In this series, we explore the genetic evidence that indicates humans became a separate species as a substantial population, rather than descending uniquely from an ancestral pair.
In the last post in this series, we began to explore the ideas of apologist William Lane Craig as they pertain to Adam and Eve. Specifically, we saw that Craig holds to genetic monogenesis – the hypothesis that humans descend uniquely from an ancestral couple, rather than a population – in the face of population genetics evidence to the contrary. Part of his reasoning, as we saw, was to suggest that the human mutation rate was once far higher than it is now. For Craig, this includes the possibility that God may have supernaturally increased the mutation rate of the human lineage to account for present-day genetic diversity. With such a miraculously accelerated rate, Craig argues, the genetic diversity we see in present-day humans could in fact arise from just two individuals – the historical Adam and Eve:
In order to calculate whether this amount of genetic diversity could arise from an initial human pair, you have to assume a certain mutation rate that is constant over time. One might deny this conclusion by postulating accelerated rates of mutation in the early human population. One could see this as a result of divine intervention – that God accelerated the evolution of early humans so as to produce greater genetic diversity.
As we have seen, this argument fails to account for calculations of human ancestral population sizes that are not dependent on estimates of mutation rate. An additional problem with this argument is that humans are not merely genetically diverse, but that we observe patterns of diversity in different human populations. Genetic variation in humans is not uniformly distributed across the globe. If you examine Northern European populations, for example, you find certain genetic variants connected to other certain variants in predictable ways. The same holds for any other human population – Han Chinese, populations in sub-Saharan Africa, and so on. It is these patterns of variants linked on chromosome sections that allow us to use linkage disequilibrium to estimate our ancestral population size. And as we have seen, this method returns the same value (~10,000) that other methods do.
The question for Craig, then, is not merely “why are humans so genetically diverse?” but rather “why do humans have their abundant genetic diversity in the particular pattern we observe in the present day? ”While the first question might appear to have a simple answer, however farfetched it may be – i.e. increased mutation rates – this answer does not even begin to explain the second question. Mere mutation alone would not put certain combinations of variants into various populations. These patterns are inheritance patterns, not mutation patterns.
Let’s look at an example to help us understand this issue. Suppose one small segment of a human chromosome has four regions where variation is present in human populations. We can represent the chromosome as a line, and mark off the four locations with letters (a, b, c and d). Each of these locations has multiple DNA sequence variants within the population. Suppose for one individual, they have the following combination on one of their chromosomes: they have DNA variant 4 at position “a”, variant 7 at position “b” and so on:
Now suppose we look at other individuals from the same population for these same four chromosome locations. In them we observe new DNA variations present at locations “c” and “d”, and we observe different combinations of those variants:
At locations “c” and “d”, we see two options at each location, and we see a few combinations: (c2 with d6), (c2 with d1), and (c7 with d1). Now the question for the geneticist arises: how did these three possible combinations arise? The basic issue is this: are we looking at the results of multiple, independent mutation events, or mixing and matching of the same mutations into various combinations through genetic recombination? Take for example variant c2: we see this exact variant (down to the precise mutation at the DNA level) in the top two chromosome combinations. The probability that these two chromosomes independently mutated to variant c2 is possible, but tiny. Far, far likelier is that these two chromosomes have the same c2, from a single mutation event in the past that has been recombined into two combinations with different variants at the d location. For example, suppose this population has many individuals with the (c2 d6) combination (first chromosome), but fewer individuals with the (c2 d1) combination (second chromosome). This observation suggests that the (c2 d6) combination is older, and has been passed down from a distant ancestor to more offspring over time. The rarer (c2 d1) combination is likely more recent, and it likely arose by a recombination event. Suppose the third combination (a2 b9 c7 d1) is also common in the population – this is the likely source of the d1 variant seen in the second chromosome combination. The alternative hypothesis is that the d1 variant arose independently twice (or that the c2 variant did) – i.e. the exact same mutation happened twice over. This is much less likely than assembling the second chromosome through the mixing and matching of recombination using the common first and third combinations.
And here’s the rub: the recombination events are also infrequent events. The more closely linked together two locations on a chromosome are, the less frequent recombination becomes. We can directly measure recombination rates in present-day humans and other organisms, and we have a good understanding of how physical separation of two locations on a chromosome is proportional to the recombination frequency between them.
So, not only do we need to account for rare mutation events, we also need to account for rare recombination events - in this case, the breakage and rejoining of chromosomes through the biological process called “crossing over”. While this example uses only four locations, imagine chromosomes that have hundreds or thousands of locations with these sorts of patterns – which is what we see in present-day humans. There are, after all, about 300,000 locations in our genome of 3 billion DNA base pairs that have this sort of variation present, in a staggering array of combinations. That amount of diversity requires a large ancestral population, because it is too improbable to postulate that a huge number of rare recombination events occurred in a small number of ancestors. That number of rare events requires a large population for the probabilities to be reasonable – just as the number of DNA variants we see in present-day humans requires a large ancestral population to allow for rare mutation events to occur.
So, an accelerated mutation rate alone is simply not going to account for those patterns. A normal mutation rate followed by genetic recombination of mutations over time in a large ancestral population, however, easily explains the pattern.
To what end, variation?
I suppose that Craig might reply that we could expect miraculous governance of recombination as well as miraculous acceleration of mutation to maintain the view, in spite of the evidence, that humans uniquely descend from an ancestral couple rather than a population. The question I have, aside from the ad hoc nature of such arguments, is why? Why would God engineer this variation into human populations in such a way to appear to be a result entirely consistent with processes and rates we observe in the present, when there is no functional reason for this variation? After all, we are discussing variation in present-day humans, which means that none of the variation we are discussing contributes to our being human. Humans could be absolutely uniform for these locations in our genome with no virtually impact on our biology, except in rare instances. This is variation that we don’t need or use – so the suggestion that God engineered it into present-day human populations, giving the strong impression with multiple, converging lines of evidence that we are a species descended from a large population, but for no apparent purpose, is frankly baffling.
In the next post in this series, we’ll continue to explore Craig’s arguments about population genetics models and their reliability.