Many people argue that suffering predominates in nature. A really simple form of the argument, supported by people like Brian Tomasik, is what one might call the argument from life history. In general, in most species, females produce many more offspring than can survive to adulthood; in some cases, a female may produce thousands or millions of offspring in a single reproductive season. Therefore, one can assume that most animals die before they are able to reproduce. In many cases, the offspring die before they can reasonably be considered conscious (for instance, an egg is eaten shortly after laying). However, even if half of animals die unconscious, the other half are a large source of disutility. Since death is generally quite painful, they may not have had enough positive experiences to outweigh the extraordinarily negative experience of death. It can therefore be assumed that there is more suffering than happiness in nature.
While this argument is intuitively compelling, I am not sure that it accurately reflects most people’s opinions about how happiness works, so I have decided to write up three thought experiments that might help people think about it. These thought experiments are quite preliminary; I hope to spark a discussion so that people who are concerned about wild-animal suffering can debate.
- The Human History Thought Experiment
Although the human population has been growing for thousands of years, for most of history the growth was fairly slow, suggesting the argument from life history applies to us as well. In part, that was because many humans died in childhood: for example, in 1800 four-tenths of humans died before they were five years old, a quarter of humans before their first birthday. (Note that 1800 is fairly late, and the statistics may have been even more stark in, say, 1 CE.)
I do not mean to deny that pre-modern human life was miserable in many ways: people were hungry, diseased, and poor. And I certainly don’t mean to claim that high child mortality rates weren’t a tragedy. However, my intuition is that– whether or not human lives were worth living before modernity– the high child mortality rate does not single-handedly prove that human lives were not worth living. Other information must be gathered to prove that. I suspect many other people’s intuitions will agree.
To the extent the human history argument is misleading or anthropomorphizing, it strengthens my point: for instance, humans grieve their infants, while fish do not, so high infant mortality rates are worse for humans than for fish.
2. The Babykillers Thought Experiment
Humans have relatively few children and are growing (albeit slowly), perhaps suggesting that they are misleading as a thought experiment. Consider, therefore, a sapient species of aliens, the Babykillers. This species lays a thousand live young at a time. The young devour each other and only the single strongest offspring survives. All of the non-surviving thousand have miserable lives: they are tremendously hungry until they are eaten alive. The Babykillers have no way of modifying themselves to lay only a single offspring. To set aside issues of the Babykillers being replaced by humans, assume that Babykillers are not aware of any other species. Would it be ethically required to have a Voluntary Babykiller Extinction Movement?
Personally, I put some weight on the argument that diversity is intrinsically valuable and therefore it is harmful to eliminate a sapient species. Otherwise, my moral intuitions are conflicted about whether Babykillers are net-negative and should be extinct.
3. The Long-Lived Babykillers Thought Experiment
The Babykillers are a species as specified above, except that the Babykillers are also extraordinarily long-lived: the average Babykiller who survives infancy lives for a thousand years. The average Babykiller who is eaten lives for only an hour. Babykillers are also a happy and fulfilled species. Therefore, while 999/1000 Babykillers experience a short life of great hunger followed by a painful death, only about one in ten thousand hours experienced by a Babykiller consists of great hunger and a painful death. The rest are quite happy.
The Long-Lived Babykiller thought experiment is supposed to get at an alternate method of assessing well-being. Instead of thinking about whether the average member of a species has a life worth living, we instead think about whether the average hour experienced by a member of a species is worth experiencing. For species with high juvenile mortality and/or long lives, these may be very different metrics.
Intuitively, I think the Long-Lived Babykillers should not go extinct. I think that also goes along with my intuitions about the human case. Since humans are fairly long-lived, a relatively small percentage of human hours are spent being a sick infant. I’m also tempted by the practical benefits. You could, in theory, figure out whether an animal species’s existence is net positive simply by randomly sampling animals (although of course seasonal changes such as winter starvation or mating seasons would make this more complicated). However, many people have prioritarian intuitions. In that case, the experiences of animals who die young and painfully should be given more weight than the experiences of happy animals.
To be clear, I’m not necessarily satisfied by the “average hour” criterion. I do think we haven’t put enough philosophical work into understanding what makes a species’s existence net positive or net negative, and I hope my thought experiments will prompt some thought about the issue.
wireheadwannabe said:
Isn’t “average hour” more or less just total hedonic utilitarianism? Or am I missing something?
Anyway, I think the average hour criterion suggests that most animal life isn’t worth living, because as the original argument states there are a lot of hours spent in the process of dying relative to the number of hour spent living. And that’s not even taking into account the various sources of unpleasantness that come with being alive. Even without taking infant mortality into account, I would expect the average hour spent alive to be net negative.
LikeLike
Aapje said:
That’s what it states, but it’s a highly subjective and debatable claim that I don’t really buy. Even if a kid dies at age 5, aren’t the many happy and unhappy moments before that more significant than the negatives of dying, which generally is just an instant compared to the period before?
Even the implicit assumption that the negative utility of dying has some fixed value seems wrong, as various ways of dying seem to differ substantially in time and suffering. Furthermore, for much of history, opiates were used to relieve pain. For example, the Greek god for sleep (Hypnos) has the poppy as his symbol and is the brother of death. Pliny the Elder also wrote about narcotics for medicinal use.
LikeLiked by 1 person
sniffnoy said:
See as someone who doesn’t endorse utilitarianism (or, I guess, aggregative consequentialism more generally — thanks for indirectly teaching me that term via that link; for all that I go around complaining about how people use “utilitarianism” to mean “consequentialism”, it turns out that I’d been using it to mean “aggregative consequentialism” 😛 ) this all just reads really weird. Like — exterminating a civilization just because most of its members live terrible lives? That can’t be right, can it? Well, maybe it can; that’s a relevant factor, certainly, it has some weight in the calculation. But this just reads really weird because it makes no mention of what’s wrong with such a thing other than that yeah also you’re killing people who live good lives. Like… there’s no mention of civilizational accomplishment, of greatness, of monuments, you know? And to my mind this really counts for something, it really is a relevant factor, it’s not something that can just be ignored. I know a lot of people will say “well monuments are only good if somebody is around to see them”. (Note: By “monuments” here I don’t necessarily mean, like, physical monuments, although those certainly count… just like, you know, doing things that are hard and awesome for the sake of it. Like going to the moon, for instance.) And I just fundamentally disagree with that, I guess? Obviously it is better if there’s actually people around to see it, but it’s hard for me to say it counts for nothing otherwise.
It’s like, welll… think about alien archaeologists. Now, I don’t often talk about alien archaeologists… but it seems that when I do, people always know exactly what I mean. Imagine we’re all dead, humanity is extinct, no living descendants or AIs or antyhing, and aliens finally land on our planet and discover our remains. I want them to say “Wow, these humans were truly amazing!”, you know? Now obviously ideally we and our descendants don’t go extinct and our civilization continues just fine until the heat death of the universe or whatever. But that’s still a lot of how I think about things.
And yeah maybe I just talk like this because I think things can be improved, that humanity doesn’t need to suffer to create things that are great, and maybe if the choice were really “great monuments and great suffering, or nothing at all?” I might be more inclined to consider extinction. But I can’t imagine I’d just consider monuments to not even be a factor, to not even have to think about the question, which is, I guess, what an aggregative consequentialist would do.
LikeLiked by 1 person
vamair said:
Not All Aggregative Consequentialists.
I’m very much for “preference fulfillment”. And preferences are about reality, not necessarily about feelings or emotions. That means you don’t really need to be able to interact with stuff to have preferences about it. The statement “I’d like our civilization to leave a mark in history, even if there’s no one to see it” is a preference. A preference I like and do have. As well as a preference for diversity as opposed to finding an optimal configuration and filling universe with its copies.
It’s as much okay to have preferences about what happens after your death or after humanity’s extinction as it is to have preferences about what happens in a room you’re not in.
LikeLiked by 2 people
sniffnoy said:
Oh, oops, you’re right. Got mixed up there myself looks like.
LikeLike
sniffnoy said:
Actually: Tangent. Why on earth does prioritarianism get its own word distinct from utilitarianism? This is silly. It seems to me that the difference between a form of prioritarianism and a corresponding form of utilitarianism is tiny compared to, say, the difference between hedonic vs preference utilitarianism. Like, both prioritarianism and utilitarianism are making the assumption of aggregation, which, let’s be clear, is a big assumption. And then you have the two questions of what you’re aggregating and how. Changing what you’re aggregating seems to change the entire character of the system. But the aggregation method? Average and total utilitarianism are both considered utilitarianism. But if instead of an average or a total we take, say, a minimum, then boom, that’s suddenly prioritarian (and pretty extremely prioritarian).
Here’s an even better demonstration. Say instead of going all the way and taking a minimum, we just take, say, a sum of square roots (assume everything is positive, or adjust the example appropriately if for some reason you find that assumption objectionable 😛 ). Now once again our utilitarianism has become prioritarian with only a tiny change. Except: Taking a sum of square roots is the same as just taking the square root of each one, and then taking a total. But is the square root of a utility function meaningfully different from the original utility function? Well, yes, very much so, if we’re talking about decision-theoretic utility functions. But if we’re talking about E-utility functions, the sort that get aggregated? Who knows! The whole area is so ill-defined that there’s not a clear answer here. Our prioritarian aggregation rule might just be a utilitarian aggregation rule. It’s all a mess. Trying to draw a distinction here that assumes that these are two seriously different things, rather than small variants on the same thing, seems very wrong.
So this seems like a bad system of terminology, if one can’t say “minimum utilitarianism is a form of utilitarianism”. To hell with saying “aggregative consequentialism”, I’m just going to say “utilitarianism” and implicitly include prioritarianism as a minor variant (especially since then more people will understand what I’m talking about 😛 ).
LikeLiked by 1 person
sniffnoy said:
Er, sorry, I formulated that argument poorly. It doesn’t quite make sense as written. Let me try that again.
So the answer to my question “Is sqrt(u) meaningfully different from u?” is clearly yes if we assume an aggregation rule beforehand. Like I said that part didn’t quite make sense. But here’s the thing — let’s take the square-root prioritarian system I described above. Then this can equally be well described as a utilitarian system just by applying a square root to each utility function beforehand. And because the whole area is so ill-defined, there’s nothing to say which is the “correct” definition of a utility function. So both descriptions truly do work equally well; there’s absolutely no way to say whether the system is “truly” prioritarian or utilitarian.
Prioritarianism isn’t just a minor variant on utilitarianism; it straight-up overlaps with it. By contrast preference utilitarianism doesn’t really overlap with hedonic utilitarianism — not unless you seriously twist things. But because preferences and pleasure-vs-pain are concepts we are already familiar with, we can in fact judge when you are twisting the concept — unlike E-utility functions, where we have no way to distinguish.
So yeah. I see no reason why I should distinguish prioritarianism from utilitarianism when it’s possible to (assuming we accept that E-utility and aggregation makes any sense, which is a big assumption but if you’re any sort of aggregative consequentialist you already accept it) write down an ethical system that can equally well be described as either.
LikeLike
Murphy said:
From the names chosen and the almost certainty that the author is familiar with the work I think this is a reference to Baby Eating Aliens:http://lesswrong.com/lw/y4/three_worlds_collide_08/
Could you think of a hypothetical species which does big impressive awesome things but which is too terrible to exist under your moral system? perhaps one who’s biology rests on the endless pain and suffering of large numbers of members of their own species. Perhaps something like a Surinam Toad which needs to regularly cannibalize members of it’s brood to survive or something. (and since they evolved as such they see it as morally neutral)
The exact mechanics aren’t really the point.
LikeLike
ozymandias said:
Yep! They were Babyeaters in the first draft but I felt like that would be unnecessarily confusing, since Babyeaters are already used as a thought experiment to talk about the orthogonality thesis.
LikeLike
Brian Tomasik said:
Thanks for the post. 🙂 IMO, the more relevant metric than infant mortality is life expectancy at birth, because this more precisely measures the average number of hours of life per death. (This is basically your “average hour” metric.) Of course, life expectancy at birth is usually pretty (inversely) correlated with infant-mortality rates. But, e.g., a hypothetical insect species with no infant mortality that lived only 4 weeks would still be pretty badly off from this perspective.
Some people who talk about disvalue in nature are egalitarians, so the fraction of nonsurviving offspring may play a role for them that it doesn’t for a classical utilitarian. (And similar for prioritarians, as you say.)
Here’s a less extreme version of babykilling:
https://www.smithsonianmag.com/science-nature/baby-sand-tiger-sharks-devour-their-siblings-while-still-in-the-womb-46192985/
“This species practices a form of sibling-killing called intrauterine cannibalization. Yes, ‘intrauterine’ refers to embryos in the uterus. Sand tiger sharks eat their brothers and sisters while still in the womb. […] From what began as two uteri full of a dozen embryos results in just two dominating baby sand tiger sharks coming full term.”
LikeLiked by 1 person
Pingback: Rational Feed – deluks917
sable said:
I feel like the argument is a dead end. Whether or not there’s more suffering than happiness in nature, cutting down that suffering is good and destroying (most) animal species is awful and impractical.
LikeLike
Windlestre said:
“destroying (most) animal species is awful and impractical.”
Awful – in the eyes of the beholder.
Impractical – not true, as we have an economic incentive to do so anyway. We have little economic use for most animal species and when we replace fossil fuels with renewables, we’re going to want to use those photons that now go into wildlife.
LikeLike
Windlestre said:
Morality is relative.
That said, I would not torture one baby in order to give happiness or meaning to others. Even less for mere diversity. No matter the numbers.
We are biologically and culturally programmed to tolerate absurd injustice and evil for life or society. A pessimistic stance that violates this programming is prone to be outside the Overton Window no matter the arguments or object-level substance.
There is also an intensity asymmetry between the worst suffering and the rest of life. You‘d need Science-fiction tech in order to have more pleasure than pain in the system. History was indeed mostly a painful mess.
LikeLike