Return to Africa: Evidence of M. Eastern DNA among Khoisans


I stumbled across this this evening and thought, I bet Joshua would appreciate this if he hasn’t come across it already (which you most likely have). Studies of DNA in the Khoisan hunter-gatherer groups have found a signature of a relatively recent (3,000 years ago) return to Africa from the Middle East via Ethiopia, and Neanderthal DNA was found among Nigeria’s Yoruba. African DNA is not free of non-African admixtures, which makes a recent genealogical Adam at least marginally more plausible.


Excellent catch. The timing of that gene wave works out beautifully with the chronology using the “long way” to read the genealogies in Early Genesis. That has the scattering after Babel equated with the sudden end of the Uruk Expansion in 3100 B.C. At that point we do see that people leave off building cities and start to move to the surrounding country. I guess even the Ussherian dates would work for this particular incursion of genes.


Excellent find. Also came across this recently.

A few things are notable.

  1. It’s very difficult (read impossible) to detect the vast majority of mixing events involving less than 5% of the population and/or earlier than 4,000 years ago. Genetic data does not give a complete picture of migration.

  2. Still, even then, we see evidence over the last 4,000 years of mixing everywhere, and over very long differences, even before modern technology.

  3. So called “genetically isolated populations” are not actually genetically isolated when we look at the whole genome.

  4. Several events detected here are independently verifiable from recorded history. This is very strong validation of population genetics in general (even though some are prone to overstate the findings of population genetics).

@jongarvey in particular had skepticism about population genetics, wanting to see some independent validation. This paper is a great example of independent validation. It is also a great example of the limits too. This is really solid science, but it has limits. And even in those limits, in increases our confidence in a recent genealogical Adam.


Several excellent resources - and Joshua’s, in particular, good confirmatory evidence for the optimistic assumptions of Rohde’s work and, hence, the Genealogical Adam hypothesis.

My hesitations about population genetics are hesitations about any system of modelling, especially one in which there are so many simplifying assumptions which, when dealing with lost events in the distant past, cannot in principle be validated. That, to my mind, makes any projection into the distant past a question of “what ought to have happened, given the model” - an intrinsically weak case - how does one know that factors not visible in the short term may not skew the conclusions? And how does one even confirm what did happen? Even weaker, of course, is the argument “X cannot have happened, because of the model.” As the Venema-Bugg exchange shows, such confidence is all too easily confounded, even from within the model. In other words, it’s all about recognising the limitations of ones models, rather than assuming that “science has spoken the final word”.

The admixture history business still makes assumptions, of course, but pretty basic ones: here is an allele, there is an allele - there must be a genealogical link. It’s probably easier to see the limitations of its predictive power.

Revealed_Cosmology: I always wave a flag for Ussher and his dating, partly because I believe he was related through marriage to a distant ancestor of mine (after Adam, I mean). He’s often represented as an example of blind faith in tradition, whereas at the time he was modelling the very best science and was, in fact, a great intellectual influence on his family friend Robert Boyle. His problem was putting too much faith in the assumptions of his model, ie that biblical chronology was sequential and precise, and that the genealogies were complete. Plus, of course, that “man” in Gen 1 was “Adam” of ch2.

Given those methodological errors, it’s remarkable how closely his dates match “likely” dates for people and events from modern archaeology and so on, eg the date of the Exodus, of the Patriarchs, of the Flood (if equated to the great Shuruppak flood) and of Adam’s cultural setting, conveniently close to the known origin of written records in Mesopotamia.