BETTER ESTIMATES POSSIBLE
The estimates presented here are reasonable. They are based on the best simulation of universal ancestry available. Building confidence, the simulation results correspond closely with theoretical analysis. Moreover, the results of this simulation have stood uncontested for more than a decade in the literature. There are reasons to think that the estimates, nonetheless, are too high. Common ancestors might arise more quickly in the past.
First, population size is much lower in the past than it is. Smaller population sizes in the past are expected to reduce the wait times to universal ancestors. All else being equal, the smaller the total population size, the less time it takes to become a universal ancestor. Quantitatively, wait time estimates should scale with log n / log m, where n is the population size at a time in the past, and m is seven billion, the approximate population at present day.15 Using logarithms with base ten, the logarithm of a number is approximately the count of digits. We know m is seven billion, so log m always equals 9.84.16 At 5,000 years ago, there were about 18 million in the world,17 which means log n is about 7.25. Dividing 7.25 by 9.84, the scaling factor is about 0.74, and the corrected MRUGA estimate is the product of this number with 3,000 years, or about 2,200 years. At 10,000 years ago, there were about 2 million people in the world; we use a different number for n, giving us a corrected MRUGA estimate of about 1,900 years. The same factor can be used to adjust the nearly IAP and IAP estimate.
Second, several of the most isolated places, such as Hawaii and the Easter Islands, are not populated at AD 1, so they can be ignored. In the simulations, these remote islands are the last places for genealogical ancestry to spread. The islands are very difficult to access without modern technology, so migrants to these islands were likely particularly rare, making estimates less certain and skewing them more ancient. Before AD 1, however, these islands are uninhabited. In the past, therefore, we might expect reduced time to universal ancestry, and less uncertainty about when it arises.
To make the best estimate, we might use improved simulations. Unfortunately, simulations at the level of detail of the 2004 Nature study are difficult to implement and run. Including all the factors necessary for an accurate estimate would be even more difficult. Ideally, we would want to add in known migration and interbreeding data, making use of the vast cache of new information about intermixing in the past that is currently becoming known through analysis of ancient DNA.18 We would also want to model known events in history as much. All of this is very difficult to model effectively. Perhaps increased interest in these results might stimulate scientists to embark on these efforts.
Fortunately, all these improvements to the simulation would reduce the time to universal ancestry. So, though we use these numbers, we expect these estimates to be higher than reality. We are making cautious estimates. They still are, nonetheless, the best estimates available for when universal ancestry arises in the past.
THE GENEALOGICAL HYPOTHESIS
Under our hypothesis, Adam and Eve must be the “ancestors of everyone to the ends of the earth” from at least AD 1 onward. Their offspring interbred with the people outside the Garden. Does the evidence constrain where or when they could have lived?
Previously published estimates of universal genealogical ancestors are for those alive in the present day. Estimates for the universal ancestors of all those alive at AD 1 are not available. Estimates, as we have seen, are still possible. For Adam and Eve to be ancestors alive in AD 1, they could have lived as recently as six thousand years ago. This date is a conservative estimate. It is also possible that Adam and Eve were situated more anciently, at eight, ten, or fifteen thousand years ago, or whenever might make most sense with other concerns in mind.
As the descendants spread, there would have been three genealogical eras. In the first era, Adam and Eve’s lineage would exist alongside those outside the Garden. In the second era, nearly everyone would be a descendent of Adam and Eve, but there would be a few isolated places that do not descend from them. In the third era, everyone would be a descendent of Adam and Eve. Depending on when Adam and Eve are situated, the third era might begin long before AD 1. The brevity of the second error is an important distinction from genetic ancestry. Different genetic lineages can persist alongside one another for tens of thousands of years, or even more, often with siblings in a single family representing different genetic lineages. In contrast with genetics, in just a few thousand years, everyone across the globe would fall into Adam and Eve’s lineage. Also, unlike genetic ancestry, siblings would all be inside Adam and Eve’s genealogical lineage together, or all outside their lineage.
A key point, moreover, is that universal ancestry would arise without luck or miraculous intervention. Universal ancestors are not restricted to single lineages, rare individuals, or a single location. Instead, universal ancestors arise everywhere. If Adam and Eve are real people, our best estimate is that they would be universal ancestors of everyone alive AD 1 and afterwards.
CHAPTER SIX
THE MYTHOLOGY OF ISOLATION
WHAT ABOUT TASMANIA? What about Easter Island and Australia? What about the indigenous populations that have been genetically isolated for tens of thousands of years? This is a legitimate set of objections, already noted in the commentary on the 2004 Nature paper by another scientist.1 All these objections are concerned with the isolation of populations for thousands of years. If any population of humans was isolated for several thousand years, they might not descend from Adam.
In scientific modeling, it is common to make an initial prediction with a simplified system. We then progressively add complexity to refine our estimates with increasingly realistic models. This progression from simplified to complex modeling is our approach in this analysis. First, Chang’s 1999 paper used a very simplified model, with no barriers to interbreeding across a single intermixing population. In this model, universal ancestry arises in just seven hundred years. Next, the 2004 Nature study modeled universal ancestry across the globe, with several barriers to intermixing. This model shows universal ancestry arises in just a couple thousand years. Finally, my study in 2018 extended these results to apply to everyone alive at AD 1. We estimate that, most likely, a couple in the Middle East six thousand years ago would likely be ancestors of everyone by AD 1. There, however, was a single important assumption in this model that we will examine here.
These final estimates rely on the same assumptions as the 2004 Nature study: no populations were unable to mix with others. Now, we want to consider the validity of this assumption. First, I want to clarify how this assumption figures in the simulation. Mixing between populations everywhere across the globe was possible, at least in principle. No populations were prohibited from intermixing with neighboring populations. No populations were forced to be totally isolated. The simulations, however, did not force intermixing between different populations, used a very low rate of migration, and did not include modern transportation to boost migration. There is important subtlety here. The simulation did not assume that all populations mixed. Instead, the simulation merely assumed it was possible for all populations to mix. Still, we need to assess whether or not it is possible that all populations in the world could mix with one another.
WHAT IF THE ASSUMPTION IS WRONG?
What if this assumption is incorrect? What if one or more populations were isolated for thousands of years in our past? No universal ancestors could arise except before or after that time of isolation. If the isolation was for a very long period of time, this could substantially push back when universal ancestors would arise. What would it mean if a specific human population was, in fact, isolated? The most likely candidates might be the indigenous populations of Tasmania. If this population were isolated from six thousand years ago till AD 1, would this be a problem?
First, if Adam and Eve lived before the population was isolated, it will not matter. For example, if Tasmania were totally isolated from nine thousand years ago onward, thi
s would rule out a universal ancestor at six thousand years ago, but not one more ancient than this. If Adam lived ten or twelve thousand years ago, there would no reason to doubt that Tasmanians all descended from him by AD 1. We have not insisted on placing Adam and Eve at a specific time. Our precise estimate would change, but a recent Adam and Eve would still be likely.
Second, a small number of people that are, in fact, isolated may not be a problem, because theology does not speak with scientific precision. If a few isolated populations do not descend from Adam at AD 1, they would be rare and undetectable exceptions to the rule. As we will see, the doctrine of monogenesis teaches we all descend from Adam and Eve to the “ends of the earth” (Acts 1:8). It is possible for the doctrine of monogenesis to be valid from a theological point of view, even if there are very rare and undetectable exceptions.
Third, our specific understanding of people outside the garden could head off any theological problems that arise from exceptions to the rule of monogenesis. Take the indigenous Tasmanians as an example. For argument’s sake, just for a moment, let us presume they did not descend from Adam. As we will see, we would not be able to determine this for sure from any scientific evidence. We would not be able to pronounce them as outside Adam and Eve’s lineage, even if they did not, in fact, descend from them. Even if we could make this determination, this would not necessarily legitimize the abuses that Tasmanians suffered when they were colonized and exterminated. As long as we have a coherent way of acknowledging their human worth and dignity, it may not be a problem that they are outside Adam and Eve’s lineage. In the final part of this book, I work out just such an understanding to the people outside the Garden, affirming their worth and dignity.
Therefore, even if a rare population is genealogically isolated, we do not face an ultimate problem. At stake here is merely the difference between total universal ancestry and nearly universal ancestry, with a few rare populations undetectably left outside Adam and Eve’s lineage. We should be open to see what the evidence tells us about rare, isolated populations. Ultimately, the genealogical hypothesis does not depend tightly on what we find. As we will see, moreover, the evidence cannot tell us definitively anyway.
The goal of this chapter is to review the evidence for and against the genealogical isolation of individual populations. I want to determine the strongest evidence against total universal ancestors in the recent past. The goal here is to test the assumption of the simulation with evidence. We will find there is no definitive evidence of genealogical isolation, but we cannot be sure. The strongest case comes, perhaps, for the isolation of Tasmania, but even the evidence here is not definitive. Definitive evidence might be outside our view. We may not ever have the evidence to tell us for sure.
THREE TYPES OF ISOLATION
Three types of isolation are important here: genetic, geographic, and genealogical isolation.
Genetic isolation is observable in genetic data. Studying the information in genomes from present-day and ancient individuals, we can determine when segments of DNA are shared between different populations. We can estimate the extent to which populations mixed, and when this mixing took place. As some of the examples will demonstrate, the details are complex, and mixing events can easily be missed.
Geographical isolation is determined using several lines of evidence. Of course, the presence of barriers to migration are important, such as large bodies of water. Across a divide, such as between Tasmania and Australia, we can also study whether animal populations are different, and whether cultural developments are shared or developing in different ways. Together, these lines of evidence can demonstrate geographical isolation.
The critical question is whether or not genealogical isolation persisted for several thousand years. This question is only answerable if genetic or geographic isolation can reliably identify genealogical isolation. As we will see, genealogical isolation does not correspond with either genetic or geographic isolation. Instead, the question of genealogical isolation poses a dilemma of complementary universal negatives. A single genealogically isolated population will prevent a universal ancestor from arising. However, a single migrant or mixing event will break genealogical isolation. On one hand, it is nearly impossible to rule out the isolation of every population. On the other hand, however, it seems impossible rule out low levels of migration in order to demonstrate a population was genealogically isolated for long periods of time. Science, therefore, cannot definitively determine whether genealogically isolated populations have existed in our past or not. Genealogical isolation is not directly observable. We can rule it out with evidence, but we cannot definitively demonstrate it with evidence.
In some important senses, therefore, it is not possible to demonstrate with evidence whether or not genealogical isolation prevented universal genealogical ancestry. Remember, undetectably low levels of migration are all that is required for universal genealogical ancestors to reliably arise in the recent past. I expect, for this reason, for there to be some unresolvable disagreement between scientists. The key questions here cannot be fully adjudicated by evidence.
GENETIC MIXING EVERYWHERE
Genetic evidence can falsify the hypothesis of genetic isolation, and usually does.2 In this way, genetics is producing an increasingly strong case against genealogical isolation. This supports the hypothesis of recent universal ancestors. As data increases, more evidence of mixing is expected. Most genetics studies only consider small portions of the genome.3 Whole genome sequencing usually reveals mixing in the past, even in populations once thought to be genetically isolated. Similarly, ancient genomes provide additional evidence for ancient migrations.4 Even though human populations are fragmented and might be genetically isolated at times, the data demonstrates a pattern of pervasive intermixing everywhere.
In 2013, a study was published by Peter Ralph and Graham Coop.5 Using genetic data from over one thousand Europeans, they found genetic evidence that all Europeans share about two to twelve genetic common ancestors one thousand years ago. This is consistent with millions of shared genealogical ancestors that all Europeans share, just one thousand years back in history. They write, “Individuals from opposite ends of Europe are still expected to share millions of common genealogical ancestors over the last 1,000 years.” They find millions of universal genealogical ancestors, on the one hand, with just a handful of universal genetic ancestors, on the other hand. The vast majority of the universal genealogical ancestors did not leave DNA to everyone.
In 2014, another study used genetic data to infer populations mixing events across the globe, “revealing admixture to be an almost universal force shaping human populations.”6 It was able to detect over one hundred events in the last four thousand years. Many of these mixing events correspond to known historical events, like the Arab slave trade and the Mongol empire. Other mixing uncovered events before history was recorded. The study revealed a churn of mixing between nearly all populations all across the globe, but a few populations did not show evidence of genetic mixing. Does genetic isolation of a few rare populations demonstrate these populations were genealogically isolated? No.
GENETIC ISOLATION IS NOT GENEALOGICAL
Genetic isolation does not demonstrate genealogical isolation. There are several reasons mixing events would be missed, and Coop’s 2014 study details several of them.7 Populations can be genealogically linked even if genetic analysis cannot demonstrate intermixing in the past. Of particular importance, moreover, is that statements of genetic isolation are often made in reference to portions of the genome. One part of the genome might show genetic isolation, while another might demonstrate genetic mixing. In these cases, this conflicting result demonstrates that the population is genealogically linked to others, and not isolated.
It is sometimes claimed, nonetheless, that certain populations have been genetically isolated for long periods of time. Most precisely, these claims should be understood as genetic isolation of specific portions of genome, but not necessa
rily of a population’s whole genome. For example, portions of DNA from the Khoisan people of southern Africa8 and the Aborigines of Australia9 appear to be genetically isolated for tens of thousands of years. This evidence is consistent with substantial cultural and geographic barriers that made mixing and migration less common. This evidence, however, is often based on only portions of the genome, and only demonstrates genetic isolation of these portions. Looking at whole genomes, taking into account all the genetic information, there is usually genetic evidence of mixing.10 This means that the Khoisan people, we know from evidence, were not genealogically isolated, even though parts of their genome have been genetically isolated for a large period of time.
The Andaman Islands in the Indian Ocean are an important case study of the pattern of progress in our knowledge. They entered the news when John Allen Chau, a twenty-six-year-old man from the United States, was killed. He attempted to contact these isolated islanders in order to tell them about Jesus. News stories often repeated the claim that the islanders had been isolated for over fifty thousand years. The islanders were famously hostile to outsiders, it seems. Some of this hostility is likely caused by the very recent history of British colonization of the islands. Did this hostility, however, stretch back for thousands of years? Could they really have been totally isolated for fifty thousand years?
The Andaman islanders are among the most isolated populations on earth. Their isolation, however, has not been total. For a long time, anthropologists wondered about the origins of the islanders. They share external characteristics like skin color and hair type with Africans, even though they are in Asia. This, in addition to the geographic isolation of the islands, suggested an interesting hypothesis. Perhaps they arose from the initial expansions of African Homo sapiens over fifty thousand years ago, and then they remained isolated on the island. Two genetic studies, from 2005 and 2006, supported this hypothesis, showing divergence and isolation for tens of thousands of years.11 These two studies, however, only looked at small portions of the genome. Subsequent studies in 2013 used more genetic data from their neighbors, and variation across the whole genome. These analyses showed intermixing between the islanders and their neighbors, and a much more recent settlement of the island.12 The perception of the ancient and total isolation of the Andaman Islands was an artifact of incomplete genetic and historical data.
The Genealogical Adam and Eve Page 7
