After a pleasant sojourn with ISIS in my last post, it’s time to get back to the question of whether religion is an evolved adaptation. In my last post on the evolution of religion, I mentioned that there was a brewing conflict between group selection and inclusive fitness models in biology. Did I say conflict? I meant outright, total war. Far be it from me to over-dramatize a scientific quarrel, but this one doesn’t need to be over-dramatized; it’s already plenty dramatic. From massive letters of protest signed by hundreds of biologists to name-calling to bald accusations of irrelevance leveled against major intellectual figures, the group selection/inclusive fitness debates are the major scientific conflict of the young 21st century. Grab some popcorn, okay?
To recap what’s at stake: in biology, the evolution of cooperation, or “altruism,” has presented a serious problem since the days of Darwin. Crucially, altruism doesn’t just mean one organism helping another. Instead, it specifically applies when one individual sacrifices some of its own biological fitness to boost another’s fitness. In Darwinian terms, “fitness” refers to the likelihood of leaving behind offspring. So, biologically speaking, one example of altruism might be when a bird sounds a warning call upon seeing a predator enter a clearing. The warning call could easily attract the predator’s attention, and so the warning endangers the bird’s own fitness: if it gets devoured, it’s sure not leaving behind any offspring. At the same time, the call alerts nearby birds to the predator’s presence, cueing them to fly away or hide – which increases their fitness.
Behaviors such as this – where one animal apparently puts its own fitness at risk while helping others – are commonplace across the animal kingdom, but they seem to conflict with the basic logic of Darwinian evolution. How can a behavior evolve if it puts the animals that exhibit it at a selective disadvantage?
As I discussed last time, group selection is one possibility. Group selection is the idea that altruism can evolve in a Darwinian framework through competition between groups. For example, one group of animals might all nicely cooperate with each other, while a nearby rival group might be populated by selfish bickerers who wouldn’t lift a finger (or wing, or flipper) to help one other. All else being equal, the first group would clearly be at an advantage in a competition, right? For example, groups of birds willing to warn each other about predators would outcompete groups of selfish birds who never gave warnings. In this way, genes for cooperation might spread through a population, as the groups populated by altruists conquered and dominated groups hobbled by non-cooperators.
So that’s one possibility for the evolution of altruism. Conversely, inclusive fitness is the idea, propounded by W.D. Hamilton in the 1960s and later championed by Richard Dawkins, that groups have nothing whatsoever to do with it – altruism can evolve only among organisms that are closely related to one another. Let’s go back to the forest clearing and its bird residents. Say that all the birds in the nearby woods are relatives of the warning bird. When it emits its warning cry, then, it’s potentially saving the lives of innumerable cousins, siblings, and offspring – each of which carries identical copies of its own genes. If enough copies of those genes are saved, then genes that code for warning calls can take hold and spread in the population. Yes, each time a courageous bird makes a warning call, it runs the risk of getting pounced on – but it’s also helping out lots of relatives who carry the same genes, and who now might live to see another day.
The key to Hamilton’s insight about inclusive fitness is that the cost paid by a self-sacrificing individual has to be made up for by how closely related it is to those who benefit from its sacrifice. As the biologist J.S. Haldane famously declared: “I’d jump in the river to save two brothers or eight cousins.” Haldane’s joke meant that, due to gene mixing in sexual reproduction, full siblings (as well as parents and offspring) carry one-half of one’s identical genetical material, while first cousins carry one-eighth. Grandparents and grandchildren, meanwhile, share one-quarter of each other’s genes. In a mathematically ideal world, then, grandparents should be twice as willing to sacrifice for their grandchildren as for their cousins, parents should be roughly twice as likely to self-sacrifice for children as for their own grandparents, and everyone should be more likely to sacrifice for cousins than for total strangers. Hamilton codified this set of predictions in a famous mathematical rule, rB > c, where B is the benefit of an altruistic or sacrificial act, r is the relatedness between the benefactor and the beneficiary, and c is the cost of the altruistic action.* Easy enough, right?
But at first, Hamilton’s equation – and his two seminal papers, published in 1964 – were ignored. Ironically, it was entomologist and public science writer E.O. Wilson who first brought Hamilton’s ideas to the attention of evolutionary scientists. In 1965, as a young professor on a train ride from Boston to Miami, Wilson happened to glance through Hamilton’s two famous articles – by now a year old – that laid out the theory of inclusive fitness. By the time Wilson arrived in Miami, he was convinced that Hamilton was correct – and he quickly used his growing influence in the scientific community to bring everyone else into the fold. From the 1970s on, inclusive fitness was king of the roost, helping biologists and evolutionary theorists make sense of a dazzling galaxy of seemingly altruistic and cooperative behaviors, from bird warning calls to sterile worker bees to human helpfulness.
Of course, one key effect of Hamilton’s victory was that biologists abandoned group selection theories en masse. Inclusive fitness, everyone agreed, made much more biological sense than unrelated groups of animals sacrificing for each other. Mathematical treatments showed that group selection might be possible in theory, but that in the real world competition between individuals would always overwhelm any cooperative efforts at the level of the group. In other words, no amount of competition between groups could ever compensate for the fact that selfishness would win competitions within groups. Group selection was a non-starter for the evolution of altruism. It became accepted wisdom that altruism can only evolve by benefitting close relatives – not social groups of unrelated peers.
I said that E.O. Wilson’s role in spreading the doctrine of inclusive fitness was ironic because now, in the 2000s and 2010s, Wilson is shaking up the scientific establishment by utterly, totally recanting his former support for Hamilton’s theory. Like, big-time. In 2007, Wilson published an infamous paper in The Quarterly Review of Biology with evolutionary biologist David Sloan Wilson (no genetic relation), claiming that inclusive fitness models were inferior to group selection theories, and that evolutionary biology made a huge mistake in abandoning group selection. In 2010, Wilson co-authored an even more infamous paper in Nature with mathematical biologists Martin Nowak and Corina Tarnita, questioning the very mathematical foundations of inclusive fitness.
This paper aroused an astounding response from the community of evolutionary biologists. More than 130 scientists co-signed a blistering response, published in a subsequent issue of Nature, in which they repudiated Nowak, Tarnita, and Wilson’s claims and accused them of “misunderstanding evolutionary theory.” Make no mistake – them’s fighting words, especially when you’re talking about E.O. Wilson, one of the most influential and respected biologists of the last hundred years. You don’t accuse a towering figure in your field of misunderstanding the field’s most basic concepts unless you’re completely convinced he’s gone off the deep end. The bellicose shock with which many evolutionary biologists responded to the 2010 Nature paper made it clear that – as Richard Dawkins expressed in his pan of E.O. Wilson’s most recent book – this is exactly what many of them think.
However, the 2010 paper seems to have had some of the effect that Nowak, Tarnita, and Wilson were hoping for. In the five years since its publication, debates about group selection and inclusive fitness have become more commonplace. For instance, in 2012 a large group of scientists went toe-to-toe in an Edge.org discussion entitled “The False Allure of Group Selection,” hammering on each other’s ideas in full view of the educated lay public. More recently, a flurry of articles in PLOS Biology has drilled down on the relevant mathematics. Nobody is really changing anyone else’s mind – yet – but the debate has gone from a moribund, back-burner simmer to a full roiling boil in the past half-decade. Things are, at least, exciting.
This series was supposed to be about the evolution of religion. It still is. It’s just that these rollicking debates about evolutionary theory are crucial for understanding cutting-edge theories about religion’s evolution, because many of the most fertile hypotheses about how religious behavior evolved are directly related to questions of group cooperation. The mechanisms by which cooperation and altruism evolve in animals – whether kin selection or group (or “multilevel”) selection – make a difference for how humans behave, because humans are animals. Studying religion in humans is, therefore, analogous to studying birdsong in birds – it’s just something humans do. So in order to understand religion properly, we need to get our biology straight.
Which is the reason this series is running to several parts. Next time, we’ll investigate how and why mechanisms for the evolution of cooperation bear on religion. This will give me an opportunity to skewer Steven Pinker, an activity that has proved productive and enjoyable in the past – although I don’t disagree with him on group selection nearly as strongly as I do on religion, and think his contributions to the group selection debates are useful. We’ll also investigate whether humans are basically good or basically evil (answer: yes). Suggestions for particular points to highlight are welcome in the comments.
So, here’s the thing. Posts will be more infrequent than usual for the rest of the summer, as I cram to finish my dissertation. Thanks for your patience. Care packages, including coffee and motivational drawings, are welcomed.
* This prediction has been well-corroborated in humans, by the way. It may seem commonsensical that we’re more willing to help closer relatives than our more distant ones, or strangers, but if you forget common sense for a moment (often a useful habit in science), you’ll find yourself wondering why, really, we’re more self-sacrificial toward kin than toward non-kin. Hamilton’s rule is one answer.