Does evolution select for mortality?
At a recent Reddit AMA, Eric Lander, a professor of biology who played an important part in the Human Genome Project, answered this question:
Do you think immortatility is technically possible for human beings?
I don’t think immortality is technically possible—evolution has installed many many mechanisms to ensure that organisms die and make room for the next generation. I bet it is going to be very hard to completely overcome all these mechanisms.
This seems to me, at first blush, to exhibit the Evolution of Species Fairy fallacy. Evolution doesn’t work to benefit species, populations, or the “next generation”. If a mutation arises that increases longevity, and has no other downsides, then animals with that mutation should become more common in the gene pool, because they die less often. I remember reading that the effect would not be very strong, because most animals don’t die of old age. But why would there be the opposite effect?
I am loath to attribute a very basic error to a distinguished professor of biology. Is there another explanation? Is the claim that evolution selects for mortality true?
Note: Eric went on to add:
I’m also not convinced immortality is such a good idea. A lot of human progress depends on having a new generation with new ideas. Immortality may equal stagnation.
This seems to be blatant rationalization of a preconceived idea that death is good. (I doubt he truly believes that extra progress is worth everybody dying.) So perhaps his first statement is also a form of rationalization. But it seems improbable to me that he would make such a statement about biology if he didn’t think it well-founded. More likely there’s something I’m misunderstanding.
ETA: one of the first Google results is this page at nature.com, The Evolution of Aging by Daniel Fabian, which goes into some depth on the subject. The bottom line is that it agrees with my expectation that evolution does not select for mortality. Choice quotes:
The Roman poet and philosopher Lucretius, for example, argued in his De Rerum Natura (On the Nature of Things) that aging and death are beneficial because they make room for the next generation (Bailey 1947), a view that persisted among biologists well into the 20th century. [...]
A more parsimonious evolutionary explanation for the existence of aging therefore requires an explanation that is based on individual fitness and selection, not on group selection. This was understood in the 1940′s and 1950′s by three evolutionary biologists, J.B.S. Haldane, Peter B. Medawar and George C. Williams, who realized that aging does not evolve for the “good of the species”. Instead, they argued, aging evolves because natural selection becomes inefficient at maintaining function (and fitness) at old age. Their ideas were later mathematically formalized by William D. Hamilton and Brian Charlesworth in the 1960′s and 1970′s, and today they are empirically well supported. Below we review these major evolutionary insights and the empirical evidence for why we grow old and die.
How could a distinguished professor of biology, a leader of the HGP and advisor to the US President, get something so elementary wrong, when even a biology undergrad dropout like myself notices this seems wrong?
ETA #2: Gwern points to the Wikipedia article on Evolution of Ageing, which lists several competing theories of the evolution of aging (and therefore mortality). This shows the subject is more complex than I had thought and there may be good reason to believe mortality is selected for by evolution (or at least is reliably linked to something else that is selected).
I should be glad that I didn’t discover an obvious error being committed by a distinguished professional, even if he may be ultimately wrong!
I don’t know about true. I think it would be more accurate to say “here is an academic who has adopted a particular side in the very long-running debate on what aging is and why it exists, a side with advantages and disadvantages, which everyone admits is correct in a very few instances like salmon, and which it would be rather arrogant to write off as a simple conceptual error or fallacy rather than a complex empirical question”.
Hm, the evidence summarized there is quite weak, especially when we compare programmed ageing to something like a damage model of ageing, under which about half of the items are just as likely.
So I’ll try and follow up on the claims that actually look useful: programmed ageing (or more accurately, death) in “semelparous” organisms (“Semelparous organisms are animals and plants whose life histories are organized around a single burst of reproduction.” Source here, I’m basically just reading this.).
Anyhow, there does seem to be more to it than the wikipedia article would suggest, but I think the “all ageing is programmed ageing” claim is still just obviously wrong. To supply a potentially novel counterargument to a claim from the book I cited, chromosomes having telomeres, and thus “built-in” ageing, is actually very weak evidence that ageing is an adaptation, because chromosomes with telomeres just do so many other things. This same claim of adaptedness is apparently a problem in lots of evolutionary psychology, where a study finds that something is genetically determined and then goes “and therefore it’s an adaptation.”
I don’t personally buy programmed aging (as the explanation of most organisms’ aging), but I think this is a situation more like interpretations of quantum mechanics: no theory is entirely satisfactory, and this theory I am skeptical of (Copenhagen, or programmed aging) is not so obviously wrong that I should dismiss it as irrationality.
Hm. I admit that in the historical picture, it makes a lot of sense. There’s no fundamental reason why things have to die of old age, and the first microscopic “damage model” of ageing cited turned out to not fit the facts. And in a certain gene-driven viewpoint, it makes total sense that if you want the body to do something new as it ages, you need additional genes to make it do the new stuff.
But we get to take advantage of modern knowledge about how microscopic damage accumulates (see e.g. Aubrey de Grey’s stuff) to go beyond genes a bit, and when you see stuff like the metabolic rate and lifespan being really close to inversely correlated (the seed of the idea dates back to the 1930s, but I think the connection between metabolic rate and lifespan may be more recent), it’s good confirmation.
So, I guess what I’m actually confident about is that a damage model of ageing is correct for most animals, and that we’re not like octopuses or annual flowering plants. But that doesn’t actually mean that our ageing can’t be an adaptation—it would just mean that rather than dying at some specific time, the human body plan ages on some time scale that made sense for some long-gone ancestor, e.g. an early multicellular animal. This does have a slight falsifiability problem, though—either we simply never stumbled upon and spread the genes that would fix our damage, or we did stumble upon those genes, but didn’t spread them because ageing is good for us. Both leave essentially identical fossils and genomes behind.
One piece of evidence may be mutations that would have changed the time scale of ageing without granting immortality—it certainly seems unlikely that all these different species are still best served by dying with the proportionality to metabolic rate set down by our ancestor.
Thanks! That makes the debate appear more complex and nuanced. It would take me a while to go through the discussion and links there and I don’t have that time today. There seems to be a real question to be answered in “why do almost all animals age?” and accumulated damage theories are not obviously a complete explanation.
A fairly complete explanation of most aging:
Disposable soma theory
Antagonistic pleiotropy
Damage accumulation
Reliability theory
And none of these argues that evolution selects for mortality or aging.
That’s true. However the antagonistic pleiotropy and disposable soma theories have senescence as the product of selection. Basically, all the theories are bad news for anyone hoping that aging and death will somehow go away. Fighting aging is fighting against the natural tendencies of systems to age and die.
In one sense, mortality is definitely selected for, though indirectly. Natural selection selects on the basis of number of offspring who go on to reproduce, not length of life. If 95% of the organisms in a species die from predation before 5 years, then an adaptation that causes muscular dystrophy after that age but increases the vigor of the organisms before that age so that they can escape predators and reproduce 10% more before that age, that trait will be selected for. Once there is a trait that causes one system to become fatally nonfunctional at 5 years, any trait that increases fitness for the first 5 years but causes another system to fail around 5 years out will also be selected for. So you would expect to see a large number of system failures around the maximum lifespan of an organism. This is actually what we see with humans—at 70-100 years, hearts, livers, and kidneys all start to fail. Neurological diseases like Alzheimer’s also become far more common around that age. My prediction is that when we figure out how to replicate the functions of hearts and kidneys well enough to allow artificial versions to function indefinitely, we will discover that the human body fails in new and interesting ways shortly thereafter.
It may well be possible to overcome these challenges, but it’s not likely to be a simple matter of tweaking the concentration of a couple of compounds that increase in quantity as we age. Instead, I’m guessing it will involve engineering replacement organs specifically designed to operate indefinitely.
As for death being net positive, I disagree, but can see his point. Historically, large societal changes have coincided with the old guard dying off, though not every instance of the old guard dying off caused a large societal change. But it’s a very easy and natural conclusion to draw, and reflects the society he was exposed to more than it reflects any huge irrationality on his part.
That’s one of the major theories of aging. The difference is that it doesn’t say aging itself is selected for, the way Eric Lander claims. It just says aging is accidentally genetically linked to something else that’s being selected for.
The failure of many systems at approximately the same time is selected for. Death as a result of the failure of those many systems is not itself selected for, but the reproductive benefits of systems that only have to function for a set amount of time are selected for. It’s not exactly correct, but it’s a reasonable approximation for talking to laypeople, as it’s close enough to the truth to allow people to make reasonably accurate predictions about the world (e.g. this new wonder drug won’t cure aging). Whereas if you say that aging is “accidentally linked” to another trait, that makes it sound like we just have to identify one or two traits and we can cure aging, so while it may be closer to being literally true, it’s also more likely to cause misconceptions in people who don’t have more than a layperson’s grasp on the mechanisms of evolution.
Maybe in a few cases. For instance, immobile organisms often take their parasite load out of circulation when they die, reducing the risk of their nearby relatives becoming infected.
However the main reason that things die is that making them long-lived is expensive and difficult. Anti-aging features are thus usually selected against—they cost too much. It is simpler, easier and cheaper to start again—now and again.
Just wanted to note that, as any programmer or experimental biologist knows all too well, designing a program or an experiment to last a long time is expensive, with the effort required often scaling superlinearly with run time. And most features not needed for genetic fitness get optimized out during evolution, like eyesight of a cave dweller.
But longevity does contribute to genetic fitness and is selected for if present.
I don’t see how this is a “but”.
I don’t either. I may have replied to the wrong comment by accident. Sorry.
No, I think you’re right. Eric Lander is a well-respected scientist in his field, but his field is not ethics.
Reddit is not an academic paper. Biology has many subdisciplines and there’s no reason to expect an expert in one subdiscipline to know everything about other subdisciplines. Your impressions are not necessarily accurate. Take your pick.
The two alternatives (which I think are correct) is that immortal varieties either never arise in the first place, and so can’t be selected; or else are always linked to some tradeoff or disadvantage which is selected against.
Your explanation is isomorphic to saying “most organisms die if heated to 100 degrees C, therefore evolution selects for organisms that die if heated”. But the simpler explanation is that resistance to extreme heat never arises in the first place, or has a large associated cost whenever it appears. Not that such resistance is in itself harmful and selected against.
His field is biology, and this (seems to be) a fundamental error in biology (evolution).
Well, this is not true. Some organisms actually are immortal.
Yes, I noticed this, which is why I removed that part of my comment.
Again, biology has many subdisciplines and there’s no reason to expect an expert in one subdiscipline to know everything about other subdisciplines. And even experts occasionally make mistakes. I mean, seriously.
Yes, and that is weak evidence against theories that say evolution selects against immortality.
Because there are so many things, people are going to get many things wrong.
Though I agree that as soon as someone starts thinking that one of the options is “incomprehensible,” they’re probably not going to think very clearly from there on out.
I just tried ctrl+F here and at the linked thread—who used the word “incomprehensible” and where?
You have successfully figured out that nobody used the word incomprehensible in the reddit thread :)
I was only attempting to mindread—sorry for poor use of quotes.
Perhaps interestingly, my model here is the reverse of my model for humans failing at hypotheticals: if you ask people to consider the hypothetical where punching kittens is a good idea for some low-probability reason, and ask them what they’d do about it, they can’t do it properly. It’s not in the design specs of their brain. Similarly, if someone encounters the same feeling of finding it hard to think about something, I suspect they conflate that with it being a hypothetical of low probability.
Boy, it’s a good thing we have biology undergrad dropouts to correct them! Reasonable default: the distinguished professor is probably right, and the undergrad dropout probably hadn’t thought about the problem hard enough.
Physicists have this problem too (http://xkcd.com/793/), but at least physicists have a track record for insightful, generalizable mathematical models (the Ising model is a good example).
I starting typing stuff, then I realized I was assuming that people couldn’t stay fertile arbitrarily long. Of course if you’re no longer fertile at 50 Azathoth won’t care much whether you’ll live until 60 or until 400 (except insofar as you can help/hamper your children in having children themselves), but why couldn’t you be fertile past 50 in the first place?
Most animals are fertile throughout their lives, although fertility often declines with age, like most functions.
Humans have the (female) menopause, and it is unclear what the evolutionary reason for it may be, but for most animals this is irrelevant.
Interesting.
From the last sentence of that article, it might be the case that captivity/domestication would increase total lifespan (unsurprising, given better nutrition and healthcare), but not reproductive lifespan, which in the wild are usually about the same.
I seemed to recall that human females are born with a finite number of ova which aren’t replenished in adults and menopause was simply due to them running out of ova (which would explain why there was a maximum reproductive age, which if it was longer than typical lifespans in the EEA wouldn’t be subject to any evolutionary pressures and so it would be stuck to an arbitrary value until modern times), but apparently it isn’t uncontroversial that that is the case and anyway the numbers don’t seem anywhere near small enough for that to be possibly an issue. Huh.
This is my first post in a long time and I’m having trouble formatting it. It seems the editor assigns different fonts to different paragraphs; this is not visible in the WYSIWYG post editor, but it’s visible in the actual post. In the HTML editor I see that each paragraph specifies an explicit font, which is sometimes different.
For instance, right now the paragraph saying “Note: Eric went on to add”, and also the paragraphs starting with the final quote (“The Roman poet ….”) are in a smaller font than all the rest. This is even though I used the “remove formatting” button in the WYSIWYG editor.
Can an experienced editor please help me? How do I reset all text to a single, default font?
In the HTML editor, remove all of the font tags.
Thanks, that did it. (There were <span style=… tags around each paragraph.) But why were they there in the first place?
Text that’s copy-pasted into the post editor has its formatting preserved by default.
Note for others: some browsers (like Chrome) give a ‘Paste as Plain Text” option to help with this. Just right-click and select that option rather than ‘paste’. Presumably it works in this editor, too, but I haven’t tried it.
That explains it, thanks!
I think that you are right and Lander is wrong.
However, it is curious that most mammals such as dogs and horses die much younger than we do, despite being made of essentially the same stuff. Certainly we could not exist if we died under twenty years because it takes us that long to mature our minds and breed. But what advantage for a dog to die young? If it lived twice as long it would (presumably) produce twice as many grandchildren.
I suspect that it is simply that dogs and horses can breed after a couple of years. So once they live more than 6 or so times their breeding age their is not that much advantage in them living any longer. But there would still be some advantage. Is there some cost to living longer, such as needing to have a slower metabolism, or is it just that natural selection does not produce unneeded features?
Evolution doesn’t select against immortality in any specific case. How could it select against it in general?
Selecting against immortality itself would (in my view; apparently not in everyone’s) be almost an oxymoron.
But there are other ways that shorter-lived variants might be preferred by selection. The major proposed ones have in common the idea of tradeoffs: that the genes for mortality at an old age also increase fitness at a young age more than enough to compensate. The mechanism might be e.g. investing a lot of energy into producing eggs, versus investing a lot of energy into building a durable long-lived body.
I don’t think that a durable, long-lived body (for the purposes of this discussion, call it a mammal that does not lose reproductive fitness with age) has ever existed in large enough numbers to influence the gene pool.
There is little evolutionary pressure to remain reproductive for an entire lifetime if that lifetime is not much longer than the current reproductive lifetime; likewise there is little evolutionary pressure to extend the overall lifetime far beyond the reproductive lifetime. It’s possible, but by no means certain, that the two are unrelated.
Almost all animals can reproduce as long as they’re alive. They may become less fertile with age, but not totally unfertile. The menopause of human women is not uniuque, but it is very rare.
How fertile is a 30-year old dog? I meant ‘lose reproductive fitness with age’ to be inclusive of all semi-deterministic effects of age, not just menopause.
I would expect a 30-year old dog to be as fertile as it is generally healthy. In other words, I expect fertility not to decline much sooner than the rest of the body. If it does, we recognize it as something special, like the menopause.
There’s a problem predicting purely from theory. No single component or function (whether fertility or something else) would normally age faster than the others, because evolution would select against that. But if a crucial function like fertility stopped working after a certain age and there weren’t any selectable variants that kept it working longer, then animals that lived longer but were infertile wouldn’t have a much greater fitness than those who died as soon as they became infertile, and we would see the same result—animals being fertile roughly as long as they live.
Exactly- There have never been selectable variants that are immortal (or even nearly so). Whether that is because immortality is hard (unlikely to result from random drift) or not is irrelevant for the purposes of natural selection if it never happens.
The premise is false. (Most straightforwardly: Females eat food and have a finite number of eggs.)
Therefore females who need not eat food and produce additional eggs die?
Therefore, there are specific cases where such an individual’s inclusive fitness will be higher if she dies than if she lives. On account of the reproductive offspring not being starved.
Note that this is a rejection of the argument, not necessarily the conclusion.
What would happen to the offspring of an ageless mother in that circumstance?
Parsimony. It takes chemical resources (ie food) to produce any given thing the body uses, and there’s selection pressure to use as little as it takes to do the job (ie produce further generations) because on the margins those who require fewer resources should outbreed those who need more. If most members of a species die in xty years or stop reproducing in those years, this creates parsimonious selection against the different needs of either an aged creature or a regenerative one.
… Except that it strongly selects FOR organisms that have an infinite number of offspring, regardless of other pressures.
Depends on when the reproduction happens. Say the organism becomes sexually mature in 20 years, then as far as evolution is concerned the organism 40 year old self is the equivalent of two addition offspring in the first litter. And that’s assuming the organism is as good at reproducing at 40 as at 20.
It’s all about the death rate from things other than aging, or how much immortality extends expected lifespan. Side effects of immortality which reduce competitiveness aren’t inherent with immortality.
You seem to be cherry-picking quotes. From the same paper:
Aging clearly exists, and is (almost) universal. But the fact that organisms have a certain property is only weak evidence that property was selected for. As I commented here, the other possible explanations are that immortal varieties either never arise in the first place, and so can’t be selected; or else are always linked to some tradeoff or disadvantage which is selected against. And both options seem quite plausible and I believe have been demonstrated in specific cases.
The mechanism discussed in the article is (from memory, having read about the linked article once) roughly as follows. A cell accumulates damage over time (‘ages’). When it divides, it can split damage between both child cells, making them both somewhat younger. Or it can place most of the damaged organelles in one child cell, producing a young and an old cell. This is a tradeoff that (unsurprisingly) is resolved differently by different cells at different times.
But the fundamental idea is that aging is damage accumulating over time. That’s what the quote about “a fundamental and inevitable property of cellular life” refers to. There isn’t a gene or behavior ‘for’ aging, and therefore there isn’t an available mutation that doesn’t age because it doesn’t have that gene, and so evolution can’t select it.
Contrary to Lander’s view that “evolution has installed many many mechanisms to ensure that organisms die and make room for the next generation”.
Complex animals nonetheless have a method of producing young cells in their offspring—which capability is obviously a requirement for gene survival. That we don’t have the same sort of capability to refresh the cells in our own mature bodies is either an accident[1], or the costs/risks[2] of it aren’t worth it for our own benefit, or (per Lander, I guess) is counterproductive to your genes compared to short aging-enforced reproductive generations.
(your argument is good and I agree with it)
[1] (more complex design simply hasn’t been reached yet, or it’s impractical for adults, requiring a womb, etc.)
[2] I have no idea what risks would be present in a world where we replenish ourselves with young cells, but I can imagine at least novel types of bad growth and rejuvenation, aside from the normal cancers
My comment was about the tell tale signs of motivated cognition by the OP, not about the substance of the topic.
I don’t understand.
I quoted the paper in support of my view (and in opposition to Lander’s). Then you said I cherry-picked that quote and here was another quote from the same paper that said something different. But I’m saying, no look, that other quote isn’t different, it doesn’t contradict what I’m saying at all.
So how am I exhibiting motivated cognition? Is it just because I initially picked the quote that best supports my case? The quote I didn’t pick didn’t contradict my case...
Originally I was reacting to
before your second edit. LWers have a strong anti-deathism bias, at least anti-deathism for humans (chickens might be another matter entirely), and your hasty conclusion was an indication of it, as confirmed by your quote selection.
Your second comment appears to be factually wrong:
There are definitely are genes which affect aging, so there is something to select for.
I agree that when Lander says
he shows a deathism bias, since he probably does not have the data to support his statement. A more charitable reading would be that prevention of aging is hard and requires a lot of genetic changes, something evolution never needed to bother with, so researchers have their work cut out for them.
When I wrote,
That sentence should be read with a ‘but’ at the beginning. I was saying that I couldn’t believe he would make such a basic mistake. I thought I must be missing something. That’s why I posted here instead of dismissing the whole thing as him being wrong.
I recognize I’m emotionally biased against deathism and thereby against people who appear to be pro-deathism. But I did make an effort to find out the actual truth.
You’re right; there are definitely genes which affect aging and longevity and can be selected. And I expect that if they have no other effect, then the selection prefers those variants which confer longevity. It’s only when they are tied to tradeoffs or other effects caused by the same genes, that the selected variant may not be the longest-lived.