David Brooks, New York Times columnist and author of The Social Animal, has never met an evolutionary psychology argument he didn’t like. I haven’t read his book, but I did read a long excerpt in The New Yorker and found it credulous, tedious, and lame. P.Z. Myers, who reviewed the book, had the same opinion. So did philosopher Thomas Nagel, who, reviewing the book in the New York Times, pretty much ripped it apart, noting that “Brooks seems willing to take seriously any claim by a cognitive scientist, however idiotic. . ” (It’s quite unusual for the Times to publish bad reviews of books by their own columnists.)
In yesterday’s New York Times, Brooks writes about recent scientific “advances” in the understanding of human altruism. And he signs on to the idea that altruism evolved by group selection.
I disagree, and see Brooks as ignorant about the true scientific issues. If true altruism (which I define here) is indeed a trait that’s deleterious to an individual’s reproductive fitness, then it could, as Brooks envisions, evolve only by the differential survival and reproduction of groups.
That form of evolution would work like this: although genes for altruistic behavior would be constantly weeded out of populations (for altruists, by definition, sacrifice their own genetic heritage for others), those genes might survive if groups that contained higher proportions of altruists were the groups that persisted, giving rise to descendant groups more often than groups lacking altruists. (The idea here is that groups without altruists wouldn’t flourish very well.) That’s group selection, and it’s how Brooks sees altruism as evolving:
In his book, “The Righteous Mind,” to be published early next year, Jonathan Haidt joins Edward O. Wilson, David Sloan Wilson, and others who argue that natural selection takes place not only when individuals compete with other individuals, but also when groups compete with other groups. Both competitions are examples of the survival of the fittest, but when groups compete, it’s the cohesive, cooperative, internally altruistic groups that win and pass on their genes. The idea of “group selection” was heresy a few years ago, but there is momentum behind it now.
Let’s be clear about what biologists really know about group selection and altruism. If true human altruism has a genetic basis, it is individually disadvantageous and could have evolved only by differential propagation of groups. That’s very unlikely, since it requires that the rate at which altruist-containing groups reproduce themselves must be high enough to counteract the substantial rate at which altruism genes disappear within groups. It’s unlikely because groups reproduce much less often than do individuals! Further, once a group consists entirely of altruists, any non-altruistic genes would rapidly invade it, as their carriers reap the benefits of altruism without sacrificing their reproduction.
Now if we’re talking about apparent altruism, in which individuals appear to sacrifice their reproductive interests but actually reap hidden genetic benefits, then we don’t need group selection to explain it. As I’ve written in a longer post on this topic, kin selection (“inclusive fitness”) can do it, as can simple individual selection based on reciprocity or, simply. selection for the advantages of cooperation, as in hunting lions.
Humans, after all, evolved in small social groups, which provide the ideal environment for the evolution of “reciprocal altruism” (“I scratch your back and you scratch mine”). That kind of altruism, which isn’t “true” altruism in the sense of hurting one’s reproductive prospects, evolves most readily in small groups where individuals know and recognize each other, and have a big brain for remembering and reciprocating good deeds. Living in groups, particularly of kin, facilitates the evolution of apparent altruism, but that is not group selection since it doesn’t require differential propagation of groups. Genes that are selected in groups based on relatedness or individual advantage will spread throughout a species without requiring differential reproduction of groups. That is, selection occurs in the context of groups, but doesn’t occur through selection among groups.
What do we know about human altruism? First of all, we don’t know whether true altruism, in which individuals behave in ways that help others by hurting their own reproductive prospects (firemen are one example), has any genetic basis in human society. True altruism like that isn’t known in any other species, and I suspect that, to the extent it occurs in ours, it’s an epiphenomenon: a byproduct of our general social cooperativeness. As far as whether we are genetically cooperative (rather than truly altruistic), that seems quite likely, but it doesn’t require group selection. It requires selection that occurred in groups, which is different. And we almost certainly have some behaviors that evolved by kin selection, parental care being the most obvious.
So Brooks misrepresents the views of biologists in his piece. There really isn’t much momentum in the evolution community behind the idea of “group selection.” There is increasing realization that selection can occur in groups, that being in groups can affect how selection operates on genes, and that there can be group effects (“multilevel selection”) that influence the evolution of genes. But there is no general feeling that “group selection” is widespread or important. And there is no widespread agreement that true altruism, or even apparent altruism, evolved by the differential propagation of groups.
In short, we know nothing about the evolution of true human altruism except that it probably didn’t evolve. And we don’t know much more about the evolution of human cooperation. It almost certainly has a genetic basis—we’re social animals, after all—but we’re ignorant about the form of natural selection that favored such cooperation, and about the social and environmental circumstances that promoted that selection.
Brooks makes one more biological error, asserting that the evolution of cooperation necessarily entails evolved morality.
But the big upshot is this: For decades, people tried to devise a rigorous “scientific” system to analyze behavior that would be divorced from morality. But if cooperation permeates our nature, then so does morality, and there is no escaping ethics, emotion and religion in our quest to understand who we are and how we got this way.
Cooperation also permeates the nature of honeybees, termites, naked mole rats, and lions, but they don’t have morality. Morality is the result of having a big brain that, in a social species, can remember other individuals and make calculations about their intentions. Whether that “result” is genetic, so that our moral feelings are encoded in our DNA, or simply an epiphenomenon, in which we’re taught rules that enable us to function, is an open question. I suspect that some of it is genetic, but we just don’t know.
Credulity, tediousness, and lameosity is pretty much Brooks’ m.o. whether it’s science or politics. Sadly, he gets *big* bucks for being credulous, tedious, and lame.
Good post.
I am still confused, apparently, about the claim that parental care is best viewed as an example of kin selection.
Specifically, it seems to me that the claim here that
does not jibe with this earlier distinction about the evolution of apparent altruism:
.
If you are willing to distinguish ‘simple individual selection’ from ‘kin selection’ (and, let’s face it, ‘inclusive fitness’ is really only a factor for behavioral phenotypes, pretty much everything else being limited to straightforward vertical inheritance and individual selection), then parental care seems to me to belong squarely in the individual-selection slot.
I understand that the exact same inclusive-fitness mathematics apply to the effects of behavior on both inclusive fitness and direct reproduction, but although that makes for convenient modelling I think it need not dictate the conceptual conflation of direct offspring and (say) cousins.
What am I missing?
I am not sure you are missing anything. Parental care and “horizontal” kin selection (e.g., helping relatives who are not offspring, at a cost you oneself) both work because of genic-level selection. Parents are behaving altruistically because alleles that predispose more or some parental care outcompete those predisposing little or no parental care in certain environments – despite the costs to the parent displaying the care. So it may be good to remember that only replicators persist and it may not be helpful to refer to parental care as kin selection rather than individual selection (ultimately, is there really such a thing as individual selection – given the breakup of multi-locus genotypes in populations with recombination?).
Can’t say for sure that you’re missing anything; part of it might be that you’re trying to put kin selection and individual selection into separate “slots” where in reality they’re both manifestations of gene selection. In some instances, gene selection could lead a parent to leave as many eggs around as possible while leaving the offspring to take care of themselves and in others, gene selection could lead to parental behavior. In the latter case, the specific instance of gene selection could be called “kin selection” (because the adult gene bag sacrifices opportunities to copy its genes while enhancing the opportunities of its children to copy their genes).
That’s essentially how I’d put it. Children are less efficient at producing copies of the parent’s genes than the parent itself is, so the parent nurturing its children instead of serial-birthing constitutes a slight short-term disadvantage to itself, making it technically kin selection.
But I’m no biologist. Please let me know if I’m not making sense.
Right, not missing anything. An often-overlooked statement in Bill Hamilton’s original paper on kin selection is “there is nothing special about the parent-offspring relationship except its close degree and a certain fundamental asymmetry”.
What he was saying here, I think, is that we’re all perfectly ok with parental care having evolved even though it can be seen as costly to the parent, simply because taking care of one’s offspring means helping one’s genes be passed along. (The fundamental asymmetry is that it’s far more likely for parents to help offspring than vice versa, simply because they are older and more able to do it.)
So, he reasons, why can’t the same principle be logically extended to other individuals who are likely to pass along one’s genes as well? The answer is that it can. Ergo, kin selection.
I guess the NYT is not content with misrepresenting science in their Science section, by publishing such incompetents as John Tierney and Nicholas Wade, now they feel obligated to misrepresent science on the op-ed page as well.
One seldom sees anything by Brooks that isn’t full of nonsense. How did he get that job, anyway?
You answered your own question.
I donated blood today, and while I was waiting as my lifeblood was sapped I thought about Dawkins’ proposal that blood donation was a possible example of a truly altruistic act — because the donor or his/her kin are highly unlikely to benefit personally from the activity.
On the other hand, my wife’s and mother’s lives were saved by blood transfusions, and it doesn’t take much imagination for me to see donation as “enlightened” self interest: it could be me or a close relative who needs the stuff tomorrow. Actually I just enjoy the good feeling that comes helping other people, even though I don’t know them personally.
It’s also not hard to imagine such feelings evolving when one was related to most people in one’s social group — and so the roots of these feelings might be from kin selection as well.
The donation centre was full to capacity today, as usual, and so these feelings are probably common.
I have to put this link – British comedian Tony Hancock in his famous Blood Donor role…!
That “good feeling” is probably what Coyne is referring to as an epiphenomenon. You have a “sloppy” instinct for helping your own genes persist. You inherited a desire to help people who have your genes, but no biochemical “kin-detection” equipment to go with it. So you treat many unrelated people like they have your genes.
Think of it like baby ducklings imprinting on a kitteh. They got an instinct to love the first thing they see. 99% of the time, this is just as good as a “love your parent” instinct, because in the wild their parent IS the first thing they see. And it’s much easier to evolve because it doesn’t require any kin-detecting equipment.
You might say that chicks and humans both got cheap knock-off versions of the instinct we needed, and feelings of support for unrelated beings is the result. 🙂 Your good feeling giving blood is like a chick looking at the cat and feeling love.
Awww… fluffy kitties & ducklings!
And, let’s not forget: you share the overwhelming majority of your genes with everything in the animal kingdom, and a huge number with everything in every other kingdom.
We are all kin; it’s just that some kin are closer than others.
Considering the universal cross-species impulses amongst mammals to nurture mammal infants along with physical characteristics shared by mammals…well, when you cuddle and care for that cute little kitten, you’re propagating the “cute baby” genes the two of you share.
Cheers,
b&
Exactly how does donating blood effect your lifetime reproductive fitness?
I suspect there’s a greater chance you’ll meet a reproductive partner of similar attitudes while donating blood than you have of missing an opportunity due to a temporary period of limited incapacity/serious damage through an accident during transfusion.
Marios
This seems to be a growing market. Philosophers and journalists being complete dilettantes in biology but getting a mouthpiece by the mainstream media.
This is obviously the biologists fault for not making sweeping generalizations about humankind every time they make a minor adjustment to (or even just have findings that support) a well established theory.
See, if they’d just be sensationalist and wrong more often we wouldn’t have to deal with all these wrong, sensationalist evo psych people gumming up the works.
And that would be the solution to unscientific America!
I’ve read quite a few pieces by Richard Dawkins and David Sloan Wilson, who seem to have been arguing since the 70s, on group selection, and I’m still not sure the extent to which they are actually disagreeing.
It seems that when Dawkins (and others) criticise “group selection” they mean the “naive group selection” which you criticise above. However, when Wilson (and others) defend “group selection” they really mean the “multilevel selection” you refer to. Is this a reasonable summary?
So, is Brooks mischaracterizing Wilson, Sloan Wilson, and Haidt? What is the situation in biology without the filter of a newspaper columnist?
I have never been impressed with Brooks, nor with EP. The group selection supported by Brooks is implausible.
There is, however, another possible meaning for “group selection” which might have played a role. That would be when a group follows a path that requires social cooperation within the group, and that group exiles (or kills) any member of the group who is not sufficiently cooperative. That would be selection by the group, rather than selection of the group. It would be somewhat analogous to sexual selection.
I see the point, but I would not call it ‘selection by the group.’ I think this rather is a case of positive frequency-dependent selection. An individual exhibiting a rare, uncooperative phenotype is unable to ‘persist’ in or ‘invade’ a group consisting of phenotypes that tend to enforce social cooperation.
Im a grad student in molecular biology so I dont know all the literature- but I’ve been reading Razib Khan over at discover blogs, and one thing he’s been mentioning a lot lately is that recent human evolution seems to be largely about near total sweeps where one agricultural group moves in and almost completely replaces all the indigenous hunter gatherer populations. Africa, India, and Northern europe all seem to have complete demographic replacement of this kind visible in the genetic data. I’m wondering if group selection theories work differently in this kind of framework than they do in the old ‘differential reproduction of groups’ notion? I suspect it still wouldn’t save an allele that was actually unfit, but as to the broader implications, im really curious.
I had a similar thought when reading Jerry Coyne. Groups don’t reproduce much, but they do compete. Over time, couldn’t repeated group triumps result in the selection of individuals with social traits/skills that favor the group over the competing groups?
If I can paraphrase a bit of The Selfish Gene, the genes for “group” altruism (if ay) would be outcompeted within the group by those for simple selfishness. Those genetic traits that put the group preservation ahead of their own genetic reproduction wouldn’t last long.
I think we can make a case for group-preserving memes (e.g, those that form a culture) beating purely selfish memes, but it’s much harder to make that case for genes.
There seems to be a pattern in which non-biologists – and some biologists – find traditional group selection much more appealing, intuitive, and therefore plausible than genic-level selection. In other words, they seem to depart from individual selection in the wrong direction when trying to understand a phenomenon like altruism.
Even though we recognize the unfortunate tendency for some to accept any evolutionary psychology argument, I live in a part of the country where most folks are loathe to consider any biological/evolutionary basis of human behavior – and may be too quick to dismiss all evolutionary psychology hypotheses. They would prefer to think of human personalities as either emerging with a blank slate (so all variation is imposed by culture) or reflecting a non-material pre-formed ‘soul.’ That is, they ascribe to two of the myths Pinker debunks so well in the The Blank Slate.
Definitely a fine line to walk. Behavior certainly has a genetic component, but proving that in specifics for humans is daunting to say the least…and hypothesizing about it so prone to individual biases.
Prof. Coyne, thanks for a great post! It really helps clarify some of the recent furor over group selection.
“First of all, we don’t know whether true altruism, in which individuals behave in ways that help others by hurting their own reproductive prospects (firemen are one example), …”
Supposing a fireman gets paid a good wage which raises his social status and which he uses to support his family. Is it then “true” altruism in that definition?
No, I agree (was writing below while you posted this) – s/he gets a societal reward of pay & kudos as well as the benefit of well-being, & I would say it is also a competitive job because they will weed out the person who does not pull their weight. The fact that is has gone beyond immediate reproductive fitness does not mean that it does not emerge from a biological basis (but then all behaviour must have a biological basis??)…
Another wrinkle–the “thrill-seeking” personality is probably maintained because it is adaptive under some conditions. If a person is thus prone to being a firefighter, is that altruistic or just a part of our genetic variation?
This is interesting stuff. I think perhaps an element in understanding “Apparent” altruism is competition and handicap. Assume for example that running a race might have been originally to show that you are better than a rival, equally expending a large amount of effort on any activity demonstrated that you were easily able to carry the loss and still retain fitness or an ability to garner resources. It is a short cultural step from there to the altruism of a fireman or a soldier saving comrades. Why do people like Michael [previous post] save cats? I think we are ‘programmed’ to look for cuteness, equally domestic cats have co-evolved with humans to be cute (I know Marta will cite Evil but it is a continuum), & I bet Michael felt good about helping the cat so this is not “True” altruism either. I think that cannot exist as the person doing whatever the altruistic act is, even if anonymous, will always feel something positive, obtaining some benefit I would guess.
Another thing I would note is that we are not either altruistic or not, we are a mixture, depending on the circumstances, but perhaps that is a given?
Nice exposition, Jerry. Illuminating.
There is no “altruism” it has been repeatedly demonstrated to benefit the genes of, largely related, individuals.
Here is the latest research we have heard: ”
May 09, 2011
Is Kin Selection Dead and Is It Time to Move On in Understanding the Evolution of Cooperation?
Peter Nonacs, UCLA Department of Ecology and Evolutionary Biology
Hamilton’s theory of inclusive fitness broadly states that whether or not a trait increases in frequency is dependent on both the direct reproductive success of individuals having that trait and the help that such individuals can provide to other trait bearers for their reproduction. The latter portion of inclusive fitness is commonly known as kin selection and has become the dominant paradigm for the evolution of cooperation: I.e., helping genetic relatives reproduce can create a net increase in inclusive fitness even with a substantial loss in direct reproduction. Recently, however, Martin Nowak has argued that the mathematical foundations of inclusive fitness theory are inappropriate for predicting the evolution of cooperation (1). Edward O. Wilson has gone even further and claimed that, “Kin selection is wrong” and a “gimmick” (2). Instead, Wilson proposes cooperation evolves through group selection. Not surprisingly, their claims have drawn considerable criticism (3), with Richard Dawkins going so far as to pronounce that he has “never met anybody apart from Wilson and Nowak who takes it seriously (2).” I will look at both sides of this issue and attempt to separate the scientific concerns from the heated clashes of personalities. At issue appears to be the question of the evolutionary advantages of genetic diversity versus kinship. Both can be advantageous, but they are simultaneously incompatible. Their resolution requires a multi-level approach as nepotism favoring kin can be selected for within groups, but genetic diversity is selected for only across groups.”
http://bec.ucla.edu/presentation.php?id=237
Brooks’ goal is mainly to bolster “happy talk” social and political ideology using “science” — it’s silly and dishonest. Thus, he is lionized and has a NYT column.
Don’t get me started on the misunderstandings of Wilson and Wilson regarding inclusive fitness theory.
I saw the younger Wilson (David Sloan) speak on the topic of group selection, inclusive fitness, and the evolution of religious behavior in humans a few months back here in Welly.
He’s nuts.
sorry, but that’s putting it both succinctly and accurately.
I am almost finished reading The Social Animal and it is truly an awful book. A lot of it is just a fictional account of a couple as they move through life that makes my scalp itch as I read it. His syllogisms are very misdirected.
“There really isn’t much momentum in the evolution community behind the idea of ‘group selection.’ … [T]here is no general feeling that ‘group selection’ is widespread or important.”
Meaning “group selection” as it specifically applies to “true altruism,” correct? The evolution community accepts that “group selection” operates in other contexts, doesn’t it?
Would that not heavily depend on how you delineate the group?
I thought “group selection” as a general principle was non-controversial — say, for example, where one group has an adaptation that gives it an advantage over another, competing group. That’s why I wondered if there is an implied qualification to Professor Coyne’s statement that “group selection” isn’t “widespread or important” limiting it to the context of altruism.
no, the evolutionary biology community does not accept that there is sufficient evidence to indicate group selection operates to any significant degree, if at all, in the field.
It’s the theoretical modelers that make the case for multilevel selection, and while at times interesting, and has lead to some cool experiments, evidence in support of those models is distinctly lacking.
Has anyone else seen this nascent meme?
Thanks, that’s fun!
I just got through the chapter in “The God Delusion” where Dawkins deals with the same subject. It was nice to get another perspective on altruism in humans. A thought occurred to me while reading this though, and I’m sure it won’t matter to you Prof Coyne, but isn’t some of what you wrote in this piece just begging to be misrepresented by some unscrupulous creationist? Especially the parts about how true altruism isn’t explained by evolution, and how human beings are the only species to exhibit such behavior. From there it’s a short leap to “science can never explain altruism,” followed soon by “only Jesus can explain altruism.” Christians should probably know better by now but they still seem easily swayed by the God of the gaps argument.
Would religiously motivated altruism constitute “true altruism” — especially for otherwise-selfless acts motivated by the desire to earn one’s way into heaven? Maybe there’s another category, aside from “apparent” and “true” altruism; call it “delusional altruism.”
I wonder if they asked Nagel to review it and then were surprised he didn’t like it. That would be odd though – why would they expect him to like it?
BTW, I posted this in the sloth thread, but it actually works even better here!
There was an entire issue of the Journal of Evolutionary Biology a few years back devoted to reviewing the work looking at the evolution of altruism and cooperation.
The articles in it are all still free, AFAIK:
http://onlinelibrary.wiley.com/doi/10.1111/jeb.2006.19.issue-5/issuetoc
a treasure trove of information for those interested in the topic.
Link saved, many thanks!
How altruistic of Wiley?! …to make it free.
heh.
I can think of a couple of alternative hypotheses to explain it.
I’d like to think it’s mostly a case of altruism though.
🙂
Actually, making it “free” just generates more awareness of their other offerings thus increasing potential revenues for the firm (group) thus leading to greater resources for the owners, and some workers, allowing them to have more children and increase the reproductive attractiveness of their offspring…..
Its a really confused and waffly article, and the old New Yorker article is just painful. He’s over here at present
http://www.guardian.co.uk/politics/2011/may/19/david-brooks-big-idea-society
and the Tories seem to be listening to him, maybe because at one point he mentions their favourite point of blaming everything on single mothers.
He was interviewed by Paxman on Newsnight last night, which I saw a bit of purely as I’d tuned in for Stewart Lee.
I think Stewart Lee put it all much better: “I do 50-60 charity gigs a year. This isn’t out of altruism, I just do it because it makes me feel good about myself”.
Reposted from the death-of-bin-Laden thread, where Jerry defines “true altruism”:
“Defined biologically, this form of altruism involves individuals making sacrifices that are not repaid. By “sacrifices,” evolutionists mean “reproductive sacrifices”, that is, you forgo future reproduction through your behavior.” – Jerry Coyne
Right. So an infertile person who murders all their relatives in order to inherit their money is showing “true altruism”. Tell it to the judge.
A better term is “genetic altruism”, to distinguish it from what the word means in everyday language.
uh, no.
you’ve got that totally wrong.
instead a more correct analogy would be:
you give up your ovaries to your sisters so they can have more kids.
Jerry has it absolutely correct.
Nonsense: you’re misunderstanding Jerry. He says above:
“If true altruism (which I define here) is indeed a trait that’s deleterious to an individual’s reproductive fitness, then it could, as Brooks envisions, evolve only by the differential survival and reproduction of groups.”
This makes no sense at all if a propensity to behaviour resulting in the sacrifice of your own ability to produce offspring in favour of your sister’s ability to do so is an example of “true altruism”, since this could obviously increase inclusive fitness. It only makes sense if he is referring to behaviour which reduces inclusive fitness. My example suffices to show that such behaviour could be the very opposite of what is meant by “altruism” in everyday-speak, so “true altruism” is a poor choice of terminology.
Further to the above, we know many examples of behaviour which clearly is genetically-based, and which would be “true altruism” under your interpretation: consider any species in which adult offspring remain with their parents and help raise sibs.
Apparent altruism, apparent design. I see a theme, nota bene not an apparent one.
“True altruism”, “true design” both looks like examples of over eager pattern search.
Our problem is how can any “group” trait be be reproductively a positive outside of the individual? So what is the carrier of the group benefit?
“Robot altruism:”
http://www.technologyreview.com/blog/guest/26772/?nlid=4495
this is getting embarrassing for Brooks, what a nitball. Probably a bit of mid-age guy disinhibition going one here.
– The Dumbest Story Ever Told: On David Brooks
http://www.thenation.com/article/160752/dumbest-story-ever-told-david-brooks
his writing has just been getting more weird, there was that piece on the epiphany at a Broadway show!!
“Firemen for example” makes me laugh because that group is literally the calendar boys for “can get laid whenever they want.” That type of altruism is SO sexy, I’d be astonished if their increased opportunity for sex did not compensate for the loss of reproductive opportunities to premature death.
Another perspective is that, in the science ecosystem, Brooks is a top carnivore. His work makes far more money then any others, especially the people whose work he reports and other worker bees, and the writing and speaking has far more policy and public influence. Apparently the Brit government got the book for all departments or ministries — groan.
However, if Brooks didn’t exist we would have to invent him because, apparently, the demand for bromides sprinkled with pop science is growing. Brooks’ writing is not the problem — the demand for this kind of writing is. Actually, the use of these kinds of ideas for making policy/political decisions is the problem.
Once again, we are often injured more by our “friends” of science trying to “help” — for fun and personal profit — the the “enemies.”
It’s easier to dissect and poo-poo Brooks’ gaffs and journalism. It’s real hard to unpack and combat the hunger for pop sci. That’s the problem we’re interested in.
As a person who has been researching multilevel selection for over 20 years, I am saddened and dismayed by this dismissal of group selection. We have experimentally examined group selection in the lab to the point of having nothing very important left to study. It works, usually better than individual selection. We know why, the traditional models don’t take into account indirect genetic effects. We know the conditions when it works, as long as there is at least some structuring. We can measure it in the field. Arabidopsis Impatiens, imported willow leaf beetles, and water striders are known to be under group selection in natural populations. (there are others, but I would have to do some digging to find them). Group selection is now the accepted means of improving production in caged chicken populations, and increasingly in hog improvement. Yes, this morning you ate the products of group selection.
The bottom line is that group selection works, and is rampant in nature. Academics have the luxury of “not believing” in group selection and justifying their beliefs with 50 year old models. Nature doesn’t care, nor do people who actually have to produce a product. For agronomists failing to believe in group selection is money lost. Guess what, they don’t like losing money.
What is it specifically you feel you are disagreeing with here?
Inclusive fitness and multilevel selection are referenced in the text. There’s a vast and tedious literature involving people with various levels of mathematical training arguing about whether inclusive fitness or group selection/multilevel selection are isomorphic (and now, which one is *more* isomorphic).
Most biologists I’ve talked are getting royally tired of the non-debate (although it’s very popular among audiences of philosophers) – but there are still plenty of people arguing either side so I can understand you wanting to assert that there are *two* sides to the debate.
What seems unhelpfully disingenuous is presenting your position as if there is no controversy. If I’ve misread you, would you agree wholeheartedly with Stewart West’s paper “Social semantics: how useful has group selection been?”? If you do not, then evidently there is controversy and it is both unscholarly of you to misrepresent this – doubly so, given that you began your post by telling us how many years you’d been engaged in the subdiscipline.
This is not appropriate scholarly behaviour between scholars. It is *certainly* not appropriate behaviour when communicating with an audience most of which does not have the benefit of a 20 year familiarity with the literature.
The methods used by industrialists are not normally viewed as having bearing on scientific discourse because industrialists are not principally concerned with science, falsification or critical experiments. If tomorrow they decided it wasn’t group selection at all, but inclusive fitness, the pork would, I imagine, taste and sell pretty much exactly the same.
Marios
I was not specifically commenting on Stewart West’s work (with whom I tend to disagree in any case), but with this paragraph from the blog entry:
“Let’s be clear about what biologists really know about group selection and altruism. If true human altruism has a genetic basis, it is individually disadvantageous and could have evolved only by differential propagation of groups. That’s very unlikely, since it requires that the rate at which altruist-containing groups reproduce themselves must be high enough to counteract the substantial rate at which altruism genes disappear within groups. It’s unlikely because groups reproduce much less often than do individuals! Further, once a group consists entirely of altruists, any non-altruistic genes would rapidly invade it, as their carriers reap the benefits of altruism without sacrificing their reproduction.”
This is quite simply mis-informed. In particular, the fairly elegant laboratory studies of group selection have clearly established that group selection works because of genetically based interactions among individuals. Arguments that are based on a “gene for altruism” really should not be used in discussions about multilevel selection. The important point about the laboratory work is that in my opinion experimental data trumps theory. That is, when an experiment and theory come up with different results the experiment is probably the one that is correct, and it becomes the theorists job to discover how the theory needs to be modified to account for the experimental results. The bottom line with group selection is that it works so well that it shocked even the people that did the experiments. These experiments have been done repeatedly, and in EVERY case group selection is effective, and in most cases more effective than individual selection. We are thus in a position where we have to ask what is wrong with the models.
To be more specific, lets parse the paragraph I quoted a little:
“Let’s be clear about what biologists really know about group selection and altruism.”
This was what I was trying to comment on. It turns out that we know a great deal about group selection, but what we know does not agree with what was subsequently stated.
“If true human altruism has a genetic basis, it is individually disadvantageous and could have evolved only by differential propagation of groups. That’s very unlikely, since it requires that the rate at which altruist-containing groups reproduce themselves must be high enough to counteract the substantial rate at which altruism genes disappear within groups.”
Based on the 50 year old models that is true. Based on experiments that is not true. This is where experiments should trump theory. What is said here has been experimentally called into question. That needs to be addressed.
“Further, once a group consists entirely of altruists, any non-altruistic genes would rapidly invade it, as their carriers reap the benefits of altruism without sacrificing their reproduction.”
As one of my colleagues is fond of saying, “if it can evolve the first time, it can evolve the second time.” Again, experiments clearly show that group selected adaptations are robust to migration (as long as the group selection continues). This seriously calls this statement into question. By the way, this is not just a an experimental result, it is also a theoretical result!
I also take great umbrage on discounting agricultural work. There are reasons not to be happy with the methods used by them in their experiments, but there are some amazingly fine scientists and statisticians in their midst. To dismiss their findings is to discard some fine work. We should not forget that no less than Charles Darwin, Sir Ronald Fisher, and Sewall Wright relied heavily on the agricultural literature in developing their theories.
So, were humans group selected? I have no idea, but I do know that the arguments given in the blog entry have no information that I would consider useful in deciding whether or not that was the case.
“(group selection) works, usually better than individual selection.”
By what measure can one selective pressure ‘work better’ than another, and what does it mean to say that group selection ‘usually’ works better. ‘usually’ with respect to humans, to fish, mice or to all biological evolution? Are you telling me that if a mice was born blind that she would be alright because group selection was ‘usually better’ than individual selection?
“We know why, the traditional models don’t take into account indirect genetic effects.”
So the traditional models didn’t include things that they didn’t include. I’m not sure that this refutes anything.
“Academics have the luxury of “not believing” in group selection and justifying their beliefs with 50 year old models”
Are you attacking the academics right to disagree with you, or do you really have a problem with the old models? Are you saying that we should abandon the gene-centered view of evolution all together? Or is it just the emphasis on individual centered view of evolution? Both individual and group level selection models can be built using the gene centered view of evolution and, importantly, the approaches are isomorphic (they both fall under the larger class of inclusive fitness models). Should we abandon inclusive fitness models all together, and if so, what models natural selection better.
““If true human altruism has a genetic basis, it is individually disadvantageous and could have evolved only by differential propagation of groups.
That’s very unlikely, since it requires that the rate at which altruist-containing groups reproduce themselves must be high enough to counteract the
substantial rate at which altruism genes disappear within groups.”
Based on the 50 year old models that is true. Based on experiments that is not true. This is where experiments should trump theory. What is said here
has been experimentally called into question. That needs to be addressed.
I dissagree. Firstly, the quote is mathematically correct: the heritable information (the gene) that correlates with the trait (altruism) must have been positively statistically associated with survival (selected) by some mechanism. If it is not selected through direct forces (altruism isn’t by definition), then it must be selected through indirect selection. What other option is there? Secondly, what piece of evidence is there that refutes the claim above, or more generally, what piece of evidence *can* refute the claim? The claim is *not* a theory put forwards to explain the evolution of altruism, but a verbal version of a matyhematical model. How can a model be true or untrue if it is mathematically consistent? Again, are you are saying that we should abandon gene centered evolution models all together, and if so, then you need to say what is a better model of natural selection.
By what measure can one selective pressure ‘work better’ than another, and what does it mean to say that group selection ‘usually’ works better.
Umm, by the fact that every experiment in which both group selection and individual selection have been measured the group selection was more effective; by the fact that plant breeders abandoned individual selection decades ago and now practice “strain selection” exclusively; By the fact that Griffing showed both theoretically and experimentally that in interacting populations individual selection will frequently fail, but that group selection (he calls it strain selection) will always give a positive response.
Are you attacking the academics right to disagree with you, or do you really have a problem with the old models?
A little bit of both. I am saying that academics (of which I am one) have the luxury of not actually having to produce a product. As a consequence, they can choose not to agree with something, even if it works. Agronomists actually have to produce a product (better chickens, hogs and corn), and they will use what works regardless of what theory says. However, I am also disagreeing with the old models. The problem is that all models are an abstraction of the real world. There is nothing wrong with the models except that they are such an abstraction that they have abstracted away things that are of fundamental importance. The difference between individual and group selection is that group selection can act on indirect genetic effects. The traditional models assume that those don’t exist, and as a result to not provide useful insights into the efficacy of group selection.
How can a model be true or untrue if it is mathematically consistent?
I am just saying that in experimental studies group selection works in a way that is inconsistent with the models. They may be mathematically consistent, but if the underlying assumptions are incorrect they will not describe the real world. Experimental data tells us that the models are not describing the real world. Ergo, we need new models.
Secondly, what piece of evidence is there that refutes the claim above, or more generally, what piece of evidence *can* refute the claim?
I dissagree. Firstly, the quote is mathematically correct: the heritable information (the gene) that correlates with the trait (altruism) must have been positively statistically associated with survival (selected) by some mechanism.
As pointed out above, it may be mathematically correct, but it does not agree with experimental data. The fact is that we can set up situations in which traits are opposed by individual selection and favored by group selection, and in those cases group selection often, if not usually wins. The models may be mathematically correct, but they are not biologically correct.
I stumbled on this site looking for inflammatory quotes to use in my evolution course. I found what I was looking for, and used it. I felt as a public service I should comment on what was said. Somehow it did not dawn on me that this was Jerry’s blog. Years ago I promised myself never to respond to Jerry except in a peer reviewed article, I have inadvertently broken that promise. This will be my last post on this blog site, I will re-establish my promise to myself, and only discuss these issues with Jerry in a more formal setting.
I came to this discussion late (20 years late?) and have no idea who Jerry is. I just saw somebody putting out bile on the internet and thought that I would do some public good of my own and correct them. If you think that population structure is important, fine, but there is no reason to attack a straw man on a public forum, and then act surprised when somebody pulls you up on it.
I guess I’ll do your trick and say that I won’t discuss any more on this topic (but only after I’ve had the last say).