Friday, December 28, 2007

The 99.9% truism

There has been much comment on a recent finding that human evolution has accelerated over the past 40,000 years, i.e., the period during which our species has spread out of Africa and differentiated into the populations we see today (Hawks et al., 2007). There has been less comment on a related finding: at least 7% of the human genome has changed over the same 40,000 years.

This second finding seems to challenge a truism that has become widespread in academia and even in our political culture. In a speech earlier this year, Hillary Clinton cited genetic research showing that human populations are 99.9 percent the same and “that the differences in how we look -- in our skin color, our eye color, our height -- stem from just one-tenth of 1 percent of our genes.

Isn’t there a contradiction here? How can human populations be 99.9% genetically identical if at least 7% of the genome has changed since they began moving apart some 40,000 years ago?

First, the 99.9% figure is not the number of genes that are the same. It’s the number of nucleotide sequences that are the same. A single gene is a long chain of nucleotides, often a very long one, and a single nucleotide mutation can significantly alter how the entire gene works. In theory, each and every human gene could differ by 0.1% from one population to another. And such a difference could make a big difference.

Second, the 99.9% estimate doesn’t capture higher-level nucleotide variation:

The technique originally used … could read the sequence of letters of a genetic code. But it couldn't detect repetitions of some parts of the code, which also occur. Differences in the number of these repetitions, called copy number variants, have since turned out to account for much of the variation in a species' DNA. Another type of variation recently found to be important is called insertion-deletion variants, snippets of code that are either extra or missing in some genomes compared to others. (World Science, 2007)

This higher-level variation has caused geneticist Craig Venter (of the Human Genome Project) to revise the 99.9% figure downward:

The find­ings re­veal “hu­man-to-hu­man varia­t­ion is more than sev­en-fold great­er than ear­li­er es­ti­mates, prov­ing that we are in fact very un­ique in­di­vid­u­als at the ge­net­ic lev­el,” Ven­ter said. The 99.9 fi­gure might need to be lowered to about 99, he added. (World Science, 2007) (also see original article: Redon et al, 2006)

So our nucleotide sequences may be closer to being 1% different, and not 0.1%. And don’t be fooled by small numbers. Whether it’s 1% or 0.1% the difference is still big in absolute terms. As John Hawks points out: “one-tenth of 1 percent of 3 billion is a heck of a large number -- 3 million nucleotide differences between two random genomes.”

Finally, there is a third reason why we should not read too much into any of these estimates. When the 99.9% figure first came out in the 1970s, geneticists had also discovered that nucleotide sequences were 98.9% the same between humans and chimpanzees (King & Wilson, 1975). And yet, humans and chimps exhibit a wide range of anatomical and behavioral differences. How come?

There is of course the aforementioned ‘small percentage fallacy’: a tiny sliver of the genome still amounts to a lot of DNA. More importantly, humans and chimps seem to differ the most in ‘regulatory genes’ whose effects are many times greater than those of ‘structural genes’ (the ones that code for the building block proteins of body tissues). A single regulatory gene has such a disproportionate impact because it can control the expression of many other genes.

These less numerous regulatory genes have gained importance as organisms have grown more and more complex. This has especially been so during human evolution. Whereas humans and chimpanzees are almost identical in the proteins that form their tissues, they differ radically in the way their brains and bodies develop. This point is summarized by King and Wilson (1975, p. 115):

The genetic distance between humans and chimpanzees, based on electrophoretic comparison of proteins encoded by 44 loci is very small, corresponding to the genetic distance between sibling species of fruit flies or mammals. Results obtained with other biochemical methods are consistent with this conclusion. However, the substantial anatomical and behavioral differences between humans and chimpanzees have led to their classification in separate families. … A relatively small number of genetic changes in systems controlling the expression of genes may account for the major organismal differences between humans and chimpanzees.

Interestingly, King and Wilson see this paradox as applying not only to human-chimpanzee genetic differences, but also to genetic differences within our species:

This [human-chimpanzee] distance is 25 to 60 times greater than the genetic distance between human races. In fact, the genetic distance between Caucasian, Black African, and Japanese populations is less than or equal to that between morphologically and behaviorally identical populations of other species. (King & Wilson, 1975, p. 113)

Yet human races are not identical populations, anymore than humans and chimpanzees are sibling species. These measures of genetic distance are not comparable because the nature of genetic change can vary dramatically. In one case, there is simply tinkering with an existing body plan through mutations in structural genes. In another, there is radical developmental change through mutations in regulatory genes.

Since the time that the ancestor of these two species lived, the chimpanzee lineage has evolved slowly relatively to the human lineage, in terms of anatomy and adaptive strategy. According to Simpson:

Pan is the terminus of a conservative lineage, retaining in a general way an anatomical and adaptive facies common to all recent hominoids except Homo. Homo is both anatomically and adaptively the most radically distinctive of all hominoids, divergent to a degree considered familial by all primatologists.
(King & Wilson, 1975, p. 113)

This is the context in which the 99.9% statistic was initially presented to the academic community … way back in the 1970s. Even then, researchers thought it misleading and went to great pains to explain why it was misleading. Yet their caveats were to no avail. The 99.9% truism has taken on a life of its own, much like those stories we hear of alligators living in sewers or evil people sticking razor blades in Halloween apples. It seems to meet a deep-seated need to affirm our sameness and to give this affirmation a stamp of scientific approval.

But science it is not.


Anon. (2007). Finding said to show "race isn't real" scrapped

Elliott, P. (2007). Clinton tells grads only minor genetics make them different.

Hawks, J. (2007) Disagreeing with Hillary Clinton on human genetic differences.

Hawks, J., E.T. Wang, G.M. Cochran, H.C. Harpending, and R.K. Moyzis. (2007). Recent acceleration of human adaptive evolution. Proceedings of the National Academy of Sciences (USA) early view.

King, M-C. and A.C. Wilson. (1975). Evolution at two levels in humans and chimpanzees. Science, 188, 107-116.

Redon, R., S. Ishikawa, K.R. Fitch, L. Feuk, G.H. Perry, T.D. Andrews, H. Fiegler, M.H. Shapero, A.R. Carson, W. Chen, E.K. Cho, S. Dallaire, J.L. Freeman, J.R. González, M. Gratacòs, J. Huang, D. Kalaitzopoulos, D. Komura, J.R. MacDonald, C.R. Marshall, R. Mei, L. Montgomery, K. Nishimura, K. Okamura, F. Shen, M.J. Somerville, J. Tchinda, A. Valsesia, C. Woodwark, F. Yang, J. Zhang, T. Zerjal, J. Zhang, L. Armengol, D.F. Conrad, X. Estivill, C. Tyler-Smith, N.P. Carter, H. Aburatani, C. Lee, K.W. Jones, S.W. Scherer & M.E. Hurles. (2006). Global variation in copy number in the human genome. Nature, 444, 444-454.

Friday, December 21, 2007

Thoughts on the EEA

For the past twenty years, a key concept in evolutionary psychology has been the ‘environment of evolutionary adaptedness’ (EEA). This is the ancestral environment within which our species first evolved and whose selection pressures shaped our current psychology and behavior. Usually, writers situate this environment in the Pleistocene before Homo sapiens began to spread out of Africa some 50,000 years ago.

The EEA concept has increasingly come under fire in recent years, especially with the recent Hawks et al. study. It now appears that human genetic evolution did not stop 50,000 years ago. Nor has it since slowed down. In fact, it has accelerated by as much as a 100-fold. In light of these findings, can the EEA concept be salvaged? Should it?

Interestingly, its earliest proponents, John Tooby and Leda Cosmides, have always been reluctant to narrow it down to a specific place and time:

Although the hominid line is thought to have originated on edges of the African savannahs, the EEA is not a particular place or time. The EEA for a given adaptation is the statistical composite of the enduring selection pressures or cause-and-effect relationships that pushed the alleles underlying an adaptation systematically upward in frequency until they became species-typical or reached a frequency-dependent equilibrium (most adaptations are species-typical; see Hagen, Chapter 5, this volume). Because the coordinated fixation of alleles at different loci takes time, complex adaptations reflect enduring features of the ancestral world. (Tooby & Cosmides, 2005, p. 22)

According to Tooby and Cosmides, there are potentially as many EEAs as there are human adaptations. Therefore, some human characteristics may have originated in very old EEAs and others in more recent ones.

How recent? For Tooby and Cosmides, the limiting factor is complexity. The more complex the adaptation, the more genes it will involve, and the longer the evolutionary time to coordinate all those genes. Therefore, recent human evolution has probably only involved simple traits, certainly nothing as complex as behavior.

The problem with this argument is that complex traits do not arise ex nihilo. They arise from changes to existing traits that may be just slightly less complex. A point mutation can greatly alter the functioning of a trait that involves thousands upon thousands of genes. Keep in mind that genes vary considerably in their effects. At one extreme, a single ‘structural’ gene may code for one protein. At the other, a single ‘regulatory’ gene may control the output of numerous structural genes … or even numerous regulatory genes like itself. As Harpending and Cochran (2002) point out:

Even if 40 or 50 thousand years were too short a time for the evolutionary development of a truly new and highly complex mental adaptation, which is by no means certain, it is certainly long enough for some groups to lose such an adaptation, for some groups to develop a highly exaggerated version of an adaptation, or for changes in the triggers or timing of that adaptation to evolve. That is what we see in domesticated dogs, for example, who have entirely lost certain key behavioral adaptations of wolves such as paternal investment. Other wolf behaviors have been exaggerated or distorted. A border collie's herding is recognizably derived from wolf behaviors, as is a terrier's aggressiveness, but this hardly means that collies, wolves, and terriers are all the same. Paternal investment may be particularly fragile and easily lost in mammals, because parental investment via internal gestation and lactation is engineered into females but not males.

In all fairness, when the EEA concept was first developed, few people were arguing that natural selection has modified human behavior over the last 50,000 years. In fact, the dominant view was the opposite: that natural selection has not shaped any specific human behavioral traits, not now, not over the past fifty thousand years, and not over the past fifty million. Not ever. The mind was a tabula rasa. Even sociobiologists, often castigated as biological determinists, commonly thought that people were simply predisposed to learn adaptively: “natural selection has produced in humans a general motivation to maximize one’s inclusive fitness—i.e., a domain-general psychological mechanism” (Buss, 1991, p. 463).

The EEA was part of a new paradigm, now called evolutionary psychology, to move away from the domain-general approach of sociobiology and to search for specific innate mechanisms within the human mind. Its earliest proponents saw the EEA not as a dogma, but as a guide—as a way of making people look at human nature from a broader evolutionary perspective, and not from the narrower one of modern industrial life.

The EEA concept has served us well. But it is now time to move on.


Buss, D.M. (1991). Evolutionary personality psychology. Annual Review of Psychology, 42, 459-491.

Harpending, H. and G. Cochran. 2002. "In our genes", Proceedings of the National Academy of Sciences 99(1):10-12.

Tooby, J. and L. Cosmides. (2005). Conceptual Foundations of Evolutionary Psychology. In David M. Buss (Ed.) The Handbook of Evolutionary Psychology. (pp. 5-67), Hoboken, NJ: Wiley.

Friday, December 14, 2007

The rising curve

It was long thought that human genetic evolution pretty much ended with the advent of culture. As Paul Ehrlich (2000, p. 63) wrote:

The evolution of that body of extragenetic information—cultural evolution—has been centrally important in making us the unique beasts we are. Cultural evolution rests on a foundation of genetic (or biological) evolution—especially that of our brains and tongues—but can proceed at what by comparison is a lightning pace. … cultural evolution can vastly outpace genetic evolution because it’s not constrained by generation time. Our genes are passed only from one generation to relatives in succeeding generations. In contrast, the units of culture—ideas, basically—are passed among both relatives and nonrelatives not only between generations (in both
directions) but also within generations.

So did cultural evolution make genetic evolution obsolete? Paul Ehrlich seemed to draw this conclusion … only to pull himself back. “There are many ways in which culture can alter selection pressures,” he says, noting that genes have co-evolved with changes to diet, farming practices, and shelter (Ehrlich, 2000, p. 64). Indeed, the same properties that make cultural evolution so fast have also been diversifying the adaptive landscape at an unparalleled rate. Whenever our species came up with a cultural innovation—a new technology, domestication of a plant or animal, or the advent of agriculture itself—our environment changed as fundamentally as if we had moved to a new ecosystem.

So which factor has mattered most in determining the pace of human genetic evolution? Has cultural evolution been resolving more and more adaptive problems that were formerly resolved by genetic evolution? Or has genetic evolution been resolving more and more adaptive problems because human environments have been diversifying more and more?

The second factor, apparently. A recent study has concluded that genetic evolution has actually accelerated over the past 40,000 years and even more over the past 10,000-15,000. This is partly because there are many more of us and partly because we are spread over an increasingly diverse range of natural and man-made environments. At least 7% of the human genome appears to have changed since the advent of Homo sapiens. And the rate of change has increased a 100-fold since the advent of agriculture (Hawks et al., 2007).

These are high numbers. As one of the study’s authors observes:

Personally, I can't believe that nobody noticed how extreme these estimates of recent selection really are. I guess that folks doing genomics just weren't as primed in evolutionary theory to perceive how weird the human estimates looked compared to what is measured in the wild on other species, or even over the span of human evolution!

In the earliest studies, when people were finding that 3 or 4 percent of a sample of genes had signs of recent selection, those numbers were already extremely high. They got even higher, as more and more powerful methods of detecting selection came online. Our current estimate is the highest yet, but even this very high number is perfectly consistent with theoretical predictions coming from human population numbers.

These figures, if anything, err on the low side. They do not capture recent selective pressures that are just emerging above noise in the data. Nor do they capture older selective pressures that have already pushed many alleles to fixation. The real figures won’t become known until we’ve retrieved the human genome that existed 40,000 years ago—something that is certainly within the realm of possibility.

All this underlines a point I made in an earlier post: human evolution is not a straight line. It’s a logarithmic curve with most of the evolutionary change in the recent past. If we met a Homo erectus face to face, or even a Neanderthal (who was probably just an arctic-adapted Homo erectus), we wouldn’t consider it to be human. It would look to us like an overgrown ape. Nor would its behavior reassure us otherwise.


Ehrlich, P.R. (2000). Human Natures. Genes, Cultures, and the Human Prospect. Penguin: New York.

Hawks, J., E.T. Wang, G.M. Cochran, H.C. Harpending, and R.K. Moyzis. (2007). Recent acceleration of human adaptive evolution. Proceedings of the National Academy of Sciences (USA), 104(52), 20753-20758.

Friday, December 7, 2007

Facial recognition and skin reflectance

Richard Russell, a postdoc at the Harvard Vision Sciences Laboratory, has just published a study that shows that minor differences in skin pigmentation are more critical to facial recognition than facial shape. In this study, “subjects were asked to recognise color images of the faces of their friends. The images were manipulated such that only reflectance or only shape information was useful for recognizing any particular face. Subjects were actually better at recognizing their friends’ faces from reflectance information than from shape information” (Russell & Sinha, 2007).

Mihai Moldovan has another paper showing that ruddiness acts as a signal of male dominance in humans, especially in a context of male-male competition (see my earlier post on this topic). I’ll have more to say when the paper comes out.


Russell R and Sinha P, (2007). Real-world face recognition: The importance of surface reflectance properties. Perception 36(9), 1368 – 1374