. The most interesting ethics happens when intuitions conflict. My intuition is that it is never permissible to seriously harm one in order to provide a pure benefit to many; Robin Hanson's intuition (and that of many others) is that this is fine, under some circumstances. My intuition is that we must be very risk-averse on behalf of others if we may harm them seriously without their consent; Robin Hanson's intuition (and that of many others) is that we can be utility-maximizing without any special regard for risk-aversion. In other words, there are real ethical disagreements regarding the basic intuitions underlying the ethics of reproduction.

In addition to my two dilemmas, I pose a third:

Dilemma of Ethical Uncertainty: Given ethical disagreement between epistemic peers, what is the proper course of action in the real world regarding reproduction?

See also Chip Smith's One Man's Exquisite Treasure.

Correction: I incorrectly refer to risk aversion (preference for certainty) throughout this piece when I mean loss aversion (desire to avoid harm is greater than desire to realize gain of the same magnitude). I leave the text as is since comments were made before I noticed my error. In other news, I have a hard time telling left from right and I tend to pronounce "scourge" phonetically.


  1. I don't really think "risk aversion" captures the key issue well. The key issues are closer to agency costs. When would we want empower others to force benefits on us? We would have to have high confidence that they would use this power with little corruption and with excellent judgement. The more we expect them to use such powers with bias and ignorance, the more reluctant we would be to accept their role. When we each consider whether to force our help onto others, we should take seriously such doubts about our own bias and ignorance.

  2. Perhaps not what you have in mind, but in his recent critical review of BNTHB, Joseph Packer extrapolates from David Benatar's central argument to show how the asymmetry could be construed to justify vast human sacrifice on utilitarian grounds. Packer is by no means posing a "strong defense" of this position (quite the contrary), but someone's bound to run with the scissors sooner or later.

    I think extreme risk aversion is the sound stance with reference to reproduction (especially given ethical disagreement), but I think Hanson's focus on agency costs makes intuitive sense for those who are already alive. If I were hospitalized in a delusional state, I would prefer to have Gregory House calling the shots on my behalf.

    Since I don't believe it can be demonstrated that anyone is meaningfully "benefited" by being brought into existence, "ex-post ratifications of similarly-situated others" fail to justify procreative force in the face of profound risk and uncertainty. At the same time, it seems better than a crapshoot. It seems better that Bryan Caplan should clone himself - given his unique epistemic vantage - than play the genetic lottery.

  3. Certainly there are agency cost issues here (parent-offspring competition, for example, is a serious problem), but I think the degree to which we'd want someone to force benefits on us must vary with the magnitude and likelihood of harm the benefit might entail. I suppose the risk aversion issue could be seen as a sub-issue of agency costs, but I think it is an especially salient one.

    You say (correct me if I'm misstating) that a single chance in a thousand of the ordinary harm of rape is worth 999 chances in a thousand of considerable sexual and social benefit to the subject. Presumably, again correct me if I'm wrong, 999 chances of the ordinary harm of rape would not be justified by a 1 in 1000 chance of a sexual or social benefit. What about 99% instead of 99.9%? 90%? 75%? 50%?

    When people are in a position of trust and must act on behalf of others (attorneys, trustees, etc.), the law generally requires them to operate at a greater level of risk aversion than if they were acting on their own behalf (fiduciary duty versus ordinary negligence). And these are usually situations of consent - or at least consensual proxy, rather than the self-appointed proxy situation we have in my hypothetical and in the case of procreation.

    Ideally, the risk aversion of the proxy would mirror the risk aversion of the principal. Since it is impossible to know the risk aversion of a principal in these cases, my position is that prudence compels us to err on the side of greater risk aversion.

    Is your position that failing to provide a benefit of a given magnitude is indistinguishable, morally, from causing a harm of a given magnitude?

  4. And Chip, I definitely think cloning is better than the sexual reproduction crapshoot - too bad cloned individuals don't seem to behave any more similarly to each other than do siblings, in animal studies. Once we get the brainwave-copying thing down, it'll be slightly less troublesome. (Though I think even a brainwave-cloned exact copy of a happy individual is harmed by being brought into existence.)

  5. Curator,

    You make an important point regarding the practical reality of cloning under present technological conditions. I should have stipulated as much, though I think it's mostly understood.

  6. I knew you knew, Chip.

    I realized why I'm confusing RH - I am using the term risk aversion when what I'm really concerned with is loss aversion - not merely a preference for certainty and predictability, but a disproportionate preference to avoid harm rather than seek gain. It's an economic realization of something like Shiffrin's asymmetry.

    I constantly conflate the two, sorry!

  7. "cloned individuals don't seem to behave any more similarly to each other than do siblings, in animal studies"
    I did not know that.

  8. I didn't find anything more recent than the 2003 pig study, but I didn't find anything contradicting that, either.

  9. What a HEAP of bullshit on either side!
    Trying to solve moral dilemmas via "logic" is guaranteed to bring even greater evils.
    I know, I know, this is the legacy of the Greeks, they invented logic for this very purpose but this is nevertheless a failure.

  10. Without logic, or, for that matter, reason in general, the word "failure" has an air of meaninglessness about it... And using meaningless words is kind of a failure...

  11. Like . . . ending slavery? Democratic government? Female suffrage? Laws against child abuse and animal abuse?

    All "greater evils," apparently, brought to us by applying logic to moral issues . . . damned logic!

  12. Robin Hanson wins a rationality and intellectual integrity gold star for sure.

  13. Like . . . ending slavery? Democratic government? Female suffrage? Laws against child abuse and animal abuse?

    I love that!
    So all this is the result of "logic"?
    You are drunk on your own delusions: things are "good" because you like them (not that I object to any of the above except may be democracy).
    Trying to hijack logic to support moral/political causes is a sure way to disastrous blunders brought about by self-deception.
    And when I say disastrous I mean disastrous with respect for your own goals, like bringing the opposite of what you wish for.

  14. Anonymous- I'd agree that things are 'good' because we like them-at least, that's what I believe things boil down to- but logic is a tool by which we pursue the end of that which we like, no? When you say 'Trying to solve moral dilemmas via "logic" is guaranteed to bring even greater evils.', I have to disagree. Moral dilemmas usually entail competing impulses or edicts whose contradictory natures aren't always clearly recognized, ofttimes because of emotionally driven valuations. Logic is ideally used to sort things out, exposing inconsistencies and displaying where we're working against our own purported ethical ends. Admittedly, things don't always work out :) But if you think 'self-deception' is rooted in logic itself, instead of in misapprehension and misapplication, I believe your case is highly overstated, at best.

  15. *rolls eyes*

    Anonymous, if you're against using logic to debate ethics, why are you debating us about ethics in the first place?

  16. Anon,

    Surely you have goals and interests of your own, and never mind what. How do you justify or promote these without logic?

  17. @metamorphhh
    Logic is ideally used to sort things out, exposing inconsistencies and displaying where we're working against our own purported ethical ends.

    Totally agree on that but logic doesn't go further: it can spot trouble but cannot give you the "right answer".

    So you think discourse is purely logic?
    Rhetoricians and sophists knew better...

    How do you justify or promote these without logic?

    Excellent! You (unwillingly I think) hit right at the core of my argument.
    I say: pretending you own position ensues from pure logic is CHEATING (or at best delusional).
    Anyone has a lot of private interests from just his/her situation in the world and those are legitimate even though they may conflict with other people interests.
    Trying to deny or snuff out the confrontational aspects just makes them worse.
    Instead of longing for a "perfect" world learn to live with the iterated prisoners dilemma (Google this).

  18. Except for the hyperbole about "greater evils", I entirely agree with Anon. I can't think of any substantive ethical belief I hold or feature of my general moral outlook that does not easily admit of likely explanations which have no recourse to the reasoning I invoke to defend or promote them. (And something along similar lines is available for each of the items in Curator's retort to Anon. Even Singer, I think, concedes that it's not the arguments but the descriptions and graphic revelations of the suffering our hamburgers require that have motivated reform.) And I very much doubt that the real source of most people's antinatalism is not some variant of Schopenhauerian hedonic pessimism.

    Maybe this is harder to appreciate for you more intellectually gifted and dialectically dexterous folks for whom it may really be the case that substantial shifts in belief and outlook are mediated by articulated reasoning, but for the rest of us (if I may speak for the folk), it looks pretty much like more or less successfully unwitting, and thus the better-crafted, post-hoc-ism. Not that I'm knocking the latter -- I savor it, insatiably, as a peon at the colosseum of cerebral performance art.

    Haidt and Bloom, signatories of the recent Consensus Statement, got into it not to long ago.

  19. @Anonymous Um, debates are SUPPOSED to rely on logic and minimize rhetoric, even if they often do not.

    Now, tell everyone what *your* position is and why you are less deluded than we are. I'm waiting.

  20. Oops, I skimmed over anon's position. But it still doesn't address the point that some people may be more motivated to find the truth for its own sake. And anon still hasn't shown us how he/she is less deluded and driven by ulterior motives in putting forth the argument.

  21. This is a specious argument.

    Rape is, by definition, unenjoyable for the victim. It does not give pleasure. It is not chosen.

    It is, by definition, a one-sided affair, perpetrated by the rapist, as per the will, desires, and choice of the rapist. It is, by definition, forcing a human being to submit to another human being, physically, for the exclusive benefit of the forcer, and to the detriment of the forcee.

    What you may be TRYING to talk about is "rape fantasy," a well-known sexual kink wherein consensual partners simulate the rape experience; but the TRUE experience of rape, the terror and unwantedness--I can't imagine any ethical doctor ever thinking that it was all right to induce all the other ramifications and long-term effects, as well as committing the CRIME of rape upon a patient, in exchange for some projected sexual pleasure. No doctor in his right mind would administer treatment with such deleterious, risky, negative "side effects."

    One might also wonder about the bias of the writer, in assuming that the gain of sexual pleasure is worth the loss of personal bodily sanctity, respect, dignity, physical harm, etc.--exactly how obsessed with rape are you? I couldn't consider this question realistically for a moment and it's a painful reminder of how fucked-up people are to see people here casually mulling over a fantasyland rape-so-i-can-be-pro-rape scenario like a pleasurable tea biscuit.

    I'd personally love to see all these Great Minds of Logic!!!!! applying their self-congratulatory brainpower to solving a realistic issue that could be of some use, as argued above, rather than sitting around circle-jerking 'cos you've found a way to make rape OK.

  22. 1. I think rape is wrong, even under the facts given here.

    2. Do you understand the concept of a thought experiment?

  23. Curator, don't argue with people like the commenter before you. They are generally too stupid to grasp their own lack of reading comprehension.

  24. For a while I've been growning skeptical of using analogies (the Rape Doctor in this case) to "discover" deep moral truths like this. Bryan Caplan finally helped me put my finger on why when he discussed Daniel Kahneman's book, Thinking Fast and Slow. His two posts discussing it are here:
    The relevant Kahneman quote is:
    "I propose a simple account of how we generate intuitive opinions on complex matters. If a satisfactory answer to a hard question is not found quickly, System 1 will find a related question that is easier and will answer it. I call the operation of answering one question in place of another substitution..."

    In other words, what "Nasty thought experiments" like Ecstasy and Peanuts, Children of Earth, or Rape Doctor are doing is substituting the intuitive snap judgement for those situations for actual moral calculations. That kind of snap judgement might be useful in day to day reasoning, but in the Big Questions like natalism it leaves something to be desired. In this case in particular it gives whoever gets to choose the thought experiment a huge framing effect they can use to take control of debate.

    Basically the whole Anti-natalism vs. Repugnant conclusion debate is a giant metaphor fight where each side argues that normal procreation is more like their bizarre abstract metaphor and therefore they deserve to get the framing effect. If you reason it out, however, it seems much more likely that normal procreation is like none of those metaphors, and that reasoning by metaphor in this situation is fundamentally unsound.

    I think instead of metaphors I'll use some good old-fashioned utility calculation. You should create life if the total amount of utility (which is not the same as happiness or pleasure) it will likely experience is likely to be greater than the total amount of disutility (which is not the same as sadness or pain) it experiences. No silly metaphors.

    1. Evan do you have any idea how to actually perform your calculation? How many ordinary lives are enough to outweigh one person who burns to death?


Tweets by @TheViewFromHell