In the last part, we finished our discussion of William Lane Craig’s incorrect use of an experimental population genetics study in his attempt to build a scientific case that all humans descend uniquely from Adam and Eve. Craig’s claims were made in the context of an apologetics class, and in the ensuing question and answer period a perceptive member of the audience asks Craig how species closely related to humans, such as Neanderthals, fit into the picture:
Question: I have heard Fuz Rana talk about on some of the Reasons to Believe podcasts that there is evidence for human Neanderthal interbreeding. I was wondering what that would do for the genetic diversity.
Craig: I wasn’t going to talk about that but that is a very, very unsettling question. When we talk about this human lineage, this includes not just modern human beings but that includes these other – I am hesitant to call them humans – but it includes these other organisms like Neanderthal man, there is this other group called Denisovans, and then of course earlier forms like Homo erectus and Homo habilis. These are all in the human line as well. So the question is: where do you want to insert Adam and Eve?
The challenge, as Craig sees it, is this: if Neanderthals are fully human, then they descend from Adam and Eve. The problem with this option, however, is that Neanderthals are present in the fossil record stretching back 300,000 years or more, far further back in time than his preferred date for Adam and Eve, which he places at about 150,000 years ago, where some studies indicate a population bottleneck. This is also further back in time than even the earliest Homo sapiens fossils, which are about 200,000 years old. If Neanderthals are a separate species, however – meaning they are not descended from Adam and Eve in Craig’s thinking – then the evidence for human-Neanderthal interbreeding is a problem, because humans are then interbreeding with a non-human species:
Craig: But apparently modern human beings interbred with Neanderthals as you say. I remember being taken aback when one of these population geneticists said to me when I was in Canada earlier this year that you, yourself, carry Neanderthal DNA. In my own genetic profile, I carry the DNA of these Neanderthals who interbred with human beings. Now, if they weren’t humans that meant that the descendants of Adam were literally committing bestiality, right? They were interbreeding with animals. Well, maybe that is possible. Maybe that is part of the fall of man into sin – that they engaged in behavior like that…
Certainly, Neanderthals create an issue for Craig’s apologetic. If indeed Neanderthals are human, then their genetic variation also needs to be accounted for in Craig’s model. This would add even more genetic variation to the human population, even more than standard evolutionary biology would need to account for. As such, Craig’s efforts to explain away the variation within present-day humans would also need to be expanded to include all Neanderthals, not just those that interbred with humans. The alternative option – that of bestiality that nonetheless produces viable offspring whose descendants are with us to the present day, myself and Craig included – is fraught with theological issues. Did the first generation of hybrid individuals have the image of God in Craig’s view? Were they subject to original sin, which Craig thinks is biologically inherited from Adam in some sense? Solving the Neanderthal problem in this way is no easy out – and it’s not surprising that Craig has not settled on an answer to this thorny question.
Census size and effective population size
Despite these obvious problems, there is another problem lurking under the surface related to these issues that Craig is seemingly not aware of: the difference between census size (the actual number of individuals in a population) and effective population size (the number that population genetics techniques estimate). Effective population size is the minimum population size needed, on average, to account for the amount of genetic variation we observe in the present – but census size is the actual population size that existed over time. Here’s the issue: census size is almost always larger than effective population size – substantially larger. For example: consider all the individuals in a population that do not reproduce – these individuals are part of the census size, but not part of the effective population size, since they did not contribute to the genetic variation seen in the present day. Or consider the case of close relatives that inherit the same genetic variation from their parents. Such “duplicate” individuals would not be included in the effective population size, but are part of the census size.
To return to Neanderthals, we know that most of them did not contribute their genetic variation to present-day human populations. Neanderthals, as a species, persist in the fossil record for at least 300,000 years – yet only a tiny fraction of human DNA in some present-day populations descends from them. This shows us that interbreeding between Neanderthals and humans was limited, and had only a small effect on present-day humans. If Craig were to accept that Neanderthals were indeed fully human, however, he would need to include all of them in the human census size, even though they have only a very minor contribution to the human effective population size. Even if Craig were to exclude Neanderthals, there are many humans in the past who have not left descendants to the present day – their lineages have died out. As such, they are part of the census size, but not the effective population size.
A concrete example of this was recently reported in the scientific literature, and I have discussed it previously. The study reported the sequenced genome of an ancient human male (named Oase 1, for the location where his remains were discovered) who lived about 40,000 years ago in Europe. His genome sequence is noteworthy because it contains very long stretches of DNA that are Neanderthal in origin – stretches that had not yet been broken up through crossing-over events, exactly like what we discussed recently. As such, this individual had a very recent Neanderthal ancestor – likely within 200 years of his birth.
A second finding, and the one relevant to our current discussion, was that Oase 1 did not contribute to present-day human populations – for any of his DNA. Even those regions of his genome that are “human” rather than “Neanderthal” were not transmitted to the present day. As such, Oase 1 is a concrete example of an individual who needs to be accounted for in the census size, even though we now know he is not accounted for in the effective population size. Even if Craig excludes Neanderthals as human, the problem doesn’t go away, because Oase 1 is a member of our species. If Craig does accept Neanderthals as human, then there is even more human genetic variation to account for.
So, the apologetic focus on the minimum effective human population size of 10,000 misses the point: the census size was larger than this. The 10,000 are merely those who have descendants in the present day, but they surely come from a population that was larger than 10,000, and likely much larger. As such, even if Craig’s attempts to minimize the effective population size were valid (and we have seen that they are not), they would not address the issue that the actual human census size at this time was larger.