How We Evolve

Feature by Benjamin Phelan / October 7, 2008

A growing number of scientists argue that human culture itself has become the foremost agent of biological change.

When the previous generation of life scientists was coming up through the academy, there was a widespread assumption, not always articulated by professors, that human evolution had all but stopped. It had certainly shaped our prehuman ancestors — Australopithecus, Paranthropus, and the rest of the ape-men and man-apes in our bushy lineage — but once Homo sapiens developed agriculture and language, it was thought, we stopped changing. It was as though, having achieved its aim by the seventh day, evolution rested. “That was the stereotype that I learned,” says population geneticist and anthropologist Henry Harpending. “We showed up 45,000 years ago and haven’t changed since then.”

The idea makes a rough-and-ready kind of sense. Natural selection derives its power to transform from the survival of some and the demise of others, and from differential reproductive success. But we nurse our sick back to health, and mating is no longer a privilege that males beat each other senseless to secure. As a result, even the less fit get to pass on their genes. Promiscuity and sperm competition have given way to spiritual love; the fittest and the unfit are treated as equals, and equally flourish. With the advent of culture and our fine sensibilities, the assumption was, natural selection went by the board.

Moreover, evolution had never been observed in humans, except in a few odd cases, so the conclusion was drawn that it wasn’t happening. One can’t fault the logic. The most famous case of adaptive change in humans, that of sickle cell trait as an evolutionary response to malaria, seemed to prove the point that human evolution must be rare: Even in as dire and malaria-stricken an environment as West Africa, the only response evolution has been able to come up with is an imperfect defense that can cause serious health problems along with its solitary benefit. Selection pressures as strong as those brought about by endemic malaria are uncommon, and civilization was thought to wash out those less powerful.

But since the turn of the millennium, genomics has undergone a revolution. With the completion of such landmark studies as the Human Genome Project and the publication of HapMap, scientists finally have access to the particles of evolution. They can inspect vast stretches of DNA from people of all ethnicities, and the colossal amount of information suddenly available has spurred a revision of the old static picture that will render it unrecognizable. Harpending and a host of researchers have discovered in our DNA evidence that culture, far from halting evolution, appears to accelerate it.

John Hawks started out as a “fossil guy” studying under Milford Wolpoff, a paleoanthropologist who is the leading proponent of the faintly heretical multiregional theory of human evolution. Coming to genetics from such a background has perhaps given Hawks the stomach to wield unfashionable hypotheses. In December of last year, he, Harpending, and others published a paper whose central finding, that evolution in humans is observable and accelerating, would have been nonsensical to many geneticists 20 years ago. Up to 10 percent of the human genome appears to be evolving at the maximum rate, more quickly than ever before in human history.

“Seven percent is a minimum,” Hawks says. “It’s an amazing number,” and one that is difficult to square with the prevailing view of natural selection’s power. Because most mutations have a neutral effect on their carriers, making them neither fitter nor less fit, neither more fertile nor sterile, only slightly different, those changes are invisible to natural selection. They spread or don’t spread through a population by chance, in a process called genetic drift, which is often thought of as the agent of more change than natural selection. But the changes that Hawks detected, if he is correct, are too consistent from person to person, from nationality to nationality, to have been caused by genetic drift alone.

By looking at the data from HapMap, a massive survey of the genetic differences between selected populations from around the world, Hawks identified gene variants, or alleles, that were present in many people’s DNA, but not in everyone’s. These alleles seemed to be moving, over time, through populations in a way that matched mathematical predictions of what natural selection should look like on the genomic level. And though Hawks doesn’t know why possession of the new alleles should be advantageous, he doesn’t need to know. The signature that natural selection inscribes on the genome is legible even when the import of the message is unclear.

Fixation of the allele that allows European adults to digest lactose is nearly complete, after arising some 8,000 years ago and undergoing strong positive selection. Following the curve predicted by R.A. Fisher, the allele spread rapidly long after it first appeared. Data for the spread of different lactose-digestion alleles in Africa is incomplete, but what is available suggests the gene is far from fixed there.

HapMap can reveal where natural selection has occurred thanks to the tendency of DNA “neighborhoods” to be inherited in blocks that do not change much, if at all, from parent to offspring. When an organism reproduces, adjacent DNA sticks together and is passed on as a unit to the offspring. Such sections of linked DNA are called haplotypes; HapMap is a directory of them. A given haplotype is nearly identical among family members, but populations that have had recent contact with each other, such as the French and the Spanish, or the Cherokee and the Inuit, also tend to share it.

One of the characteristics of this linkage is that it is strong over short distances on a chromosome and weak over long distances. This is because mutations are rare but equally likely at every location, so they happen less often in a small region of DNA than in a large region. Over many generations, mutations nibble away at the edges of haplotypes and poke holes in their interiors, and the routine reshuffling of nucleotides, called recombination, can move linked sections of DNA far from one another, thereby breaking the linkage. Thus, the length of a haplotype roughly indicates its age, as does the amount of variation within it. Since mutations and recombination occur at a predictable rate, by comparing haplotypes from two populations, one can determine their degree of relatedness and thus estimate how long ago they diverged. So, for example, the San of southern Africa and the Han Chinese, would tend not to share haplotypes because their populations diverged long ago, and those they did share would be short or contain a great deal of variation.

Because haplotypes are similar from population to population, differences are easy to spot; a variation on a familiar haplotype background is like a smear of red paint on a white wall. If one looks at a haplotype in 100 individuals, and 90 of them are identical but 10 show the exact same variation, the odds are vanishingly small that random processes generated the same mutation 10 different times. Such a site is a candidate for one undergoing natural selection, because only a mutation that confers some kind of advantage will be propagated reliably through the population.

If the trait under selection produces a significant enough adaptive advantage, the allele responsible for the trait will rise in frequency so quickly that it will drag a long haplotype along with it before recombination and mutation can break it down to a short haplotype. So a rare allele on a long haplotype is an indication of strong and recent selection.

Hawks’s analysis of the HapMap data yielded many such candidate sites, but some of his colleagues were unimpressed. “They didn’t like the idea,” he says. An anonymous reviewer of his paper claimed not to think that natural selection could possibly be important in recent evolution, “so much so that they said positive selection happens rarely, if ever.”

An oft-cited example of evolution in historic times is the spread of the mutation that allows humans to digest milk in adulthood. It seems to have arisen around 8,000 years ago and has since spread to all parts of the world, though there are still plenty of us without it: One in 50 Swedes and nine out of 10 Asian Americans lack the mutation. The lactose intolerant are, at least in this respect, as the first Homo sapiens were.

“There are five versions of the lactase drinking gene, so five different populations have mutations that let them drink milk,” says Hawks. Because of this, many mutations conferring the same benefit are unlikely to have become common by genetic drift, but Hawks knows of practicing geneticists who find the idea that natural selection was the agent of their propagation to be preposterous. He’s incredulous at what he sees as such scientists’ fundamental misprision of the field’s core principle. “This is Darwin’s field,” he says. “Darwin talks about evolution, and Darwinism is about natural selection. But these people don’t believe in natural selection — except way back when, when chimps and humans were the same.”

By Hawks’s own description, his research “depends on a view of evolution that’s dominated by natural selection. When I look at the evolutionary processes leading to humans I’m thinking, what’s the adaptive change that’s happening? What are the constraints on our adaptation?” One of those, he says, “is demography.”

By invoking ancient demography via the anthropological record, Hawks believes he has identified what has been driving all the adaptive evolution he detected: an explosion in the global human population roughly coincident with the agricultural revolution of some 10,000 years ago. We invented agriculture, started eating different food, and began dwelling in cities. Our numbers swelled, our world changed, and our DNA is still catching up.

Spencer Wells, director of the Genographic Project, an attempt to reconstruct human migration patterns by sampling DNA from the world’s populations, has studied humanity’s transition to agriculture extensively. Hawks’s result was no surprise to him.

“The biggest change in our lifestyle as a species has happened in the past 10,000 years,” Wells says. “We spent most of the past million or so years of evolution living as hunter-gatherers, hunting game on the African savannas, or gathering shellfish on the coast, gradually moving out to Eurasia. Then, suddenly, in the past 10,000 years, we become a species that settles down. The diversity of food sources drops precipitously from over 100 in the hunter-gatherer diet to fewer than 10 in the average agricultural diet. And then, of course, you build up the population densities and disease takes off.”

Such changes to environment, diet, and disease load are classic agents of natural selection. The three acting in concert could certainly accelerate evolution. But it might seem odd that a larger population is required to produce a faster rate of evolution, especially if you happen to be American.

Scientists have finally been granted access to the particles of evolution.

The early and mid-20th century witnessed a tension between two interpretations of evolutionary theory. Sewall Wright, an American, argued that for rapid evolution to occur, what was required was a small, semi-isolated population through which a mutation could spread quickly, even by genetic drift. Thereafter, that population could migrate and spread the allele in other populations. R.A. Fisher, a Brit, argued that, in fact, a large population was required, because only a large population can produce large numbers of mutations. Because most mutations are neutral, he reasoned, it takes a large number of mutations to produce one beneficial allele. American biologists were most influenced by Wright, but Fisher’s work is where Hawks and Harpending find their support.

Fisher developed a mathematical model of how beneficial mutations should move through a population toward fixation, the point at which all members of a species have the allele. The shape of the curve is characterized by slow dispersion at first, because the mutation initially exists in only one member of the species. It takes a long time for a new allele to reach an appreciable frequency in a population, but at a certain point the growth rate becomes much steeper; many carriers bear many offspring, and the gene becomes widespread. But during the last leg of the push toward fixation, the rate decreases and begins to resemble a curve approaching an asymptote.

When anthropologists analyzed caches of ancient Eurasian skeletons, they found evidence that Fisher’s model was correct. In the DNA of a group of 5,000-year-old skeletons from Germany, they discovered no trace of the lactase allele, even though it had originated a good 3,000 years beforehand. Similar tests done on 3,000-year-old skeletons from Ukraine showed a 30 percent frequency of the allele. In the modern populations of both locales, the frequency is around 90 percent.

“This is the curve that Fisher predicted,” says Hawks. “The frequency [of the lactase allele] that we have at different times fits this curve. This means that the maximum rate of change in frequency of this gene was within the past 3,000 years, even though the gene originated 8,000 years ago.”

Seeing the mathematical model he was using borne out in data other than his own was encouraging to Hawks: Many of the alleles he’d identified as being under selection seem to show a similar trajectory toward fixation.

“My attitude about recent human evolution comes straight out of mathematics,” he says. “I can say, this is population growth, and these are the effects it should have. And as long as I keep observing data that’s consistent with that idea, I think it’s a strong model…. Once you can connect history with genes, you can build up knowledge from the standpoint of anthropology, then let the biochemists work out what each gene does.”

Being able to understand the purpose of a given gene, however, is perhaps the main challenge facing the current generation. Hawks doesn’t know what function the genes he identified as evolving perform, but such information isn’t important for his purposes. He is content with linking demographic history with mathematics and gene surveys and hypothesizing natural selection based on the confluence of those streams of evidence. A biochemist, though, might balk at saying that a gene is under selection without knowing what the gene actually does.

“Human genetics made a major leap forward at the turn of the millennium,” says Pardis Sabeti, an evolutionary geneticist at MIT’s Broad Institute who has done a great deal of work on methods for assessing genomic surveys like HapMap, the first draft of which was published in 2005. HapMap is a leaner and in some ways more powerful version of the Human Genome Project, as it compiles only those regions of the human genome — less than 1 percent — that have the potential to differ from person to person. In comparing different populations’ genetic information, it’s possible to tease out patterns of gene inheritance, how certain genes correlate with certain diseases, and even the likely geographic origin of some mutations.

One of the methods that Sabeti has developed to identify selection is to search for rare alleles on long haplotypes, which is useful for identifying selection in the past 30,000 years or so. Using the long haplotype test on HapMap data, Sabeti was able to find what appears to be a signature for recent natural selection on genes that are associated with resistance to lassa, a hemorrhagic fever that’s endemic to parts of central and western Africa. She is perhaps more cautious than Hawks in her conclusions, though; they are in different fields and have different standards of proof.

“I’m a little guarded on the findings for lassa, because the question is, is the finding real?” she says. “The strongest signal of selection we’ve detected in a West African population is on a gene called Large, which has been biologically linked to lassa.” Lassa is a poorly understood and infrequently studied pathogen, she says, so there was not much literature to consult about genes possibly associated with it. However, a microbiologist named Stefan Kunz had demonstrated that if Large is deleted from a mouse’s DNA, lassa is unable to infect it.

“That was exciting, because otherwise we’d look at the gene and say ‘I don’t know what it does,’ and that would have been the end of it. But now we could see a link,” she says. “But when you look at selection you never believe your results completely because it’s circumstantial. We have basic evidence that it seems to be evolving and we can link it to this disease, but we don’t have a real biological link.”

The molecular record, for all its overwhelming garrulousness, its babel of A’s, C’s, G’s, and T’s, is ambiguous. But the fossilized skulls of our ape lineage seem to tell a clear story, with respect to one trait, anyway. The past few million years have witnessed a steady, plodding increase in the volume of the human lineage’s brains and, presumably, the sophistication of their contents. High intelligence is to great apes as the wing is to birds.

But where are we in that process? Is intelligence still being selected for? Parsimony and uniformitarianism would compel one to answer yes; things in the present are, by and large, as they were in the past. But the way evolution works, whereby mutations arise in one person and slowly spread throughout a population, makes such a question difficult to frame, for if intelligence is still under selection, that could mean that some populations at this very moment are slightly smarter than others — that, perhaps, even certain ethnicities are slightly smarter than others. In the West, speculation on the subject almost automatically tars the speculator as a eugenicist or a racialist.

Bruce Lahn is an evolutionary geneticist and a lab director at the University of Chicago, but he was born and completed the early part of his education in China. A heightened sensitivity to imputations of racialism doesn’t afflict most Chinese, according to Lahn, who, by his own admission, has yet to fully internalize the finer points of Western political correctness. In a pair of 2005 papers, he presented evidence that two genes known to play a role in brain development, microcephalin and ASPM, appear to be undergoing continuing natural selection in historical times. In the penultimate paragraph on microcephalin, he observed that “Sub-Saharan populations generally [have] lower frequencies than others.” And after noting that the ASPM mutation, which he refers to as haplogroup D, is most common in Europeans and Middle Easterners and least common in Sub-Saharan Africans, he speculated, “Although the age of haplogroup D and its geographic distribution across Eurasia roughly coincide with… the development of cities and written language 5,000 to 6,000 years ago around the Middle East, the significance of this correlation is not yet clear.”

It makes sense that some alleles present in Europe, Asia, and the rest of the world wouldn’t appear in Sub-Saharan Africa, and vice versa; population flow has not yet had time to spread all alleles to all parts of the world. However, it’s hard for many of us not to hear in Lahn’s musings on brain genes the ugly implication that Africans are inferior. But such was not Lahn’s intention, nor was that his finding. It was not even what he was investigating.

“Some interpret it as meaning, this is the civilization gene, which is clearly not what we’re trying to say. Maybe we should have said it with more qualifications, to avoid the misconception,” he says. The belief that minor mutations to two genes could bring about a profound and essential difference to an abstract quality as polymorphous as intelligence Lahn sees as springing from America’s confusion about race, its desire to overcome a shameful past, and a fear that old racist beliefs might be given empirical support. Nevertheless, Lahn and his group did ultimately investigate whether possession of the new alleles correlated with intelligence. It did not.

Indeed, possession of Lahn’s variants might have nothing to do with intelligence. “It could impact emotionality, the ability to be patient, for example,” he says. “Our understanding of brain evolution at the phenotype level is so rudimentary right now. We’re very far from actually breaking down the difference between human and other species, let alone among humans.”

Lahn’s result was criticized in subsequent papers, not on ideological grounds, but on technical ones. It was claimed that the signal for selection he thought he’d found was not there. He took the criticism in stride and reanalyzed his data. “We stand by our conclusions,” he says. “We have more unpublished data to support them. We’re convinced that what we published is real.”

Even if Lahn could prove to everyone’s satisfaction that ASPM and microcephalin are under selection, whether intelligence is the trait being selected for would be far from a settled question. It could be, as Lahn suggested, that some other mental trait is being selected, or that the activity of ASPM and microcephalin in other parts of the body is what is under selection. More work will certainly be done. But one can speculate with far more confidence about what drove the dramatic increase in intelligence attested by the fossil record: the advent of human culture.

“Intelligence builds on top of intelligence,” says Lahn. “[Culture] creates a stringent selection regime for enhanced intelligence. This is a positive feedback loop, I would think.” Increasing intelligence increases the complexity of culture, which pressures intelligence levels to rise, which creates a more complex culture, and so on. Culture is not an escape from conditioning environments. It is an environment of a different kind.

Lahn says there could be “some deep-down information theory perspective” that underlies both the rapid increase in human intelligence and an event like the Cambrian explosion, the unequaled diversification of life forms that occurred about 500 million years ago. In an eyeblink, almost every modern body plan came into existence. “It may take a long time to evolve certain components of the body plan, but once you have them, minor tinkering that requires not many changes and very little evolutionary time could give you great diversity in body plans and species,” Lahn says. “The brain may be similar, because it takes a long time to get to a certain level of intelligence, but once you get there, it makes possible a cultural explosion.”

Both events inched toward a threshold that, once crossed, was soon left far behind. The 20th century, in which it took us a mere 60 years to elaborate the horse-drawn carriage into a vehicle that carried us to the Moon, and the howitzer into a 50-megaton nuclear weapon, was another threshold. The forces that we created are on a different scale than those of nature, which works slowly. It seems possible that as our technology grows more subtle, genetic manipulation, gene manufacturing, and even cloning could finally carry us clear of natural selection, but such a commanding position can be maintained only with the survival of a technological society, and that is hardly a foregone conclusion.

The Bleakness of that vision exerts a strong hold on Paul Ehrlich, a professor of population studies at Stanford, who finds in the 20th century a minefield of near misses with extinction. We were saved as often by cunning as by dumb luck: intended to save sleeping families from exploding refrigerators cooled by ammonia, chlorofluorocarbons nearly fried the entire planet. As often as not, some solution creates a new problem.

“The fate of our civilization, and maybe our species,” says Ehrlich, “may be determined by the next five generations. So I don’t really give a shit what’s happening to our genetic evolution.” The global climate is changing too violently for DNA to respond by fiddling around with heat regulation and hair thickness; forests everywhere are being clear-cut too quickly for their inhabitants to adjust, and so food chains are coming undone; the collapse of global fisheries has been identified as an imminent calamity; and a nuclear disaster would constitute a catastrophe many orders of magnitude larger than what nature could readily absorb. If any of these nightmare scenarios comes to pass, Ehrlich fears, evolution will be unable to help us. It may be operating faster than we thought, but it’s not that fast. Problems like smog and acid rain seem almost quaint, and even to be longed for.

Species are transient. There is no question that the day will come when humans are no longer on Earth. But the transience to which we are subject has two faces. The first is extinction. Unlike our forebears, we are aware of how tenuous is our perch atop the food chain. It remains to be seen whether that knowledge has been acquired too late to be of use.

The second face of Homo sapiens’ eventual exit from history is the more hopeful possibility that we may yet evolve into our own successors. Unlike our forebears, we are aware of evolution, which changes our relationship to it, if only by a little, for we are still natural creatures. We continue to evolve, in the face of hunger, disease and a changing ecosystem; but our virtual habitat of culture could enable us to become both subjects of evolution and conscious co-directors of it. “It’s occurring,” says Ehrlich. “There’s no question about it. What’s frightening is the questions we’ll have to ask.”

Science must evolve new tools to raise us to such a commanding vantage, as well as to avert a self-inflicted extinction. Technology might some day enable us to control aspects of evolution, or it may prove to be the ultimate selection regime, culling all of us. Perhaps we already find ourselves wishing we’d lacked the intelligence to monkey with howitzers. Either way, the culture that we’ve created is, strangely, evolution’s most powerful tool and its potential nemesis, the womb of human nature and perhaps its grave. By our own hand: this is how we evolve.