Before Darwin, the only problem with altruism was that there wasn’t enough of it in the world. After Darwin, altruism emerged as a genuine scientific problem. If animals, including human beings, evolved by natural selection—a merciless process in which organism struggles against organism and all that matters is outcompeting everyone else—how could altruism arise? How could natural selection promote, or even allow, behavior that is costly to the individual that performs it but that benefits someone else?
This is the problem at the heart of Oren Harman’s new book, The Price of Altruism. The punning title refers to the protagonist of his book, George Price, a brilliant figure who performed fundamental work in the 1970s on several problems in evolutionary biology, including altruism. (Price committed suicide in 1975.) Harman is a professor at Bar Ilan University in Israel, where he chairs the graduate program in science, technology, and society. He is the author of The Man Who Invented the Chromosome (2004), an intellectual biography of the botanist Cyril Darlington. In both that book and the present one, Harman takes up the tale of a scientist who is largely unknown to the general public but admired among scientific specialists. The subject of The Price of Altruism provides particularly rich material, both scientifically and biographically, for such a project. Price’s work in evolutionary biology, while abstract, was breathtakingly original and deep. And Price the man makes for a fascinating, if often disturbing and ultimately tragic, story.
Though Harman’s writing is a bit bumpy in places and the book sometimes feels too self-consciously in the tradition of A Beautiful Mind, it is, on the whole, a remarkable achievement. Harman has done his homework. He not only documents the details of Price’s scientific efforts but also mines his personal correspondence as well as the recollections of his associates and family. Harman also skillfully handles some of the most abstract ideas to emerge from evolutionary biology, including the notoriously subtle “Price Equation.”1 Although I will argue that Harman goes overboard in a few places, there can be little doubt that his book represents a major contribution to our understanding of an important but neglected scientist.
The Price of Altruism does much more than tell the story of George Price and his science. It also canvasses the entire history of biological attempts to explain the origins of altruism, thereby placing Price’s work in its intellectual context. As Harman emphasizes, the history of these attempts is long and complex. Indeed it begins with Darwin himself. Darwin recognized that altruism posed a potentially fatal challenge to his theory of natural selection. He also saw that, while we often equate altruism with acts of human kindness or sacrifice—the saint who surrenders his belongings or the soldier who throws himself on the grenade—some of the clearest examples can be found in other species, where there can be little doubt the behavior is encoded genetically, not encouraged culturally. (If altruism or a disposition to altruism has a genetic basis, it must have evolved in some way.)
One of the most striking examples of animal altruism, and one that troubled Darwin to no end, occurs among the insects. In many species of wasps, bees, and ants, entire castes of individuals are sterile and all reproduction occurs through the queen. If it’s hard to see how natural selection could drive a worker bee to toil for the good of the colony (not herself), it’s next to impossible to see how natural selection could render whole collections of such individuals sterile. To sacrifice one’s fertility would seem, under Darwinism, a sacrifice too far. Though Darwin offered tentative explanations of this phenomenon, they were less than entirely satisfying.
Darwin’s great champion T.H. Huxley also agonized over altruism, though mostly with respect to human morality. Huxley concluded that it was romantic folly to expect kindness to emerge from the logic of Darwinism, a logic that is selfish through and through. Our challenge, he concluded, is not to seek morality in nature but to rise above nature. As Harman puts it, this is “the task of civilization—its very raison d’être: to combat, with full force, man’s evolutionary heritage.” Others saw things differently. The Russian zoologist Peter Kropotkin, for example, found an astonishing amount of altruism in nature. His book Mutual Aid: A Factor in Evolution (1902) documented many cases of cooperation, not competition, among animals in the wild. Pelicans, for example, corner fish by collectively forming an arc in the water and then swimming to shore. Contra Huxley, therefore, Kropotkin concluded that our moral challenge is to navigate a return to nature, not to rise above it. Kropotkin, an aristocrat turned anarchist, not surprisingly believed that this return would require radical economic change.
In the twentieth century, two big developments marked attempts to understand the biology of altruism. In the first, V.C. Wynne-Edwards, author of Animal Dispersion in Relation to Social Behavior (1962), argued that evolutionary biologists had partly misconstrued the biological level at which natural selection acts. Though selection might often involve competition among individual organisms, Wynne-Edwards maintained that it often involves competition among entire populations of organisms. Under this so-called group selection view, altruism seems easily explained. Within a group, selfish individuals (ones who, say, reproduce rapidly even when resources are limited) might outcompete altruistic individuals (ones who slow reproduction when times are tough). But among populations, things are different. A population comprised of selfish individuals will likely ultimately deplete its resources and crash to extinction, whereas a population comprised of altruists will not deplete its resources and will survive to another day. Natural selection acting at the group level might therefore sustain altruism in animals. Though many biologists were skeptical of group selection—especially as the models offered were mostly verbal, not mathematical—no fully convincing alternative seemed available.
Then in the mid-1960s, the second big development came along: W.D. Hamilton’s theory of inclusive fitness. Hamilton, who penned famously difficult but profound papers, argued that what matters under natural selection is not the fitness of this or that individual (where fitness is measured roughly by the number of progeny an individual produces) but the fitness of an individual averaged together, in a certain way, with the fitness of its relatives. Inclusive fitness could explain all manner of previously puzzling acts of animal altruism.
Consider, for example, alarm calls in some species of squirrels. If a predator appears, the squirrel that first sees it may emit a call that warns other squirrels, which can then flee to safety. The behavior is presumably altruistic as it draws the predator’s attention to the calling squirrel, making of him a likely lunch.2 To see how such behavior could evolve, imagine an initially rare genetic variant that encourages an individual to sound the altruistic call. (Not all individuals who carry such a gene will always sound the alarm when seeing a predator; they need only show a predisposition to sound the call.) Hamilton showed mathematically that natural selection will, under a certain condition, increase the frequency of such an altruistic gene within a population. And the critical condition is simple: the individuals saved by the act of altruism must be related to the altruist. Why? Because such relatives also likely carry the gene for altruism. (That’s what it means to be relatives—you tend to carry the same genes.) So while an altruist might sacrifice himself, he saves—by that very act—other copies of the altruism gene, ensuring the gene’s spread through the population.
Hamilton’s theory of inclusive fitness, which shifted the focus of natural selection from the individual to the gene, offered a bold new account of altruism and, by the 1970s, was becoming evolutionary orthodoxy. Group selection arguments appeared increasingly extravagant and, thanks to the efforts of Hamilton and other prominent evolutionists, they mostly fell into disrepute. As Harman stresses, George Price would help change that.
Price, born in New York in 1922, graduated from Stuyvesant High School and did his undergraduate studies at Harvard and the University of Chicago. He remained at Chicago for doctoral work in chemistry, performing impressive research on radiation for the Manhattan Project. After a period as an instructor at Harvard—Price found he hated academic life—he began the intellectual wandering that would characterize his career. Brief stints at several institutions ensued, including Bell Labs, the University of Minnesota Medical School, and IBM. During these years, Price married and had two daughters.
To the surprise of those around him, Price turned his attention in the mid-1950s to writing popular science. He made an unexpectedly big splash with a provocative piece in Science that assailed the credibility of research on extrasensory perception. (A staunch atheist, Price despised anything that smacked of the supernatural.) In subsequent articles for the popular press, he ridiculed bad economics, explained how to speed up the process of invention, and railed about the Communist threat in America. Beginning to call himself an ex-chemist, Price was an intellect without intellectual focus.
His marriage faltering, Price abandoned his family and, in 1967, relocated to London. There he planned to pursue his newest interest, evolutionary biology. Ironically, Price announced that he was most interested in the evolutionary origins of the family: Why do acts of altruism characterize interactions among relatives and why would natural selection favor such acts? Price took a theoretical approach to such questions and began to focus on the mathematics of natural selection. He initially worked alone in libraries throughout London but eventually secured a temporary and modest position in the Galton Laboratory at University College London. During his time there, he published only a handful of papers on evolution. Most of these publications attracted little immediate attention among biologists, partly reflecting Price’s obscurity and partly reflecting his unorthodox theoretical approach. Fortunately, however, Price had struck up a correspondence with Hamilton who, together with the prominent evolutionist John Maynard Smith, began to appreciate the significance of Price’s work. His ultimate reputation among evolutionary biologists was thereby assured.
True to form, Price soon started losing interest in evolution. Suddenly, and to the consternation of his small circle of colleagues, he discovered, through an apparent religious experience, a new passion: Christianity. Price’s interest in Christianity was partly intellectual, partly spiritual, and wholly radical. To Price, the Bible was a text that, while suffused with codes and ciphers, was mostly meant literally. (His biblical literalism yielded a surprising correspondent: Henry Morris, a young-earth fundamentalist and leader of the scientific creationist movement. Astonishingly, Price concluded that Morris’s brand of Christianity was too liberal.) At his London church, Price’s mentors urged him to slow down and to temper his literalism. Predictably, he ignored their pleas and began to write a long document that revealed, he claimed, errors in the traditional chronology of the Passion.
As Harman emphasizes, Price’s radical Christianity had an even more profound, and disturbing, effect on his personal life. Ultimately feeling called to surrender his possessions to the poor, he began ministering to, and often sheltering, the homeless of London, many of whom were alcoholic. Price himself quickly descended into poverty.3 He also seems to have begun to suffer from increasingly serious psychological problems. (Harman speculates that Price may have been mildly autistic and that, later in life, he suffered depression, possibly caused by failure to take prescribed medication for a thyroid condition.)
As his personal life disintegrated further, Price finally concluded that he should attempt a more conventionally Christian life. After all, he reluctantly conceded, his extraordinary experiment in altruism had failed to save a single soul. He also decided it was time to move on to his newest intellectual interest, economics. But it was too late. In 1975, Price, fifty-two and a squatter in a derelict London building, succumbed to his many difficulties and committed a gruesome suicide, slitting his throat with a pair of tailor’s scissors. His funeral was an unusual gathering, drawing only a few homeless people and two of the world’s greatest evolutionary biologists, Hamilton and Maynard Smith.
Just what did this tragic figure contribute to the study of evolution? For one thing (and this alone would have secured his reputation), Price, along with Maynard Smith, helped to introduce game theory into biology. Game theory lets biologists answer questions like: Under what conditions should an animal escalate or deescalate its aggression against another animal? Price also clarified a controversial result in population genetics called the Fundamental Theorem of Natural Selection. But his key contribution, and the focus of Harman’s book, is a mathematical result now known as the Price Equation. Though somewhat mystifying when first encountered, the Price Equation is actually fairly straightforward.4
To see the gist of the equation, it helps to begin with a simplified version, one that was discovered independently by three evolutionists, including Price, between 1966 and 1970. Their finding shows that the amount by which any biological trait, e.g., height, changes from one generation to the next by natural selection equals the statistical association between the trait and fitness.5 For example, if there is a strong positive relationship between height and fitness—with tall individuals producing more progeny than shorter individuals—the average individual in the next generation will be taller than the average individual in the current generation and by an amount that can be predicted by the equation. This result, now usually called the Secondary Theorem of Natural Selection, is so simple that it remains remarkable it was seen so late in the history of evolutionary genetics.
Price’s full equation—the Price Equation—goes further. The simplified version just described assumes that natural selection (acting at some biological level) is the only force that changes a trait. But this needn’t be true. Traits might change for all kinds of reasons, including, for instance, a new genetic mutation. The full Price Equation, published in 1970, takes these possibilities into account by including an additional mathematical term, usually called the “transmission” term.
Now comes the deep part. Price saw—and it is far from obvious when put into words—that this transmission term can be rewritten mathematically in a way that allows for natural selection at a different biological level. So, to take a simple example, the evolution of height might reflect natural selection between individuals within populations and between populations of individuals. In fact, this mathematical maneuver can be performed recursively, over multiple levels of the biological hierarchy. In the end, the Price Equation allows one to partition evolutionary change in a trait into the possible effects of natural selection acting simultaneously at multiple biological levels, e.g., the species, population, organism, gamete (sperm or egg), and gene.
The result is so simple but surprising that Price deemed it “quite a miracle.” And Hamilton, the great champion of the gene’s-eye view of evolution, was completely won over, reporting to the still-obscure Price that “I am enchanted with your formula.” Indeed Hamilton began to see that his own inclusive fitness approach to altruism might be equivalent mathematically to Price’s approach. Though biologists were slow to appreciate it, Price had discovered a deep truth about natural selection.
There is, of course, nothing special about a trait (like height) that characterizes how an organism looks. One can also use the Price Equation to study animal behavior, including altruistic behavior. Doing so reveals that accounting for altruism biologically does not require one to choose between believing that natural selection acts within groups or between groups. Instead, selection might act simultaneously at both levels, with selection within groups favoring selfishness (individuals that are selfish will out-compete those that are saintly) and selection between groups favoring altruism (groups including many cooperative individuals will do better, as a group, than those including many uncooperative individuals). Whether or not altruism evolves depends on the relative strengths of these conflicting forms of selection.
Crucially, the Price Equation showed that group selection arguments for altruism need not be offered tentatively or apologetically; they could be offered with mathematical precision.
Harman’s story, both biographical and scientific, is fascinating and he generally tells it well. His account does, though, suffer from a few problems. One is that he seems to mislocate slightly what is most interesting about the Price Equation. While it’s true that the equation is central to the study of altruism, it is, as I hinted earlier, broader than that. Because the equation applies to any trait, it is interesting more for what it says about natural selection in general—for instance, that it might act at multiple biological levels—than for what it says about altruism in particular.
The Price of Altruism also suffers from some exaggeration. One instance involves the role of the Price Equation in contemporary evolutionary biology. It would be easy to walk away from Harman’s book believing that Price and his equation are central to the field. This is not the case. While the Price Equation is undeniably important, I doubt that the average evolutionary biologist can tell you what it says. This reflects two facts, one about evolutionary biology and the other about the equation. The fact about evolutionary biology is that it has, for the last generation, been consumed with data, not theory. The flood of whole genome sequence data, in particular, has commanded the attention of many evolutionists and the last thing on their minds is abstract equations that look more like set theory than DNA sequences. The fact about the Price Equation is that, in a curious way, it doesn’t answer biological questions. Instead, as Steven Frank has emphasized, the Price Equation simply features an unusual system of notation that forces us to think more clearly about natural selection. But there’s a difference between thinking clearly and getting concrete answers to specific biological questions.
This brings me to another instance of exaggeration in the The Price of Altruism. Harman suggests that the Price Equation “solved the problem of altruism.” But such language masks a deeper point. No equation, however elegant, can solve the problem of altruism because the problem, like most in science, is partly empirical. The Price Equation shows that evolutionary change in a trait can be partitioned into the effects of natural selection at different biological levels. But this tells us nothing about the sizes of the effects at these levels. Some (say, at the individual level) might be enormous while others (say, at the group level) might be minuscule. If so, individual selection would overwhelm group selection and altruism wouldn’t evolve. No mathematical formalism can determine these critical weightings. Harman ultimately acknowledges all this, but the acknowledgment arrives late in his book.
Harman may also overestimate the extent to which evolutionary biologists have warmed up to group selection. It is true that group selection is sometimes invoked to explain biological phenomena. But it is more true that these invocations are rare. The reasons are again partly empirical. Particular biological phenomena have been explained successfully far more often by arguments involving inclusive fitness than by those involving group selection. To many evolutionists, then, hesitation about group selection isn’t so much principled as practical. Group selection just doesn’t have as impressive a record of explanation.6
It would be unreasonable, however, to suggest that these few shortcomings detract in any significant way from The Price of Altruism. Harman does many things right in his book. One of the most important—indeed more important than the above concerns—is his treatment of human altruism. What, after all, does this recondite biology have to do with our own species?
Books on the biological basis of behavior often go overboard when it comes to human beings. In too many such books, it seems to follow—and absurdly easily—that if worker bees surrender their fertility at the command of genes, the boy scout who escorts the elderly lady across the street also moves in response to DNA. Harman’s book is refreshingly free of such nonsense. Happily, it is also free of the opposite kind of nonsense, that which maintains that our behavior is utterly detached from biology.
Harman’s view, expressed most clearly near the close of his book, is nuanced, sensible, and almost surely right: the truth about human behavior, including human moral behavior, lies somewhere in the middle. It would, after all, be remarkable if our hard-won cultural inheritance, including our religious traditions, had no bearing on our moral judgments or inclinations to act altruistically. But it would be equally miraculous if our biological origins left no mark on our thought and behavior. The boy scout might move partly in response to his culture (that’s why parents push their children into scouting, the synagogue, or church groups) and partly in response to his genes (after all, one can imagine a species that, in contrast to human beings, is so innately vicious that the scout’s behavior is unthinkable).
It is not scientism to claim that biology matters, though it might well be scientism to claim that it is all that matters. Harman appreciates the distinction and resists the temptation to make human beings, and human altruism, simpler than they are.
October 14, 2010
Harman does stumble now and then on technical matters, e.g., the so-called probability of fixation of a new beneficial mutation, but these slips are of little significance to his story. ↩
This example is not universally accepted among biologists; but it captures the logic of an inclusive fitness argument. ↩
Harman speculates that Price’s life of selflessness was connected to his (troubling) scientific finding that altruism in nature invariably reflects an agenda: some entity, whether the species, population, or individual, must gain if altruism is to evolve. Perhaps, Harman suggests, Price set out to show, through his life, that pure altruism is possible. ↩
Harman wisely removes most of the mathematics to an appendix. Those interested can find a more technical discussion in his Appendix 2 or in Steven Frank’s classic rederivation of the Price Equation (Steven A. Frank, “George Price’s Contributions to Evolutionary Genetics,” Journal of Theoretical Biology, Vol. 175, No. 3, August 7, 1995). ↩
The statistical association between the trait and fitness is measured by a quantity called the covariance. ↩
Some biologists also emphasize theoretical reasons for preferring inclusive fitness approaches. Formally, models of inclusive fitness and of group (or, more exactly, multilevel) selection appear equivalent. But the former are often simpler mathematically than the latter. This seems to me a partly, though not wholly, persuasive reason for preferring inclusive fitness models. Surely the fact that the two kinds of model are equivalent is more significant than that one seems simpler to us. These issues, both theoretical and empirical, are considered in far greater detail in a set of lectures hosted by the London Evolutionary Research Network; video recordings of the lectures are available at vimeo.com/user2795743. Also see Elliott Sober and David Sloan Wilson’s Unto Others (Harvard University Press, 1998) for a vigorous defense of group selection and its relevance to altruism. ↩