Every few years it seems that the British biologist Steve Jones declares the death of evolution by natural selection in the human species. The logic here is simple even to a schoolboy: evolution requires variation in fitness, and with declining risk by death during our reproductive years humans have abolished the power of selection. But this confuses the symptom for the disease. Death is simply one way that natural selection can occur. Michelle Duggar has 19 children. The average American woman has around two by the end of her reproductive years. It doesn't take a math whiz to figure out that Michelle Duggar is more "fit" in the evolutionary sense than the average bear. Even without high rates of death, some people have more children than other people, and if those people who have more children than those who do not are different from each other in inherited traits, evolution must occur. Q.E.D. But you probably shouldn't be convinced by logic alone. Science requires theory, experiment, and observation. (If you're talking humans, you can remove the second from the list of possibilities: there are certain unavoidable ethical obstacles to experimenting on human evolution---plus we take far too long to reproduce.) But humans sometimes have something which bacteria can not boast: pedigrees! Not all humans, of course. Like most of the world's population I don't have much of a pedigree beyond my great-grandparents' generation. But luckily for biologists, the Catholic Church has long taken a great interest in life events such as baptism, marriage, and death, and recorded this info parish by parish. With these basic variables, demographers can infer the the rough life histories of many local populations over the centuries. In many European nations, these databases can go more than 10 generations back. And some aspects of human evolution are revealed by these records. What aspects am I talking about? Reproduction itself. Not only is variation in fitness one of the primary ways by which evolution occurs, but it is also a trait upon which evolution operates! How else are there rabbits which breed like...rabbits, and pandas...which don't. There is often variation within species for the odds of multiple births, age at first reproduction, and lifespan, depending upon the differences in selection pressures over a population. And that seems to be exactly what occurs in human beings. There is interesting evidence for evolution of reproductive patterns from populations as diverse African pygmies and Finns, but more recently some researchers have been plumbing the depths of the records of the Roman Catholic Church in Quebec, and they've come back with gold. Until the past generation, Quebec was a relatively homogeneous society in which the Roman Catholic Church was the dominant institution in the province. For several centuries, the Roman Catholic Church has been recording the births, marriages, and deaths of almost all Québécois. Not only that, but Quebec is also an ideal "natural experiment." It began as a frontier society with a small founding population on the order of thousands, but now has a population of over 8 million. This incredible demographic expansion gives us an insight which Steve Jones misses: evolution by natural selection can occur not only via death, but also via exuberant life! In a world where the range of offspring tops out at three or four, any variation in outcome due to genetic differences are moderated by the fact that that variation is constrained to a narrow range. In contrast, it was not exceptional on the North American frontier for some women to have more than 10 children. Such a wide potential range of reproductive output (from 0 to 10 children, or more) means that natural selection has much more latitude in driving adaptation through fitness differences. One study focused on a 140-year period on an island in the Gulf of St. Lawrence. Over five generations, the island's population increased by a factor of 10 through natural increase, while the average age of first reproduction declined from 26 to 22. Just as menarche and menopause are heritable (variation in genes explain much of the variation in the outcome of the trait), so too age at first reproduction seems heritable. Of course age at first reproduction is not necessarily due just to differences in genetic disposition across families; there are strong cultural factors which can bias this value up or down. But variation across families and within families seems to be strongly explained by the extent of genetic relationship. The fact that the median trait value changed so much over the century and a half of the survey, combined with its heritability, allows one to infer that this shift was driven in part by natural selection. Another group focused on the differences between core and frontier populations in the Saguenay–Lac-Saint-Jean region of Quebec. It turns out that the majority of the modern population descends from those in the frontier zone, not the core. Women on the frontier were ~20% more fertile, and fertility as a trait was heritable on the frontier but not in the core. (On the frontier, there was a connection between relatives and how many offspring they would have, in direct proportion to the amount of genes they shared. E.g., two sisters tended to have similar numbers of offspring, while there was no correlation between strangers.)