Summary

I'm currently not sure if I'm a negative-leaning utilitarian or a negative utilitarian. I have contradictory intuitions in both directions. Fortunately, the practical applicability is limited.

Introduction

I used to think that I was what I call a "negative-leaning utilitarian," i.e., a classical utilitarian who assigns very negative values to torture-level suffering, but not infinitely negative values. This means there's some amount of happiness that could outweigh any given amount of suffering no matter how intense.

But if I really reflect on it, I would not accept a day in hell in exchange for any number of days in heaven. Here I'm thinking of hell as, for example, drowning in lava but with my pain mechanisms remaining intact for the whole day. Heaven just wouldn't be worth it, no matter how long. I really wouldn't mind missing out on heaven, but boy would I scream during the day in hell. It almost seems like there's no comparison. Nonexistence is fine for me -- I wouldn't be around to miss it -- but hell-level suffering is just not something I would accept.

Inconsistent intuitions

This leaves me with four intuitions that can't all be true. Like the proverbial fourth man in a lifeboat, one of them has to go.

1. Linearity. Goodness/badness scale linearly with respect to duration and number of minds. (Given empty individualism, duration is just a dimension of other minds, so the "duration" part could be considered redundant.)
2. Continuity. For any intensity of suffering at a given moment S, it would be possible to experience another intensity of suffering S' slightly lower such that there exists a positive constant C such that C * S' is worse than S. For example, if you're burning in lava, S could correspond to 1000 degrees, and S' could correspond to 999.9 degrees. C could be set to 10 or 100 or 1000 or 3^^^3 or whatever.
3. Small pains are ok. I would accept many of the pains that people normally experience in life in exchange for a sufficient amount of happiness. For example, I would accept being awoken by my alarm clock, or nausea, etc. in exchange for extra days spent with a good friend.
4. A day in hell could not be outweighed, as discussed above.

The obvious reply to #4 is "scope insensitivity!" and this may be right. Indeed, it's the reason I thought I was (and might still be) just a negative-leaning utilitarian. But if you actually asked me what it would take to outweigh a day in hell, I would say there's nothing that could compensate, and this feeling doesn't go away no matter how much I think about the question.

I'm not sure which of my intuitions to throw out.

• I refuse to throw out Linearity, because this is even more obvious than all of the others, and you get into massive paradoxes if you don't accept it.
• Continuity seems like it should be true, but on the other hand, human brains are finite, and it's not inconceivable that a one-neuron difference could be lexically worse. Still, that feels like a big stretch (3^^^3 minutes of 999.9 degrees being not as bad as 1 minute of 1000 degrees?), and I'd rather not hang my hat on it.
• It seems like small pains are ok. That said, sometimes thinking is distorted: When we can't have something we want, we feel bad, and our desire causes suffering. If I didn't exist, I wouldn't really mind staying that way rather than popping into existence, experiencing some happy moments, and then popping back out. It still does seem like I'd rather have good moments with my friends than not exist, but the desire is actually quite weak, and it may be biased by the fact that I already exist, so I'm tempted.
• Even though it seems like the "most rational" answer is to accept a day in hell in exchange for some amount of extra happiness, I just can't do this. My brain refuses.

Now, we shouldn't expect my intuitions to be consistent. Brains are a combination of a bunch of behavioral modules that act in opposition to each other. Some are turned on more than others at various times. These hacks worked well enough in the ancestral environment, and there's no evolutionary reason for people to have a globally consistent set of intuitions. That said, I feel I need to make my intuitions consistent, so I have to abandon one of the above. I don't know which one it should be.

The practical implications are limited

In general, negative-leaning utilitarians will agree with negative utilitarians on almost all issues, so this distinction doesn't matter in practice. Creating massive computing power for the purpose of simulating pleasure is great, but the risk of it being used, at least in small part, to simulate hells will never become tiny enough that even negative-leaning utilitarians would favor creating such computing power unless it also serves to reduce suffering occurring elsewhere. In other words, the happiness parts of the equation can mostly be ignored from the calculations either way.

A note on Buddhist rhetoric

For the negative utilitarians in the audience: Maybe it would be worth exploring whether we could channel Buddhist ideas on this topic in order to show how mainstream these sentiments are. While people generally regard negative utilitarians with disdain, they often have higher regard for Buddhists, and as David Pearce says, Buddhism is a counterexample to the claim that negative-utilitarian ideas have no basis in intellectual history. I suppose it's arguable to what extent Buddhism aligns with negative utilitarianism vs. classical utilitarianism vs. other non-utilitarian ethical systems, though.

Anyway, it is amusing how much I might appear as a Buddhist to outsiders -- with my focus on suffering, concern for wild animals / insects, interest in neuroscience, etc.

Another note on rhetoric

NU sometimes seems preposterous when you raise the pinprick objection. Is a pinprick really so bad that it outweighs heaven?? I agree NU seems crazy if phrased that way. But the way I see the matter, as a tentative NU myself, is different: It's not that a pinprick is very bad; it's that heaven isn't morally good. Heaven feels wonderful for those in it, but nonexistence, or Buddhist meditation, would also be good. I'm fine with any of those options. I don't see it as important to bring heaven into existence from nothingness.

Logically, you should throw out #4. Pain intensities are finite. They are encoded by neuron firing rates, which are finite. Unless you think they encode an infinite intensity percept somehow, you should treat the disutility as finite.

If your brain refuses, this merely means you're not a utilitarian system. Like all actual human brains. (For instance, I would gladly accept one day in hell for a sufficiently large reward myself, but I wouldn't force the same trade-off on you - an inconsistency that is maintained because my social emotions don't actually update on "empty individualism")

Brian Tomasik wrote:That said, I feel I need to make my intuitions consistent, so I have to abandon one of the above. I don't know which one it should be.

Prediction: No matter what you decide, your intuitions will not remain consistent over time. You have no practical mechanism to commit to that permanently.

Hedonic Treader wrote:Logically, you should throw out #4. Pain intensities are finite. They are encoded by neuron firing rates, which are finite. Unless you think they encode an infinite intensity percept somehow, you should treat the disutility as finite.

Subjective numerical assignments are allowed to be whatever we want them to be. The fact that neurons scale in a finite way is an intuition in support of taking hell to be a finite cost relative to pleasure, but this needn't be what we assign if we have strong enough countervailing intuitions. In any event, if we throw out #3, then #4 doesn't imply that a day in hell is infinitely bad relative to other pains -- only that it's lexically worse than pleasures. I don't claim to have my mind made up on the matter.

Hedonic Treader wrote:For instance, I would gladly accept one day in hell for a sufficiently large reward myself, but I wouldn't force the same trade-off on you

But, for instance, if you supported expansion of computational power in the future, then there would be people like me on whom you would be forcing this tradeoff, to their dismay. I think you don't support expansion of computational power, so I'm just speaking hypothetically.

Hedonic Treader wrote:Prediction: No matter what you decide, your intuitions will not remain consistent over time. You have no practical mechanism to commit to that permanently.

Probably. But I have to make decisions in the meanwhile. Fortunately, as noted, the resolution of this isn't that important practically speaking.

Brian Tomasik wrote:The obvious reply to #4 is "scope insensitivity!" and this may be right. Indeed, it's the reason I thought I was (and might still be) just a negative-leaning utilitarian. But if you actually asked me what it would take to outweigh a day in hell, I would say there's nothing that could compensate, and this feeling doesn't go away no matter how much I think about the question.

Would it change your views to break it down? Let's say I offer you a thousand (don't want to make the number too high, to avoid scope insensitivity in the other direction ) years of bliss for a picosecond in hell?

What if I said we could renegotiate afterwards (whenever you liked), such that you could either choose oblivion at the end of your thousand years, or trade another hell-picosecond for as many years of bliss as you preferred? Would you accept the initial offer?

If you'd take the first picosecond just in case, I feel like we've established the sort of person you are and now we're just negotiating terms.

Arepo, do humans really have good enough intuitions about the picosecond level?

Brian Tomasik wrote:In any event, if we throw out #3, then #4 doesn't imply that a day in hell is infinitely bad relative to other pains -- only that it's lexically worse than pleasures. I don't claim to have my mind made up on the matter.

Sure, and if we only count Tuesdays, we can just schedule all the torture for Friday.

Of course, the arbitrariness objection is more of a problem for moral realists like David Pearce, who argues from a non-arbitrary God's Eye View, and intuits a negative bias nevertheless. If you reject moral realism, you can just be arbitrary.

But, for instance, if you supported expansion of computational power in the future, then there would be people like me on whom you would be forcing this tradeoff, to their dismay.

Right, and if I reject it, I forcibly take the option to live and have pleasure from those who want it.

I think you don't support expansion of computational power, so I'm just speaking hypothetically.

It's true that I don't support it, but that's not a rational judgment. It's irrational resentment and distrust because I'm not emotionally healthy.

Rationally, I already argued that all it takes is a 10% chance of an invention that creates pleasure of 10% agony intensity and which is used thousands of times more often than agony is felt. Easily plausible.

If "you" belive in Empty Individualism, it has no sense to talk about personal prefernces concerning exchange rates. If "you" decided to stay a day in hell to be rewarded with eons of bliss "you" are only sacrifying a number of prospectives selves to please a far greater number of prospective self. Whatever you do "you" are morally responsable for what "you" do to "yourself" so deciding in what way "we" should treat "others" by pondering how "we" would treat "ourselves" is useless. I don't know a way out.

I don't know if I can call myself a moral realist anymore after having written this post.

Small pains are ok. I would accept many of the pains that people normally experience in life in exchange for a sufficient amount of happiness. For example, I would accept being awoken by my alarm clock, or nausea, etc. in exchange for extra days spent with a good friend.

Brian , if you dare to doubt your intuitions a bit more, you might find this my following thoughts on the pinprick-argument helpful.

The last days I've made up my mind how to answer to the pinprick-argument because it contradicts my intuition that we should destroy a world full of extraordinary bliss just to avert the pain of an pinprick. Through reading about modern physics like the relativty theory I know that time, as we perceive it, is an illusion.

Time does not "pass" and what we call the past and the future does exist even if not on the same timeslice. This view is called Eternalism.Time as well as space can be divided into finitive small units called planck time and planck space. A planck time is ~ 0,5 * 10^-44 seconds and so to say a physical moment. Within the planck time there no time "passing" so you can call this physical moment "timeless" or even "eternal" (not to mistake for "endless duration of time") So if there's a planck time of pinprick-induced suffering we might also call it timeless or eternal suffering. If we do so the disvalue of this planck time should be regarded as infinitive. (But as we can count several planck times of pinprick-induced suffering it is not unimportant how much plancktimes of infinitive suffering exist.)

I think most people don't find a pinprick that harmful because they believe that there will be a time in the future in which the pain of the pinprick will not be existent anymore. But it exist. It exists in another time! And in that time it will never cease to exist! It only ceases to exist in the future.

So if you talk of the pain of a pinprick people suppose that the pain is finitive because they are presentists and think that only the present exists. But if a second has ~2*10^43 planck times a second of pinprick-induced pain means in fact ~2*10^43 timeless/infinitive moments of suffering. As soon as people acknowledge that Presentism is false they will think differently about the pinprick argument.

Humphrey, according to this view all the bliss also doesn't cease to exist. Of course, the pain and the bliss exist in different locations/perspectives, and you could have the intuition that it is unjust to cause a bad state in one to create a better state in another. But negative utilitarianism doesn't solve this problem, since it condones inflicting a worse pain in one location to alleviate more minor pains in other locations.

Yes, you are right on that, Hedonic Treader. For a classical utilitarian it is irrelevant whether Presentism or Eternalism was true. But for Negative Leaning Utilitarians who say that they give more importance on higher amounts of suffering it could make a change. If the pain of a pinprick does not cease to exist we could talk about infinitive suffering so that a NLU could see even a pinprick as an sufficiently huge amont of suffering to give it absolute prioroty so that he could be a Negative Utilitarian as well. I don't know if Brian would accept a pinprick if he thought of it causing eternal pain who just cannot recognized as such because of our illusional intuition that time (and with it the pain) passes.

Thanks for the comments, all!

Arepo wrote:Let's say I offer you a thousand (don't want to make the number too high, to avoid scope insensitivity in the other direction ) years of bliss for a picosecond in hell?

A picosecond is probably too short to even count as a coherent experience. The neural firings would take orders of magnitude longer than that to even kick around in my brain. I think we'd need something like at least a decent fraction of a second for me to get my bearings and even figure out what was going on. With that said, I would not accept a fraction of a second of hell in exchange for 1000 blissful years.

Hedonic Treader wrote:If you reject moral realism, you can just be arbitrary.

That's my boy, HT.

Hedonic Treader wrote:Rationally, I already argued that all it takes is a 10% chance of an invention that creates pleasure of 10% agony intensity and which is used thousands of times more often than agony is felt. Easily plausible.

Yeah, I haven't replied to that yet because I'm slow like a turtle. However, I did read it, and that's part of what prompted this post.

Humphrey Schneider wrote:If "you" belive in Empty Individualism, it has no sense to talk about personal prefernces concerning exchange rates.

In general, my egoistic preferences may be a decent guide to altruistic preferences when extended to others -- c.f., the Golden Rule and all that. This isn't to say other people have my specific tastes (e.g., liking mushrooms), but it is to say that my egoist motivations for myself can be a way to motivate/inform altruism.

Humphrey Schneider wrote:I don't know if Brian would accept a pinprick if he thought of it causing eternal pain who just cannot recognized as such because of our illusional intuition that time (and with it the pain) passes.

As Elijah says, you're imagining the pinprick as lasting forever, but it actually doesn't. It lasts for a moment, and that moment is real, but it's not eternal. If you wanted to regard it as infinite, then every moment would be infinite.

If you want to introduce infinities, you can already do that using the fact that I exist infinitely many times in the multiverse. So one pinprick implies infinitely many of them.

As Elijah says, you're imagining the pinprick as lasting forever, but it actually doesn't. It lasts for a moment, and that moment is real, but it's not eternal. If you wanted to regard it as infinite, then every moment would be infinite.

I have problems to express it in words, but I don't said that the pinprick lasts forever. Here we just need a tern that excludes any idea of time. I wanted to say that the pain of a pinprick of in fact beyond time and therefore timeless. I distinguish between "timelessness" and "infinitive duration of time", but for both notions I use the term "eternal" in the lack of something better hwat might lead to misunderstandings. The moment I meant is timeless eternal as for platonics numbers are timeless eternal. Sorry for not finding a better example.

I consider a moment to be infinitive in relation to pain/pleasure felt in this moment but finitive in realtion to other moments. In my notion of time every moment is timeless and could (if qualia is experienced) have infinitive (dis-)value.

To my mind, the feeling that time passes is an illusion. We just can't experience the time passing. It's a mental costruction. We compare the sensory expressions of the present with sensory expressions of some thing we call the past. Time is a explaination for the experience that the "past" can be known but the "future" cannot. I think the fact that humans usually have no idea of the so called future has to do with how the physics in our brain works. May be the second law of termodynamics just doesn't allow to turn the Arrow of time of the processes that cause our sensory expressions and our memories. If these processes were reversible we would also "remember" the future. And if we knew the future, I think we would abolish the idea that time passes.

I consider time to be similar to space. I you go from A to B, you don't clam that A doesn't exist if you are in B. A is just somewhere else. When I say that the past exists I don't mean that it still exists. Because if I maent this I would say that the past exists right now. That would be almost the same nonsense as if I said that A would exist in B. The past doesn't exist in the present. The past exist in his own time. And in this time, time will never pass.

Brian Tomasik wrote:With that said, I would not accept a fraction of a second of hell in exchange for 1000 blissful years.

Huh. If I thought this is realistic, I would work really hard to get this option.

Well, I think that many people would take a day in hell for 1000 years of heaven and I think they would definitely do it if it was just a picosecond. Your choice here seems idiosyncratic and I hope you wouldn't want to bind other organisms to make the same choice.

(Side note:) In a Facebook comment, Karn gives an interesting way to express a lexical ordering using ordinal numbers. Assign each intensity to a positive integer i. Then the badness of k organisms suffering at intensity i is k * omega^i.

I refuse to throw out Linearity, because this is even more obvious than all of the others, and you get into massive paradoxes if you don't accept it.

A little off topic, but I'm not aware of any paradoxes with non-linearity. Time and space discounting have paradoxes, but what paradox would a utilitarianism that, for instance, values duration and number of minds logarithmically run in to?

amc wrote:A little off topic, but I'm not aware of any paradoxes with non-linearity. Time and space discounting have paradoxes, but what paradox would a utilitarianism that, for instance, values duration and number of minds logarithmically run in to?

The main paradox is just that the amount something matters depends on how much suffering there already is. For instance, you think a murder is really bad, but you find out there have been googols of murders in the past, so meh, this additional murder is basically negligible. Where do you start counting? Now? The beginning of the universe? Do you count suffering in other universes?

Alastair Norcross has some discussion of paradoxes of violating linearity in section VI of "Two Dogmas of Deontology: Aggregation, Rights, and the Separateness of Persons," although this is focused on a model where different harms have different aggregation counters.

Also, anything but linearity seems ad hoc. Who came up with log? Why not sigmoid? Arctan? etc.

Brian Tomasik wrote:

Hedonic Treader wrote:Logically, you should throw out #4. Pain intensities are finite. They are encoded by neuron firing rates, which are finite. Unless you think they encode an infinite intensity percept somehow, you should treat the disutility as finite.

Subjective numerical assignments are allowed to be whatever we want them to be. The fact that neurons scale in a finite way is an intuition in support of taking hell to be a finite cost relative to pleasure, but this needn't be what we assign if we have strong enough countervailing intuitions. In any event, if we throw out #3, then #4 doesn't imply that a day in hell is infinitely bad relative to other pains -- only that it's lexically worse than pleasures. I don't claim to have my mind made up on the matter.

If you take " numerical assignments are allowed to be whatever we want them to be" and go far enough, at some point the proposition will become entirely irrational and meaningless. If my finger is crushed by 10,000 tonnes of lead, it would be no different if the same part of my finger were crusshed by 100,000,000 tonnes of lead. If you want to assign different preferences to the two, you can do so but I am not aware of any meaningful basis for doing so. There is also something called sensory adaption (or sensory adaptation)--if you experience the same thing for a long time, you stop noticing it. I'm not sure how this would work in the imaginary case of burning in hell or a pool of lava that never kills you.

One if the earliest scientific studies of sensation addresses the issue of linearity and the limits of sensation. Fechner did many studies in the 1800s, many of which are simple. He would do things like touch someone's back with two pencils and see if they could tell the difference between two pencils that were close together or one pencil alone. Our perception of sound is nonlinear (two bells rining don't sound twice as loud as one). Here is an excerpt of some of his ideas: http://psychclassics.yorku.ca/Fechner/

Hi myacct.

myacct wrote:If you want to assign different preferences to the two, you can do so but I am not aware of any meaningful basis for doing so.

"Meaningful basis" is itself a subjective judgment call. I agree that I wouldn't value the cases differently, but there's nothing to say someone can't legitimately do so.

In my 7th-grade social studies class, we sometimes had little quizzes where we had to distinguish factual statements from opinions. "Meaningful basis" falls qualifies as an opinion.

myacct wrote:One if the earliest scientific studies of sensation addresses the issue of linearity and the limits of sensation.

Yes, but the linearity I'm referring to is moral linearity across experience-moments, not linearity of perception in a given experience-moment with respect to some objective stimulus.

Some random additional notes.

I would accept a headache for enough orgasms. As far as my discussion above, I realized that maybe hedonic zero is the wrong cutoff point for me. Maybe the best resolution is that suffering below a certain threshold of badness lexically dominates that above it. Headaches and pinpricks aren't below that threshold, but brazen bull is, for example. This doesn't violate continuity below the threshold, and above the threshold, everything is essentially zero relative to what's below it.

A friend wrote: "we lean towards the view that it's the non-existence of what we call unfulfilled immediate desires/urges [...]."
You can point to a vast number of Buddhists who agree:

Estimates of the precise number of Buddhists in the world vary between 350 and 1,500 million, making Buddhism the second, third or fourth largest world religion.

Can classical utilitarians counter those numbers?
If Buddhism didn't believe in reincarnation, some might be doing pretty similar things as modern-day NUs.

Brian Tomasik wrote:Maybe the best resolution is that suffering below a certain threshold of badness lexically dominates that above it.

Now I'm second-guessing this. I would accept a huge number of headaches to avoid a day in hell, but maybe not, say, 3^^^^3 days of headaches. In contrast, I would probably give up 3^^^^3 days of happy life to avoid a day in hell. There's not really a huge loss to forgo the pleasure, but to have a headache for 3^^^^3 days would be pretty unbearable. Maybe hedonic zero is the natural cutoff point after all.

In my opinion, the continuity argument against NLU is decisive. But notice that there are other arguments against that position. For example, once you are prepared to accept that pains below a critical level of intensity lexically dominate those above it, it seems arbitrary to assume that there is just one critical level, rather than many.

More generally, I must say I'm surprised at the appeal that this position has on people I highly respect intellectually. NLU is not only extremely hard to articulate theoretically, but is also easily explained as motivated by various biases (scope insensitivity, unavailability of pleasurable memories of comparable intensity, etc.), which themselves have plausible evolutionary explanations (cf. Carl Shulman's remarks). Given this, it seems to me far more sensible to dismiss the intuitions supporting NLU as mere cognitive illusions, and desist of further attempts to construct a moral view out of them. (Even moral anti-realists agree that some preferences should be accorded less weight than others, or no weight at all, in deciding which things have value, and how much value they have. So this option of dismissing such intuitions is available to realists and anti-realists alike.)

Pablo Stafforini wrote:In my opinion, the continuity argument against NLU is decisive.

I think you mean NU, right? NLU has no continuity problems.

Pablo Stafforini wrote:But notice that there are other arguments against that position. For example, once you are prepared to accept that pains below a critical level of intensity lexically dominate those above it, it seems arbitrary to assume that there is just one critical level, rather than many.

Well, this isn't necessarily a problem. Maybe it's true, and we should focus on the absolute worst. Anyway, as I said in my previous comment, I'm backing off on the threshold proposal.

Pablo Stafforini wrote:NLU is not only extremely hard to articulate theoretically

Not at all. As I explained in another post, NLU is just classical utilitarianism. Everyone has to choose the numerical value of every experience; these are free parameters because "amount of happiness" and "amount of suffering" are not objective things.

Pablo Stafforini wrote:More generally, I must say I'm surprised at the appeal that this position has on people I highly respect intellectually.

What it comes down to is that my intuition that suffering is vastly more important than happiness is stronger than my intuition that I should be an aggregative utilitarian. The only reason I'm utilitarian is because I feel it's the right thing, but in fact, I don't feel that creating a future that contains both massive bliss and some massive torture is acceptable, and I feel the latter more strongly than the former.

Pablo Stafforini wrote:So this option of dismissing such intuitions is available to realists and anti-realists alike.

Yes, and I just exercised that option. Apparently you don't agree. Anyway, as was clear from the opening post, even my intuitions aren't consistent. It's like an optical illusion where you see either one thing or the other thing depending on how you look at it. At the end of the day, though, I know it wouldn't be right to allow hell-level suffering to be created even for the sake of additional happiness.

In a comment on "Reducing wild-animal suffering," Nonhuman Emancipation made a seemingly light-hearted but actually pretty deep statement:

Ending life (unless it's for the benefit of the one whose life it is) =
Starting new life =
Preventing new life from starting =

I don't necessarily agree with it entirely, but I think it represents a potentially persuasive way to cast NU to make it more appealing to others. In particular:

• Many people object to NU's willingness to exterminate life, both because they don't like killing intrinsically and also because they object to authoritarian policies that don't respect individual autonomy. Point #1 eliminates that worry.
• In practice, the best way to prevent suffering is almost always going to be preventing it from coming into existence. This is true both for wild animals and for future computations. Therefore, point #3 is really all we care about from a practical standpoint, and yet point #3 is not that counterintuitive to some people.
• Many people have the intuition that they'd like to keep living but don't as strongly have the intuition that they would have been displeased not to have been born. There's also a classic asymmetry between people feeling it wrong to create new suffering life but not wrong to fail to create new happy life. These intuitions make Nonhuman Emancipation's perspective more plausible to common-sense moral perspectives.

One place where Nonhuman Emancipation's proposal could get into trouble is that it might seem as though it's suggesting that death is intrinsically bad, when in fact death is only circumstantially bad because of all the other impacts it has on hedonic experiences (c.f., Hedonic Treader's anti-authoritarian arguments). But Nonhuman Emancipation's statements don't say explicitly that death itself is intrinsically bad. In any event, many people feel that it is, so we may not want to step on those toes.

Brian Tomasik wrote:Starting new life =
Preventing new life from starting =

Hm. If we had a blueprint of a happy being who never suffers involuntarily, it would seem grossly wrong to say it's irrelevant whether that being comes into existence or not. The option to live has non-trivial positive value.

Many people have the intuition that they'd like to keep living but don't as strongly have the intuition that they would have been displeased not to have been born.

This is logically broken. You have to exist in order to be displeased. I think it's more coherent to say some people would not want to repeat their past experience, or be reborn into a roughly similar circumstance. The question here is, is this due to the context of their individual past, of childhood, or of a deeper asymmetry between death and non-birth. In my personal opinion, the option of really reliable painless suicide would solve the problem of involuntary suffering (modulo speculations about hell and the early childhood stage where voluntary suicide isn't realistic).

Elijah's approach of listing things below the threshold has something going for it, and it's what I meant when I suggested the threshold proposal. The problem comes when we have to compare things slightly below the threshold with those slightly above. For example, a single rape may not qualify, but what about 3^^^3 of them? Surely that would be worse than Auschwitz? But then Auschwitz wouldn't be enough to destroy 3^^^3 worlds, for example. So in practice, this position amounts to negative-leaning utilitarianism unless we do actually break continuity.

Brian Tomasik wrote:As I explained in another post, NLU is just classical utilitarianism. Everyone has to choose the numerical value of every experience; these are free parameters because "amount of happiness" and "amount of suffering" are not objective things.

That post merely asserts these claims; it doesn't argue for them. As I point out, it is an open question whether there is an objective matching between pleasure and pain intensities. If there is such an objective matching, non-classical forms of utilitarianism (including both NU and NLU) are arbitrary in a way that classical utilitarianism isn't.

What it comes down to is that my intuition that suffering is vastly more important than happiness is stronger than my intuition that I should be an aggregative utilitarian. The only reason I'm utilitarian is because I feel it's the right thing, but in fact, I don't feel that creating a future that contains both massive bliss and some massive torture is acceptable, and I feel the latter more strongly than the former.

I am aware of that, but my point is that some of your intuitions may be susceptible to debunking explanations. In that case, the fact that these intuitions are stronger is not by itself evidentially decisive. It is unclear to me that you are taking this objection, which Carl Shulman has pressed before, sufficiently seriously.

Elijah, the intuition that pain is much more important than pleasure seems to rest partly on phenomenal beliefs about how it's like to have pleasurable and painful experiences of various intensities. If some of these beliefs turn out to be unreliable, that should undermine the original intuition.

The most intense pains we can experience are far more intense than the most intense pleasures we can experience. It is plausible to suppose that this fact about our psychology explains the intuitive appeal of NU and NLU. But this is a contingent consequence of the fact that, to quote Shulman, "in the lifecycle of some organism there are more things that it is important to avoid than to approach." So it seems irrational to allow this feature of our psychology to play a role in our moral theorizing.

(And no, the fact that you are a moral anti-realist is irrelevant. Even as an anti-realist, you want your preferences to be formed on true beliefs about the world, including beliefs about how very painful experiences and very pleasurable experiences feel like.)

[blockquote]Suppose one person is in Hell and, e.g., Graham's number are in heaven (and the pleasure and pain in both are extremely and equally intense, and infinitely long-lasting). Then the CU or NLU would conclude that either 1) such a world is better than a world with no sentience (since even an extreme NLU would probably accept a 1:G ratio for happiness:suffering) or 2) that the moral value of such a world is indeterminate (because of infinite levels of suffering).[/blockquote]

Yeah, I think CU would conclude (1) because ω*G > ω*1.

I think we should all be skeptical of intuitions that appeal to scenarios involving very large numbers, such as comparisons involving very many people, very long durations or very high intensities. As Broome notes,

[W]e have no reason to trust anyone’s intuitions about very large numbers, however excellent their philosophy. Even the best philosophers cannot get an intuitive grasp of, say, tens of billions of people. That is no criticism; these numbers are beyond intuition. But these philosophers ought not to think their intuition can tell them the truth about such large numbers of people.

For very large numbers, we have to rely on theory, not intuition. When people first built bridges, they managed without much theory. They could judge a log by eye, relying on their intuition. Their intuitions were reliable, being built on long experience with handling wood and stone. But when people started spinning broad rivers with steel and concrete, their intuition failed them, and they had to resort to engineering theory and careful calculations. The cables that support suspension bridges are unintuitively slender.

Our moral intuitions are formed and polished in our homely interactions with the few people we have to deal with in ordinary life. But nowadays the scale of our societies and the power of our technologies raise moral problems that involve huge numbers of people. […] No doubt our homely intuitive morality gives us a starting point, but we have to project our morality beyond the homely to the vast new arenas. To do this properly, we have to engage all the care and accuracy we can, and develop a moral theory.

Indeed, we are more dependent on theory than engineers are, because moral conclusions cannot be tested in the way engineers’ conclusions are tested. If an engineer gets her calculations wrong, her mistake will be revealed when the bridge falls down. But a mistake in moral theory is never revealed like that. If we do something wrong, we do not later see the error made manifest; we can only know it is an error by means of theory too. Moreover, our mistakes can be far more damaging and kill far more people than the collapse of a bridge. Mistakes in allocating healthcare resources may do great harm to millions. So we have to be exceptionally careful in developing our moral theory.

Pablo Stafforini wrote:I am aware of that, but my point is that some of your intuitions may be susceptible to debunking explanations. In that case, the fact that these intuitions are stronger is not by itself evidentially decisive. It is unclear to me that you are taking this objection, which Carl Shulman has pressed before, sufficiently seriously.

That my intuition has an evolutionary explanation doesn't necessarily debunk it. The fact that you don't like suffering also has an evolutionary explanation. Among our grab-bag mix of evolved intuitions, we pick which ones we want to let override other, contradictory ones. I can decide that my evolved impulse to regard extreme suffering as more important than extreme pleasure is allowed to override other intuitions.

Pablo Stafforini wrote:So it seems irrational to allow this feature of our psychology to play a role in our moral theorizing.

(And no, the fact that you are a moral anti-realist is irrelevant. Even as an anti-realist, you want your preferences to be formed on true beliefs about the world, including beliefs about how very painful experiences and very pleasurable experiences feel like.)

It's not irrational unless you have the intuition that it's irrational, which begs the question. As far as having true beliefs about what those experiences feel like, this gets back to our earlier disagreement. I maintain there aren't objective facts about the morally relevant numerical intensities of various experiences. (Of course there are objective facts about experiences, like how many neurons are firing for how long at what rate, etc. The question is how these objective numbers map to what we care about morally.)

I like the Broome quote.

Brian Tomasik wrote:That my intuition has an evolutionary explanation doesn't necessarily debunk it. The fact that you don't like suffering also has an evolutionary explanation.

Oh, I agree. In saying that there is an evolutionary debunking explanation of your intuition, I didn't mean to say that the intuition can be debunked because there is an evolutionary explanation; I meant to say that there is an evolutionary explanation which is also a debunking explanation. An explanation of some belief is debunking, roughly, iff (a) it actually explains that belief and (b) it explains it in a way that undermines our reasons for thinking that the belief is true.

In the case at hand, however, the evolutionary explanation plays a secondary role. The main "debunking" role is played instead by the independently plausible observation that we can remember or imagine painful states that are far more painful than the most pleasurable states we can imagine or remember. (Evolution strengthens the plausibility of this observation by providing a natural rationale for why this is so.)

Brian Tomasik wrote:Among our grab-bag mix of evolved intuitions, we pick which ones we want to let override other, contradictory ones. I can decide that my evolved impulse to regard extreme suffering as more important than extreme pleasure is allowed to override other intuitions.

Your phrasing is, I believe, misleading. It is true that, for a moral non-realist, there is ultimately no mind-independent truth to which intuitions must conform. However, many of these intuitions are sensitive to factual beliefs which realists and non-realists alike agree can be true or false. If you agree that, because of limitations of memory and imagination, you cannot imagine pleasures that are comparable in intensity to extremely intense pains, then this should obviously have the effect of undermining your intuition about "non-classical" exchange rates. This is because that intuition rests on the assumption that you are representing the relevant phenomenal states accurately.

(If you dispute that there is a determinate intensity of pleasures that corresponds to pains of extreme intensity, just understand my claims deferentially, as referring to that intensity of pleasure that people like me, who don't dispute that there is an objective correspondence between intensities of pleasures and pains, would interpret as the reference of those claims.)

Brian Tomasik wrote:It's not irrational unless you have the intuition that it's irrational, which begs the question. As far as having true beliefs about what those experiences feel like, this gets back to our earlier disagreement. I maintain there aren't objective facts about the morally relevant numerical intensities of various experiences. (Of course there are objective facts about experiences, like how many neurons are firing for how long at what rate, etc. The question is how these objective numbers map to what we care about morally.)

No, the question is not how these objective numbers map to what we care about morally. The question is a factual question about phenomenology, not a moral question about concern. Even if no one cared about anything, I could still ask if, for a pain of a given intensity, there is some pleasure whose intensity matches it. (Analogously, you can ask yourself whether the brightness of some shade of red is matched by the brightness of some shade of yellow. This is clearly not a moral question.)

Frankly, I don't understand the basis of your high confidence in the view that there is no objective conversion rate between pleasure and pain intensities. After all, presumably you believe that there is a conversion rate between the intensities of different pains (or pleasures). For example, it seems that one pain can be twice as intense as another. Or do you deny this, too?

Concerning the issue of debunking explanations, perhaps an analogy might help. Suppose I try two different types of food and, as a result, form certain beliefs about how much food of the first type I would need to be given in exchange for a given amount of food of the second type. Now suppose a friend tells me that before trying one of these foods I drank a beverage which impaired my sensitivity to certain determinants of flavor (sweetness and umami, say). This person then insists that, for this reason, I should not place too much confidence on the "exchange rates" I arrived at when considering how these foods tasted like. What would you think of me if I then gave the following reply to my friend? "See, I am a culinary non-realist, so it's really up to me what culinary "intuitions" I should reject. If I want to, I can decide that the "intuition" that my exchange rate is correct should trump the intuition I should be skeptical about it because it was formed under unreliable conditions." Clearly, this reply would exhibit a failure to understand the point my friend was making. His point was that, when I formed the "intuition' that one type of food was n times better than the other food, I was tacitly assuming that my memories about how these foods tasted like were reasonably accurate: it was on the basis of this assumption that I settled on a particular exchange rate. So obviously, the proper reaction when someone informs me that the memories were not accurate is to revise that exchange rate. The fact that, ultimately, "it is up to me" which foods I pick and how much value I place on them is neither here nor there.

Elijah, I think you need to have a good grasp of both heaven and hell in order to reach conclusions about tradeoffs between one and the other. Otherwise, your position comes perilously close to that of a classmate of mine from elementary school, who insisted that his father earned more money than mine in spite of acknowledging that he didn't know how much money my father earned.

If my classmate's father was Bill Gates, he would have been justified only if he also knew certain (well-known) facts about the wealth of Gates relative to that of the average individual. If instead all he knew was that his father had a net worth of a hundred billion dollars, but ignored whether the wealth of the average person was a hundred thousand dollars or a hundred quadrillion, his claim would have been clearly unwarranted.

But analogies aside, I think the general point stands: to make an informed choice between two options you need to have sufficient information about each option. And in any case, Brian doesn't dispute this: he thinks that pains of some intensity are much worse than pleasures of that same intensity (using 'same intensity' to mean what the classical utilitarians mean by that expression) because, when he imagines what it's like to have these experiences, he concludes that his desire for the absence of one is much stronger than his desire for the presence of the other. It seems that for such a person it should be highly relevant whether the experiences in question are imagined accurately.

Pablo Stafforini wrote:No, the question is not how these objective numbers map to what we care about morally. The question is a factual question about phenomenology, not a moral question about concern. Even if no one cared about anything, I could still ask if, for a pain of a given intensity, there is some pleasure whose intensity matches it. (Analogously, you can ask yourself whether the brightness of some shade of red is matched by the brightness of some shade of yellow. This is clearly not a moral question.)

I don't know how to decide the "intensity" of a pain or pleasure relative to other experiences without asking the question of how I would trade it against other experiences, but once I start asking that question, it's a moral decision.

I suppose I could use some concrete measure of intensity, like number of affective neurotransmitters released per unit time or whatever. If so, that's fine, but then it loses its relevance to the discussion.

Pablo Stafforini wrote:For example, it seems that one pain can be twice as intense as another. Or do you deny this, too?

I deny this too. How bad one pain is relative to another is a moral question, or if you make it into a factual question by defining a narrow measure, then it loses its moral relevance.

Brian, do you deny that one sound can feel (roughly) twice as loud as another, or that one color can feel (roughly) twice as bright as another?

In general, and as I said many times, I think it's a bad idea to "moralize" a question that can be understood descriptively, and that is meant to be so understood by the person asking the question. It seems to me that we can consider questions about the intensity of pains very much in the same way we can consider questions about the loudness of sounds or the brightness of colors.

Pablo Stafforini wrote:Concerning the issue of debunking explanations, perhaps an analogy might help. Suppose I try two different types of food and, as a result, form certain beliefs about how much food of the first type I would need to be given in exchange for a given amount of food of the second type. Now suppose a friend tells me that before trying one of these foods I drank a beverage which impaired my sensitivity to certain determinants of flavor (sweetness and umami, say). This person then insists that, for this reason, I should not place too much confidence on the "exchange rates" I arrived at when considering how these foods tasted like. What would you think of me if I then gave the following reply to my friend? "See, I am a culinary non-realist, so it's really up to me what culinary "intuitions" I should reject. If I want to, I can decide that the "intuition" that my exchange rate is correct should trump the intuition I should be skeptical about it because it was formed under unreliable conditions." Clearly, this reply would exhibit a failure to understand the point my friend was making. His point was that, when I formed the "intuition' that one type of food was n times better than the other food, I was tacitly assuming that my memories about how these foods tasted like were reasonably accurate: it was on the basis of this assumption that I settled on a particular exchange rate. So obviously, the proper reaction when someone informs me that the memories were not accurate is to revise that exchange rate. The fact that, ultimately, "it is up to me" which foods I pick and how much value I place on them is neither here nor there.

Nice example.

I think what Pablo is trying to debunk is my feeling that dolorium is more bad than hedonium is good even if hypothetical minds that can experience them report that hedonium was as good as dolorium was bad and that 2 minutes of hedonium could outweigh 1 minute of dolorium.

Let me quote a sentence from Paul Christiano that has stuck with me:

I accept that all of my values are contingent, but am happier to accept those values that are contingent on my biological identity than those that are contingent on my experiences as a child, and happier to accept those that are contingent on my experiences as a child than those that are contingent on my current blood sugar.

Could you rewire my brain such that I would accept 2 minutes of hedonium for 1 minute of dolorium? Probably, with sufficient expertise. Could you create a brain that would accept 3^^^^3 minutes of dolorium for 1 minute of hedonium? Yes, you could do that too. Brain-space is big, and lots of designs are possible.

My current design happens to feel that sufficiently intense suffering can't be outweighed by any amount of pleasure. This is a result of evolutionary asymmetry, yes, but I don't think it's debunked; it's actually quite central to who I am now. (On Christiano's hierarchy, it's actually at the most solid level, since it's due to my genes.)

So, in reference to Pablo's beverage analogy, the beverage is not just a fleeting influence like Christiano's blood-sugar level. It's actually a permanent feature of my values. I will be drinking that beverage for the rest of my life, so I want to act in accordance with what it tells me. Sure, other people don't feel the same way, and possible modifications of my brain don't feel the same way, but that's not very important to me.

---

Now, what if you could change my brain without changing my genes? What if you hooked me up to a machine that made me feel that its pleasure was worth experiencing hell for? On Christiano's hierarchy, this would be less intrusive to his values. However, on my hierarchy, I would probably still say I don't agree with it. Hooking up to such a machine would for me constitute more of a failure of goal preservation than it would constitute a victory of enlightenment.

To make this clear, suppose I hooked you up to a machine that gave you such intense agony that you felt no pleasure could outweigh something that bad. I think you too would regard this as a failure of goal preservation rather than a victory of enlightenment. There's an arms race of possible brain changes, and especially with digital minds, it can become really easy to modify an agent in any which way. This is not progress; it's more like spinning the color wheel in Twister.

Pablo Stafforini wrote:Brian, do you deny that one sound can feel (roughly) twice as loud as another, or that one color can feel (roughly) twice as bright as another?

I could probably do that for sounds and colors. If you absolutely required me, I could make some inadequate approximation about pain/pleasure intensities too, but as a friend of mine once said, emotions often feel pretty "incomparable." We must compare them because we have to make decisions, but they don't naturally fall onto the same scale. The only natural scale seems to be the moral one.

Anyway, even if we assigned pain/pleasure intensities non-moral numerical values, what would that matter to the discussion at hand?

Brian, there are all sorts of cognitive and perceptual illusions that have a genetic basis, and which you don't regard as any more veridical for that reason. To take an example that is close to the issue at hand, it is well-established that, in assessing the duration of past experiences, we don't integrate over the entire relevant episode, but instead average the final and most intense moments (the so-called peak-end rule). I suspect that, when learning about this phenomenon, you don't react by saying something like: "Well, that's how I'm wired." Instead, you try to use other parts of your brain to correct for the expected distortions.

Also, it seems odd to attach significance to whether the bias is genetic, developmental or hormonal in origin. Suppose you stumble upon a journal article that describes a bias which you hadn't heard about before. The article contains two sections, one on the nature of the bias and one on its causal origins. As you read the first section, you tell yourself: "Gee, this bias looks really weird. But wait! Before I judge how bad it is, or whether it is bad at all, I must read the next section." That strikes me as a very peculiar reaction, frankly.

even if we assigned pain/pleasure intensities non-moral numerical values, what would that matter to the discussion at hand?

It would matter because the claim that we cannot make purely descriptive comparisons between pain and pleasure intensities would then become much less plausible. If you agree that we can truly say that one pain is (roughly) twice as intense as some other pain, it seems that you should either agree that we can also truly say that one pain is (roughly) twice as intense as some pleasure, or explain why the analogous comparison cannot be made.

Incidentally, ordinary people seem to be able to make all sorts of magnitude estimations within and across sensory modalities without much difficulty. See S. S. Stevens, Psychophysics (New York, 1975).

Pablo Stafforini wrote:It would matter because the claim that we cannot make purely descriptive comparisons between pain and pleasure intensities would then become much less plausible.

Not sure why that matters.

I think different types of pain are themselves incomparable but more comparable than pain vs. pleasure. Apples and oranges are more comparable than apples and rocks.

You said:

We must compare them because we have to make decisions, but they don't naturally fall onto the same scale. The only natural scale seems to be the moral one.

If you accept that pains of different intensities do fall onto the same scale, then it would be false that "the only natural scale seems to be the moral one." If you then want to insist that pain and pleasure do not fall onto the same scale, your appeal to the supposed "naturalness" of the moral scale looses its force.

In any case, I feel like I'm arguing against a moving target. You say that "different types of pain are themselves incomparable", but you also say that you could "probably" make comparisons between different sounds or colors, and that you could make "inadequate" comparisons between pain and pleasure intensities. It is still unclear to me whether you believe that we can make descriptive comparisons between different pain intensities.

Sorry for being so difficult.

I would accept that we can compare emotions of different intensities depending on what "compare" means. Heck, I would say we can compare apples and rocks if you choose the right comparison function. An apple can be softer than a rock, more red, more full of fructose, etc. So, sure, we can compare basically anything given a comparator.

The reason I was reluctant to admit this was because it's a red herring. That objects can be compared (in all kinds of ways) doesn't inform us about our moral perspective on the comparison unless we choose to tie our moral perspective to one or more of the other comparison functions that are open to us.

Thanks for the clarification.

Some readers might have lost grasp of the structure of this discussion by now, so here's a summary of the various claims I've been making, in case it is of help to anyone:

(1) We can sometimes compare things without invoking any evaluative standard. Let's use the term 'd-compare' for comparisons of this sort.

(2) We can d-compare the brightness of colors and the loudness of sounds. For instance, we can say that one sound is roughly twice as loud as another, or that one color is roughly twice as bright as another.

(3) We can also d-compare the intensity of painful experiences. For instance, we can say that one pain is roughly twice as intense as another.

(4) We can also d-compare the intensity of painful and pleasurable experiences. For instance, we can say that one pain is roughly twice as intense as some pleasure.

(5) Claim 3 becomes more plausible if claim 2 is true.

(6) Claim 4 becomes more plausible if claim 3 is true.

(7) Hedonistic utilitarians, whether classical or negative-biased, care about pain and pleasure intensity, so if claims 3 or 4 are true, this would be highly morally relevant.

Thanks for the summary, Pablo!

I'm stuck on claim (1). What does it mean to compare something without an evaluative standard?

peterhurford wrote:What does it mean to compare something without an evaluative standard?

It means that, even if there was nothing you cared about, or nothing that was objectively good or bad, you'd still be able to make the comparison. For example, I can compare the spin of two elementary particles, and say that the spin of one (say, a neutrino) is twice that of the other (say, a boson). Clearly, no evaluative standards are being invoked by someone making comparisons of this sort.

Makes sense.

Are these statements observer-specific? For example, I can agree that jamming my finger in a door is roughly three times as painful as stubbing my toe, but I would have no way of verifying whether Peter getting his finger jammed in a door is roughly three times as painful as Pablo stubbing his toe.

I think the same might be true for colors or sounds in a few corner-case scenarios, but not on the whole.

The analogy is good, though. If people experience colors, sound, taste in roughly the same way, it stands to reason by extrapolation that people should experience pain in roughly the same way too.

Hi Peter,

peterhurford wrote:Are these statements observer-specific? For example, I can agree that jamming my finger in a door is roughly three times as painful as stubbing my toe, but I would have no way of verifying whether Peter getting his finger jammed in a door is roughly three times as painful as Pablo stubbing his toe.

Different physical events--such as getting one's finger jammed in a door, or stubbing one's toe--can cause different experiences in different people, or even in the same person at different times. But insofar as the statements mention physical events, it is only as a means for referring to certain experiences. Ultimately, the experiences are meant to be characterized by what it's like to have them rather than by what physical events cause them. If we had a public way of pointing to experiences, we could say: "An experience that feels like this is roughly three times as painful as an experience that feels like that." The comparison would not be person-relative, since it would hold true for experiences had by different people, as long as the corresponding experiences felt the same.

I wrote a comment on "Torture vs. Dust Specks" that helps make some sense of some of my intuitions here:

Maybe I'm not imagining the dust specks as being painful, whereas Eliezer had in mind more of a splinter that is slightly painful. Or we can imagine other annoying experiences like spilling your coffee or sitting on a cold toilet seat. Here again, I'm not sure if these experiences are even bad. They build character, and maybe they have a place even in paradise.

In a similar way as I feel like pleasures can't add up to outweigh a second in hell, so it similarly feels plausible that small pains can't either, not because I reject aggregation, but because small pains aren't even bad. Of course, if you imagine yourself signing up for 3^^^3 dust specks, that might fill you with despair, but in that case, your negative experience is more than a dust speck -- you're also imagining the drudgery of sitting through 3^^^3 of them. Just the dust specks by themselves may not be bad.

There are mundane pains that actually are bad, so this isn't a full threshold-NU stance.

I'm still toying with this position, and I'm not sure I accept it. It seems more akin to an aesthetic stance of the type that adults have when assessing a situation in the abstract. Maybe when you actually spill your coffee you feel differently.

I think many people share this intuition that small pains and pleasures just don't count -- if not in practice then at least in theory. Many people feel like life is about meaning and purpose and truth and love, and the amount of raw pleasure it contains doesn't really matter. Even Mill has something like this going on with his "higher pleasures," and the Greek eudaimons felt similarly.

Most people are not NUs not because they care about pleasure but because they're not fully hedonistic.

I personally don't care a lot about meaning and truth and stuff, but I do care that other people care.

Of course, the arbitrariness objection is more of a problem for moral realists like David Pearce, who argues from a non-arbitrary God's Eye View, and intuits a negative bias nevertheless. If you reject moral realism, you can just be arbitrary.

It's still a problem for moral anti-realists who care about having a coherent ( internally consistent ) world view.

small pains aren't even bad.

If pain is bad (or at least if you regard it as bad) how can you not regard a small amount of pain as mildly bad? At what point officially does pain become bad? I don't think that there is a negligible amount of pain or happiness. The tiniest amount counts just not by a lot. And I think that if it seems counter-intuitive that the largest conceivable amount of suffering could be compensated for by a greater amount of happiness or by reducing a high number of almost negligible amounts of pain felt at different moments or by different people it's because we don't have a vivid understanding of how much happiness or suffering would be felt or relieved so, not being able to imagine it, we can't have the same emotional response to it that we can to an amount of suffering that we can more easily imagine. A stronger, limitless imagination would kill scope insensitivity and all of our intuitions justifying negative utilitarianism. Imagination is more important to (hedonistic) utilitarianism than anything because we'll never have direct access to another person's private mental experience.

these are free parameters because "amount of happiness" and "amount of suffering" are not objective things.

I think that you can, in theory, objectively measure the total amount of happiness and suffering that exists in the universe at any specific moment. Two people in the same circumstances aren't necessarily affected to the same extent and if you could feel what they felt you could make an objective comparison. It's a fact that what most people feel when a loved one dies is greater than what most people feel when they get a speck of dust in the eye.

In my personal opinion, the option of really reliable painless suicide would solve the problem of involuntary suffering (modulo speculations about hell and the early childhood stage where voluntary suicide isn't realistic).

I agree. With some restrictions (like a mandatory 'waiting period' to really think the decision through, at least for those people who aren't terminally ill and in excruciating constant pain), I think the option of a painless, convenient suicide for adults would be hugely beneficial. It would give people a sense of control that might end up lowering the number of people who want to commit suicide to begin with.

I don't see it as important to bring heaven into existence from nothingness.

I consider being unconscious to be of neutral value but I still think it would be hugely desirable to create an eternal heaven that all sentient beings would experience constant, never ending and shockingly intense happiness in (something like what I imagine being high on MDMA to feel like mixed in with being in love or being an 8 year old on Christmas morning but 10 times that, something so beyond what any living human could possibly imagine).

Hi Ubuntu

Ubuntu wrote:It's still a problem for moral anti-realists who care about having a coherent ( internally consistent ) world view.

I doubt there is a non-arbitrary stance on this issue (or on many, many other ethical issues). It's not an issue of coherence. My view is coherent, just not non-arbitrary.

Ubuntu wrote:If pain is bad (or at least if you regard it as bad) how can you not regard a small amount of pain as mildly bad?

Maybe not all pain is bad. Some small pains might not be bad. This is more of a holistic, abstracted viewpoint, as opposed to, say, viewing each molecule of pain signal as inherently bad.

Ubuntu wrote:At what point officially does pain become bad?

I may pick some cutoff point between a dust speck and, say, depression.

Ubuntu wrote:It's a fact that what most people feel when a loved one dies is greater than what most people feel when they get a speck of dust in the eye.

The neural processes that fire in the former case are clearly distinguishable from and more intense than in the latter case. How we compare them is always an ethical judgment call. (I agree the loved one dying matters vastly more, but this is an opinion.)

Hi Ubuntu

Hello

I doubt there is a non-arbitrary stance on this issue (or on many, many other ethical issues). It's not an issue of coherence. My view is coherent, just not non-arbitrary.

Even if a preference for happiness over suffering, or non-existence over suffering, is arbitrary (I don't think that it is) I still think that your position is self-contradicting if you make a fundamental distinction between small pains and larger pains, different degrees of the same thing, or view happiness and suffering as asymmetrical in value.

Maybe not all pain is bad. Some small pains might not be bad. This is more of a holistic, abstracted viewpoint, as opposed to, say, viewing each molecule of pain signal as inherently bad.

The difference between a large amount of suffering and a small amount of suffering is a difference in quantity, not quality. If suffering is intrinsically bad then the property of badness doesn't emerge at some arbitrary point between mild pain and extreme pain, it's intrinsic to the smallest amount of pain. An increase in the amount of suffering a person feels doesn't change the nature of their experience, it changes the intensity of it. Again, even if you reject the idea that suffering has objective intrinsic dis-value, if you regard suffering as bad then I think it is incoherent to not regard mild pain as mildly bad. Mild pain has the same nature as extreme pain which is why we still call it 'pain'. I don't understand how your position is holistic or abstracted. I don't know what that's supposed to mean.

As for happiness and suffering being asymmetrical in value ; happiness and suffering are antithetical. Happiness is negative pain and pain is negative happiness. Good and bad are antithetical just like up and down or dry and wet. If you regard suffering as bad then I think it's consistent to regard happiness as good because they are opposite in nature. Strictly speaking the absence of pain isn't good, it's just not bad. Negative utilitarianism denies that there is a such thing as 'good' but how can you have a concept of 'bad' without one of 'good'? The idea that suffering is more bad than happiness is good seems as unintelligible to me as the idea that X degree of cold is colder then the equivalent degree of heat is hot. I think that a hypothetical person who has experienced very little pain in their life, who has an unusually strong disposition toward positive emotions and a weak disposition to negative ones, might intuitively feel that happiness is more good than suffering is bad because they don't have a vivid understanding of how intense suffering can be. I think it's a lot easier to cause someone a devastating, shocking amount of pain than it is to make them extremely happy, there might be evolutionary reasons for animals to feel pain more intensely than happiness which could explain why maximizing the happiness of well-off people seems supererogatory and trivial compared to alleviating pain.

Also, I think you mentioned somewhere that different painful emotions aren't commensurable? What all negative emotions have in common is being inherently averse and dislikeable and I think you can compare the intensity of boredom or humiliation with depression or frustration in the same way that you can make trade-offs between pain and happiness depending on how much of either is felt.

I may pick some cutoff point between a dust speck and, say, depression.

I would also make an arbitrary cutoff between some pain and a greater amount of pain when it comes to what I think of as mild pain versus extreme pain but the very nature of the thing (and my attitude toward it) doesn't change when you change the amount of it.

How we compare them is always an ethical judgment call. (I agree the loved one dying matters vastly more, but this is an opinion.)

It's not an opinion that the grief most people feel when they lose someone they love is more intense than the mild pain caused by a speck of dust in the eye. If you regard pain as bad then you would have to regard it as worse (not necessarily the event but the pain caused, maybe there are some hypothetical people who could be really devastated by specks of dust in their eyes).

Ubuntu wrote:Even if a preference for happiness over suffering, or non-existence over suffering, is arbitrary (I don't think that it is) I still think that your position is self-contradicting if you make a fundamental distinction between small pains and larger pains, different degrees of the same thing, or view happiness and suffering as asymmetrical in value.

One can value the whole more than the sum of its parts. The qualitative nature of a intense suffering can be judged different in kind from mild suffering.

Ubuntu wrote:if you regard suffering as bad then I think it is incoherent to not regard mild pain as mildly bad.

I may not regard all suffering as bad. (Toying with the position -- I don't have firm commitments here.)

Ubuntu wrote:Mild pain has the same nature as extreme pain which is why we still call it 'pain'.

There may be something qualitatively different about extreme pain. The brain is complex, and pain is more than just a negative number encoding. Even if that's not true, we could care nonlinearly about the amount of pain an organism feels at once.

Ubuntu wrote:Negative utilitarianism denies that there is a such thing as 'good' but how can you have a concept of 'bad' without one of 'good'?

Values can be negative or zero. That seems to make sense on its own.

Ubuntu wrote:which could explain why maximizing the happiness of well-off people seems supererogatory and trivial compared to alleviating pain.

Also the hedonic treadmill, which doesn't apply for certain forms of suffering (depression, anxiety, extreme deprivation, etc.).

Ubuntu wrote:Also, I think you mentioned somewhere that different painful emotions aren't commensurable? What all negative emotions have in common is being inherently averse and dislikeable and I think you can compare the intensity of boredom or humiliation with depression or frustration in the same way that you can make trade-offs between pain and happiness depending on how much of either is felt.

You're an astute reader. From "The Horror of Suffering":

In 2006, I discussed the problem of trading off suffering against other emotions with a friend. He said that in his experience, different emotions can be not just strong or weak but even "incomparable" with one another; certain emotional states can seem incompatible with memories of other emotional states. I replied that we're forced to compare them, and whatever tradeoffs we make in our decisions imply some exchange rates among emotions. While this is true, I still find a certain wisdom in the view that my friend expressed.

Emotions are like colors. How do you compare red against blue against green? Of course, we're forced to do so, and we can make rough tradeoffs, but there's a multifaceted texture to emotions that isn't fully captured when we collapse them to a scalar number.

One can value the whole more than the sum of its parts. The qualitative nature of a intense suffering can be judged different in kind from mild suffering.

The whole of something is the sum (whole) of it's parts. I don't understand what this has to do with your argument. Extreme pain has the same nature as mild pain (which is what makes both 'pain'). The difference is a difference in quantity, not quality.

I may not regard all suffering as bad. (Toying with the position -- I don't have firm commitments here.)

Do you mean based on something other than it's quantity? If you make an arbitrary distinction between some suffering and the same basic experience in another circumstance, aren't you abandoning hedonism?

There may be something qualitatively different about extreme pain.

'Extreme' and 'mild' are relative differences in quantity. A raindrop isn't something fundamentally different from the Pacific Ocean, the water in the Pacific Ocean isn't more wet there's just more of it. I don't understand how an increase in the amount of something can change it's inherent nature or quality.

we could care nonlinearly about the amount of pain an organism feels at once.

We can care about anything but if we care about the suffering of any one person felt at any moment in time then it's consistent to care about all suffering (because suffering is suffering and we have to refer to some arbitrary thing other than suffering to justify making distinctions between the same basic experience felt by different persons 'or' at different moments or in different circumstances) and only suffering (because happiness-suffering is not commensurable with other values).

Regardless of whether or not water has intrinsic value, it's self-contradicting to regard water as valuable between 10 am and 10 pm on Wednesdays and Thursdays and as having neutral or negative value the rest of time, and I don't mean because you changed your mind but because you regard the water that exists on Wednesday afternoons as more valuable than the water that exists on Friday evenings even though it's the same stuff.

Values can be negative or zero.

Yes but the opposite of negative is positive and the opposite of suffering is happiness so how can you not regard happiness as good - the opposite of bad- if you regard it's opposite - suffering- as bad?

Emotions are like colors. How do you compare red against blue against green? Of course, we're forced to do so, and we can make rough tradeoffs, but there's a multifaceted texture to emotions that isn't fully captured when we collapse them to a scalar number.

All negative emotions are commensurable in that they're all intrinsically dis-likable and aversive. They're all experienced as having negative value and being undesirable. The basic aversiveness of any emotional state can be compared to any other. It might be hard or impossible to do in practice but if you accept that all negative emotional states can be quantitatively measured in terms of intensity and duration and that they share the same property of aversiveness then how can you not, in theory, make precise trade-offs from an impartial/objective standpoint? I think our intuitions can be misleading.

I don't know if you've ever heard of Steve Roach but I just finished listening to Reflections In Suspension. I'd recommend it.

Ubuntu wrote:Extreme pain has the same nature as mild pain (which is what makes both 'pain').

There are similarities, yes, but they're not necessarily just different in some quantitative dimension. There may be qualitatively different brain processes for different intensities (and even different forms of suffering at the same intensity). Collapsing degree of suffering to a single number is something we impose; the brain itself is doing lots of complicated stuff.

Ubuntu wrote:'Extreme' and 'mild' are relative differences in quantity. A raindrop isn't something fundamentally different from the Pacific Ocean, the water in the Pacific Ocean isn't more wet there's just more of it. I don't understand how an increase in the amount of something can change it's inherent nature or quality.

Simple example: Extreme pain can trigger PTSD. Mild pain doesn't. That's a qualitative shift in how the brain responds to the experience.

The Pacific Ocean is more salty than a raindrop. The water beneath the surface may have less dissolved oxygen and more pressure on it. It contains phytoplankton and zooplankton. And so on. If you're just talking about the H2O molecule, I agree that's the same. But suffering is vastly more complex and multifarious than H2O.

Ubuntu wrote:Yes but the opposite of negative is positive and the opposite of suffering is happiness so how can you not regard happiness as good - the opposite of bad- if you regard it's opposite - suffering- as bad?

The opposite of suffering is happiness under the classical-utilitarian axiology. Under a different axiology there is no opposite of suffering. Not everything has an opposite.

Ubuntu wrote:if you accept that all negative emotional states can be quantitatively measured in terms of intensity and duration and that they share the same property of aversiveness then how can you not, in theory, make precise trade-offs from an impartial/objective standpoint?

There may be many textures of aversiveness: aversiveness that makes you say "that sucks," aversiveness that makes you cry, aversiveness that makes you clench your chair, aversiveness that keeps you awake at night, aversiveness that haunts you with terrible memories, aversiveness that makes you want to die, etc. There are may dimensions to experience that we may regard as morally relevant.

Our brains appear to have many different reinforcement/motivational systems: reflexes, model-free learning, model-based learning, planning, etc. Even if we just used the value assessments of these systems for our moral calculus, which system would we be using? Or perhaps we'd use the joint output, but this isn't the only choice. Also, the joint output in terms of motivation / action selection needn't coincide with hedonic optimality.

Ubuntu wrote:I don't know if you've ever heard of Steve Roach but I just finished listening to Reflections In Suspension. I'd recommend it.

Nope. Thanks.

There are similarities, yes, but they're not necessarily just different in some quantitative dimension. There may be qualitatively different brain processes for different intensities (and even different forms of suffering at the same intensity). Collapsing degree of suffering to a single number is something we impose; the brain itself is doing lots of complicated stuff.

Simple example: Extreme pain can trigger PTSD. Mild pain doesn't. That's a qualitative shift in how the brain responds to the experience.

There aren't just similarities, mild pain is the same experience as extreme pain hence both falling under the category of 'pain'. What's different is the 'extreme' vs. the 'mild'. The subjective experience of pain isn't the neurological activity it corresponds with. PTSD is just extreme, prolonged distress.

If you're just talking about the H2O molecule

I was.

But suffering is vastly more complex and multifarious than H2O.

How so? I don't see how the two are comparable and I would also intuitively think that an extreme, overwhelming amount of suffering felt at a single moment by a single person is worse than a very small amount felt over several different moments or by different people but I don't think my intuitions are reliable.

No offense but I think people often misunderstand their position as being complicated and 'nuanced' when it's fundamentally self-contradicting.

The opposite of suffering is happiness under the classical-utilitarian axiology. Under a different axiology there is no opposite of suffering. Not everything has an opposite.

No, the opposite of suffering is happiness. Everyone understands suffering and pain to be antithetical to happiness and pleasure. That's not a value judgment. Bad does have an opposite and it's good. The absence of suffering can't be intrinsically good because it isn't anything.

There may be many textures of aversiveness: aversiveness that makes you say "that sucks," aversiveness that makes you cry, aversiveness that makes you clench your chair, aversiveness that keeps you awake at night, aversiveness that haunts you with terrible memories, aversiveness that makes you want to die, etc. There are may dimensions to experience that we may regard as morally relevant.

I don't disagree but all aversive states are basically the same in feeling undesirable.

There are may dimensions to experience that we may regard as morally relevant.

From the hedonistic point of view, only the basic felt aversiveness or desirableness of an emotional state has direct moral relevance.

which system would we be using

The one that is concerned with maximizing happiness and minimizing suffering in the long run and generally or most effective at doing so (the model based or a mixed 2 level utilitarian approach?).

the joint output in terms of motivation / action selection needn't coincide with hedonic optimality

What foolproof approach would?

Hi Ubuntu

Apologies for my delay. I finally read your latest message. My guess is we both understand each other's viewpoints, so I won't press the argument further, since it's likely to be repetitive. I agree that your view corresponds to a certain brand of hedonistic utilitarianism (which you might call the uniquely correct hedonistic utilitarianism). My view corresponds to some other consequentialism. I would call it also a form of hedonistic utilitarianism, but you wouldn't, and this is fine because it's just a linguistic debate.

Thanks for the discussion.