The “repugnant conclusion” in question is that a few trillion lives just worth living might be better than 10 billion brilliantly wonderful lives.
This was raised in Sam’s recent podcast “Being Good and Doing Good - A Conversation with William MacAskill”, attributed to Derek Parfit (also known as the “Mere addition paradox” https://en.wikipedia.org/wiki/Mere_addition_paradox).
There seems to be an odd unquestioned assumption underlying the paradox, when talking about either “total” or “average” happiness (or wellbeing - I’ll just use ‘happiness’ from here on in, though), which is that happiness is summable across a population.
Now, even if we were to grant that levels of happiness in an individual are comparable (which seems intuitively true), it doesn’t necessarily follow that that individual’s happiness is quantifiable (in the sense of there existing some quantitative amount - leaving aside the difficulty of measurement, which I’ll just ignore for the rest of this discussion). Such comparisons may not even be transitive, which would make ‘quantifiability’ impossible. For instance, we know that in the real world, an individual’s preferences may not be transitive, e.g. they may prefer oranges to apples, apples to pears, and pears to oranges, when presented with binary choices - there may be similar problems when comparing happiness (especially given that preferences may affect happiness…). If it turns out that happiness is quantifiable, it doesn’t necessarily follow that that quantity is proportional or summable in the case of a particular individual (and certainly there’s evidence that our subjective experience of happiness is not linearly proportional to the positive or negative stimuli in our lives - at the very least, any meaningful quantification would have to be or include a measure of subjective experience). If it turns out that you can come up with some way to define the quantity of happiness for an individual linear and summable for one individual in different scenarios, it still doesn’t follow that happiness is summable over multiple individuals.
The argument is made in the podcast that a happy person existing vs. not existing (all else being equal) isn’t neutral. The reasoning goes that if the ‘total good’ of “person A exists at happiness 6” (on a 0-10 scale, 5 being neutral) is equal to “person A doesn’t exist”, and the ‘total good’ of “person A exists at happiness 8” is greater than “person A exists at happiness 6” (obviously true), therefore the ‘total good’ of “person A exists at happiness 8” is greater than “person A doesn’t exist”... so the good of their existence at happiness 8 isn’t neutral - and by implication even their existence at happiness 6 isn’t neutral. Again, this relies on the property of transitivity existing between comparisons of states for an existing person and comparisons between existence and non-existence (i.e. that X > Y and Y > Z implies X > Z, or in this case X > Y and Y is equivalent to Z implies X > Z), which hasn’t been shown, and even if we do grant transitivity for these comparisons (which I’m inclined to do - and intuitively I tend to agree with the conclusion here), that still doesn’t necessarily imply quantifiability or summability in any form.
Discussions of these arguments seem to start at the comparability of happiness, completely ignore the properties of transitivity and quantifiability / linearity, and jump directly to summability across a population. Each of these is properties is necessary but not sufficient condition for having the next property, and each has to be shown to hold for a single individual in ‘alternate universes’, between different individuals, and between an existent and non-existent individual, for summability across a population to be valid - and so far I’ve not seen a convincing proof that happiness or wellbeing has any of these properties, beyond the “intuitively true” assertion of comparability when looking at a single individual across alternate universes, and the “intuitively true” but in many cases problematic assertion of transitivity. I can’t even recall seeing anyone seriously tackling the question - as far as I can see, this just seems like a wild leap into the unknown without any solid foundation, which people are then deriving conclusions from (such as ‘the repugnant conclusion’ mentioned).
So has anyone come across a decent treatment of these properties in this context?
[edit: I just noticed that the problem of transitivity is mentioned on the wikipedia page - but as mentioned, even granting transitivity within and between all cases, we’d still be several steps away from whole-population summability, as far as I can see]
I quite like musak and potatoes.
Hi there Madacre and welcome to the SH forum,
this is a standard critique of utilitarianism, and it is somewhat related to the Utility Monster thought experiment.
But all that these things do is start with approximations of happiness and then pretend that you can arbitrarily extend, add and subtract them. We are talking models here, not accurate representations of a society: if you have an average life-expectancy of 80 years in a group, no one assumes that everyone will die before their 81st birthday but after their 80th.
Sam Harris’s model of the Moral Landscape somewhat circumvents this problem in that he doesn’t have to deal with absolute values of happiness, only in relative changes: will having more people make us more happy? Will less? Does it matter?
These are the questions we could test and a well-described moral landscape. But we would not be able to tell (as in the example) A from B, since there might not be a connecting pathway on the landscape between those two points.
For the Record, Parfit argues that comparing things like levels of happiness between others, or theoretical future selves, is inexact, but still comparable in important ways like health, well-being, avoidable suffering and so forth. My personal take on the repugnant conclusion it shows overpopulation is a bad thing, that more lives do not always equal more good.
In a world where the human race is in harmony with nature and on track for deep-time evolution, our individual lives and maximum individual well-being can be additive towards an approximate maximised human potential? Since we exist in a universe with limited options, if adding more lives leads to unsustainable resource consumption and the risk of future lives or the human race, we start to defeat the point of it all?
Thanks - I have seen a few critiques of “pretend you can arbitrarily extend, add and subtract” part along these lines - it is a common critique, but it’s easy to miss how serious a critique it is.
As you point out, there are problems in any model where a simplified measurement parameter may be misleading (particularly, as you point out, when you don’t have a clear idea of what a model parameter like “average” actually means), or where a model doesn’t fit the data very well, but this is an error (if it is an error) of a different and more serious category. It’s less like using a poor proxy variable, mistaking an average for a uniform value, or using standard deviation on a non-normally-distributed data set, and more akin to trying to multiply an apple by the anticipation of the colour purple - we may even have a quantifiable amount of apple and a quantifiable amount of anticipation for purple, but that still doesn’t mean it makes any sense, or can be treated as a valid operation to perform.
So what I’m really wondering is if anyone has come up with a decent defence of the “arbitrarily extend, add and subtract” presumption, or even for some of its necessary preconditions? Or is it just presumed for the sake of the thought experiments, and never defended?
As you allude to from Sam’s work, we don’t need all this to make comparisons. I should note that we can actually get a surprising amount of mileage even from a (fairly well connected) partial ordering, without summability, quantification or even strict transitivity (so long as sufficiently “large” differences are still transitive) - for measurement within a population, we could conceivably approximate medians or Nth-percentile measures by selecting instances at the appropriate distance along an approximated total ordering, and then compare those instances across scenarios. Questions like “If adding one very happy person makes everyone else a tiny bit less happy, is that good or bad?” are still hard to answer, though.
I just finished listening to this wonderful podcast episode. I’m not a philosopher in any trained sense, so maybe my reactions will expose some ignorance that would otherwise be dispensed with in a first year course, but here I go anyway. The section on population ethics and repugnant conclusion seemed to be missing a connection to reality. I know, as thought experiments we are trying to arrive at some greater principle that could be applied to any situation, but maybe that just falls apart in the real world and can’t be done. So, in my opinion, the rather simple answer to the population ethics question relies on maximizing well being given the carrying capacity of the earth to continue and thrive. This should also take into consideration biodiversity, and the well being of our fellow beings. This is a far more complicated equation than any thought experiment, for example a mosquito’s right to exist is far less than a chimpanzee’s, given the harm done by the former. Adjudicating such things may be difficult, given unintended consequences if we try to over-manage the situation. Nevertheless I can envision a world with some few billion humans maximally optimized for balance in the ecosystem on an earth largely turned back into a giant wilderness park for our enjoyment where biodiversity is restored and protected. Over time the optimal stable population of humans is reached, all of whom are leisure-class or pursuing knowledge and art at their pleasure. If a permanent situation were knowable, for example 2 billion humans is sustainable indefinitely but 2.1 billion, (say, according to models and predictions of our best AI of the day) will end badly, then we have our answer. The maximum number of future generations of happy people will thusly be achieved.
The discussion is entirely humancentric. Anthropocentrism throughout. Population ethics entirely ignoring the other populations on the planet, and the Earth itself. MacAskill’s rather weak and tedious argument revolves around happy Harry in the “world”. What world? Whose world? How can one have a serious discussion about population ethics and its effects on the world when the world and its other 70 billion tons of biomass is totally ignored by virtue of an argument ensconced in speciesism? Neither participant ever touches on the assumption of humans as sacrosanct. It’s all about human happiness and ignores the simple and glaring fact the the “world” would be exponentially better without humans. Very disappointing discussion. The discussion was about preserving happy humans. I thought Schopenhauer, and later, Zapffe, warned against this folly?
I enjoyed the podcast with MacAskill on the one hand, but on the other was shocked at how superficial the thinking was about the objects of altruism. And incidentally: most of MacAskill’s wealth is in his human capital. The youngest tenured professor in the history of Oxford has absolute job security. And suppose some day he decides he needs an extra $10K? He could probably earn it giving a speech or two. Most of us can only dream of the financial security he has by virtue of his tenure, fame, and intellectual gifts.
But the bigger issues are: who do you give your money to? MacAskill claims to know to the penny how much money will save a life. The discussion implies it’s easy—pay for food for a starving child. But wait, maybe subsidizing the distribution of condoms or other birth control would lead to fewer suffering children and be a better outcome? But wait, the main reason abject poverty exists is because of political pathologies, so maybe the way to mitigate suffering is to create foundations that promote the right political institutions? (Some people still think it’s capitalism that creates the suffering masses, and the way to alleviate it is to create communism. So, they might say, fund revolutions in impoverished third-world countries.)
Or perhaps more lives can be saved by financing someone’s Ph.D. in agronomy, so that more and better food can be grown to feed more people? But universities couldn’t exist if we all follow the simple MacMi model. Paying tuition would be unethical because the tuition money could save lives. (I guess MacMillan completed his education before adopting his ethical rules.)
Maybe the creation of tax deductions for Monsanto to invest in more and better BMO crops would save the most lives—the right seeds might grow more food with less water, or be more resistant to pests or disease and save millions of lives.
What causes poverty in the first place, and what if we focus on fixing those causes? No, it wouldn’t be easy, but perhaps focusing on finding and promoting the right answers would lead to better outcomes ultimately.
In short, what MacAskill presents as a simple ethical decision –“give most of your money away”—doesn’t get you anywhere. His certitude about what it takes to “save a life” is impressive but call me skeptical. What charities actions will do the good? What actions lead to the best outcome? MacMillan’s smarter, more targeted effective altruism seems to have skirted all the most important questions.
The concept of a “repugnant conclusion,” which Harris addresses in The Moral Landscape, contradicts his claim that science can determine human values, doesn’t it? Is “repugnance” something science can determine? Or is it something we humans subjectively perceive?
If science determines that the aggregate Well-Being of Conscious Creatures is maximized by maximizing total well-being instead of average well-being, but we discard that determination because we find it “repugnant,” can we really claim that our values are scientifically determined? I don’t think so.
In fact, the issue is moot because the question of aggregating WBCC is one that science cannot answer. Nor can it answer the question of whether future well-being counts less than, equal to or more than present well-being. (If future well-being counts more than present well-being, then contributing to Global Warming is immoral—as is running up the national debt.) And if science can’t determine how to aggregate WBCC, it can’t determine human values. Period.
If the concept of a “repugnant conclusion” shows anything at all, it shows that Harris isn’t using “science” to determine human values, he’s trying to use “science” to rationalize his own preexisting, subjective moral beliefs. It’s a classic case of Shermer’s “belief-dependent realism.”
Sam expressed that he particularly liked his latest podcast “Being Good and Doing Good.” I thought it was one of his worse (ignoring his spats earlier this year) but certainly respect Sam’s effort to do good and his judgement… which is normally spot on. For me, philanthropy is a futile approach to the planet’s biggest issues, and the thought experiments offered, such as the burning house painting vs children decision, are frankly useless…
Sam ended the podcast questioning how one could challenge the paradigm of giving up everything so as to better the quality of life of less fortunate humans. To paraphrase the solution offered by his guest, humans should seek to spread evenly all that enriches our lives, while the human species, and many other forms of life on Earth, barrels uncontrollably towards collapse. Our planet’s environment is on a crash course with devastation and extinction. Over consumption of resources, over destruction of habitat, over pollution of air, land & water. And for the most part, we do this while three delusions persist: 1) that our economic system is an appropriate gauge of progress, 2) that an all powerful superior being is looking after us, and 3) that as individuals we are somehow ethically responsible to procreate excessively and then to be concerned for the well-being of that fail state that ensues. What liberal garbage.
So Sam, here is the SOLUTION you seek, however difficult it may be to achieve while humans cling to their 3 delusions. Regardless of ongoing technological advancement, humans act, and will always act, like a parasite to the planet Earth. It is inherent in human nature to be selfish and parasites are only non-destructive when their populations are overwhelmingly in check. What we must do is maintain the planet in a healthy equilibrium where the environment is pristine and human enrichment is maximized. This can and will only be achievable when the available planetary wealth so greatly exceeds human needs, that human greed becomes a moot point. Just imagine the human population and environment of 8000 years ago, with the technology and moral advancements of today’s mankind? Now that would be goal worth achieving.
Now I know most will rebuke this as impossible at best, and lunatic ramblings at worst. But, such a state for the planet is both possible and achievable. it can be substantially achieved by driving our population down to a small fraction of what it is today. No more overfishing of oceans, no more over-polluting of the environment, no more habitation in uninhabitable geographies, no more humanity-forced extinctions, no more over-consumption of natural resources, and no more need for ongoing human conflict. Such a world was not possible in our past because we had not yet learned how to master our environment. And when be began to learn how, we made the mistake of allowing greed and religion to drive our populations to the level that our technology could never get out ahead of.
I challenge anyone to present a substantive argument that refudiates the premise that a 500 million human population sharing the world’s resources in environmental equilibrium using and advancing the technologies of today would not represent something exponentially better than anything humankind has ever seen. And imagine the benefit to other life forms! Yet instead we get to watch Peoples with primitive and ignorant religious ideologies/cultures populate like rabbits, and Corporations pollute and destroy our environment for short-term economic gain. And the best that Oxford philosophers can come up with is to advocate for individuals not to buy a new stereo, enjoy some seafood or play a round of golf this weekend because somewhere in the world such expense could be put to better use saving impoverished peoples whose principal role on Earth is to breed until their environment & population collapses?
Mankind’s objective should be to strive for greatness and enrichment, not strive for over-populated mediocracy and strife. The Repugnant Conclusion is NOT A PARADOX. It is a complete fallacy to believe that it is humankind’s objective to forever increase populations at the expense of anything/everything we value. And what is the Catholic Church doing these days? Indoctrinating Africa so that a People can lock in for the future their already impoverished existence. If you don’t like me picking on Africans, then you are welcome to choose anywhere else in the world where population growth is outstripping resources at the expense of a people’s average quality of life.
A note to Sam: on what ethical basis can you argue that 2X human population at even an infinitesimal lower quality of life is better than X population at a theoretically maximized quality of life? We either breed like rabbits and have the environment cull us, or we control our population and maximize our existence. Please explain to me why a population of 10 billion ( a target we are hell bent to reach within the 21st century ) is desirable over a population of 500 million? I can’t think of one single reason why the human population needs to be so large? Socially, Economically, or Scientifically, I don’t see the value in excessively large populations. Evolution and Religion may both have driven humans to breed without control for much of our past, but such needs no longer exists. Humans need to shake off these primitive chains or else we shall drown trying to swim with them attached.
Please explain to me why a population of 10 billion ( a target we are hell bent to reach within the 21st century ) is desirable over a population of 500 million? I can’t think of one single reason why the human population needs to be so large? Socially, Economically, or Scientifically, I don’t see the value in excessively large populations.
There are going to be an order of magnitude more geniuses in the former case. And there are advantages to those people interacting with each other vs. appearing hundreds of years apart. If you look at charts of population growth vs. scientific advancement they track pretty closely. In the end, the enormity of human population doesn’t matter as long as using up the Earth isn’t a constraint - that is, humans can figure out how to live off-planet. Which frankly doesn’t seem like a hard thing to do in the next 10,000 years, a truly trivial amount of time even in human history.
Even if you value animals other than humans, one terraformed moon or planet shared with non-human life is going to counteract any amount of present-day on-earth species extinction in the long run.
another reason is genetic diversity.
amusl thanks for your thoughts. Obviously this topic is not easily quantifiable, but let me make a few responses. First, it is not geniuses but rather medium to high IQ individuals with the right education and resources that help mankind make advances. Thus I don’t buy into needing 10 billion people of which most are poorly educated in order to make the advancements needed. Quality not Quantity will suffice. As for correlating scientific advance with population, although I won’t claim that there isn’t positive correlation, I suspect there are stronger correlations for scientific advancement than population. For example, a society’s acceptance of free-thinking uninhibited by oppressive environments (such as theology) would likely be a stronger correlation. I also suspect that population growth is in fact more correlated with scientific advancement than the converse!
I don’t claim that it is not theoretically possible for the planet to support current or larger human populations at a high quality, but it would require pushing many many limits that frankly are a tall order for humans…. noting my assertion that selfishness is a well entrained human trait not easily extinguished. Regardless of how optimistic one is with technological advances, the resources needed to sustain 10 billion humans at a high quality of life (ie as defined by Sam in the blog… lots of free time to develop the arts and allow philosophers to sit around at TED conferences debating the future) are not sustainable.
As for terraforming other planets, I am quite certain human populations will collapse under environmental pressure long before we attain the practical capabilities needed.
Let’s lose political correctness for a moment and admit that the best solution (ie the lowest common denominator that addresses the bulk of our problems) is to greatly reduce human population. Yes we need to continue ethical, moral and scientific growth, but we’re never going to get out in front of our over-bearing population… until mother nature does it for us.
So here’s a good question of ethics for this audience: is it ethically acceptable to continue to pump out babies into a future that as of today our science cannot possibly support at the quality of life levels currently summited? Using the technology of Today and NOT assuming there’s a silver bullet down the road is a responsible approach. Is it thus ethical to allow our near future prodigy to collapse under the weight of overpopulation rather than taking proactive actions now? I don’t buy Sam’s inference that curtailing a life not yet conceived is somehow a moral failing or net loss to our civilization. In business one can make an argument about lost opportunity cost, but to argue that controlling population is an opportunity cost for mankind’s future is quite a stretch. I could very easily argue that the opportunity cost lost is in fact the very act of disadvantaging our future by overburdening it with present day population control failings. Think of the education resources that could today be applied to a smaller population, considering the depth of our knowledge today vs our population a millennium ago!
When and if we learn how to harness free, clean energy, I’d be happy to entertain discussions on what theoretical population limits are achievable. For the time being, it’s time to start driving population down to pre-industrial age levels.
Jason66, today’s global population and your anxiety notwithstanding, the trends all look good. The U.S. has more forested acreage than anytime since the 19th century; the air and water is cleaner; people are living longer and (if they take the trouble) healthier. Similar trends in other advanced economies. Population growth in China is peaking and will soon decline. India is a bit behind but will follow a similar curve. But these salutary trends depend on continued economic growth, because the progress has been made possible by wealth.
Human genetic diversity is fully sustainable at much lower population levels… is it not?
My understanding is that the population trends indicate that we may be asymptotically approaching a population peak… but there is ZERO trend to indicate that global population will WILLINGLY drive itself to much smaller populations for the purpose of long term sustainability. Can anyone really claim that with a little more conservation/recycling etc. that planet earth can bear today’s population long term? Our Oceans are collapsing, our biodiversity is collapsing, our CO2 production is increasing and the trees needed to absorb it are disappearing. The protein needed to support human life is so overwhelming that, as Sam fears, we have turned our food chain into an ethical blackhole. Our clean freshwater is quickly disappearing.
I don’t know about Emerich’s claims concerning how better off the US environment is today vs 150 years ago ( I suspect there is some truth mixed with some failings too), but I can assure you that MOST of the remaining planet is in environmental free fall. I live in the Far East where about half of the world’s population resides, and trust me, the environment’s ability to sustain Asian populations long term is doubtful.
Raise your hand if you think Asia won’t collapse when the ocean fisheries fail and fresh air & water is a distant memory. Does anyone know what is going on with water tables around the planet? And that is without the impact of Global Warming flooding if those projections turn out to be true. A recent few hundred thousand migrants to Europe and Western political stability is faltering. Wait until migration levels reach the 100s of millions around the world.
Talking about rescuing a painting so that it’s value can be used to save 1000 fold impoverished children is completely missing the point. And by the way, if the owner wasn’t willing to sell it and offer the proceeds to charity, why would they do so after the fire?
Anyway, my humble point is that humanity should be seeking an equilibrium population that is EASILY sustainable… not trying to live on the very edge of resource limits where Sam giving up his $25 lunch is needed to sustain human well-being globally.
I too live in the “Far East” (Hong Kong). Beijing and many other Chinese cities’ air is horribly polluted. Did you know that Los Angeles’s air was horribly polluted in the 60’s, Tokyo’s in the 70’s? Both are now much cleaner. If you’ve travelled around Asia, you should know that the dirtiest cities are in emerging countries. Seoul’s streets and its air are clean. Bangkok, Mumbai, and Jakarta’s air(s) are filthy. You’re confusing yourself by using metaphoric language. What does it mean for an ocean to “collapse”? Water doesn’t compress easily. There’s nothing unsustainable about where we are. Read a little history. Paul Ehrich, author of the much acclaimed “Population Bomb”, written in 1968, predicted mass starvation by the late 70s. Why didn’t that happen? Were you around for “peak oil” just a few years ago? If you’re interested in such topics, you should also know about or the Club of Rome report, which predicted massive shortages in commodities by the 1980s. You need to understand why these pessimists were wrong.