Monday, December 2, 2013

Y-DNA Expansions And Polygamy

Both Europe and Africa have dominant Y-DNA lineages that are probably associated with populations that had major expansions sometime in the Holocene era.

In principle, Y-DNA lineages can expand much more rapidly than mtDNA lineages. The maximum number of children that a group of women with identical mtDNA can have in a lifetime on average is much lower than the maximum number of children that a group of men with identical Y-DNA can have in a lifetime.

But, how often does this happen and does the historic anthropology support the suggestion from the genetic evidence.

Commentators like Razib Khan (who has migrated Gene Expressions to a new website as of yesterday) speak of androcide - the near complete exclusion of substrate population men from having male children who go on to reproduce themselves as at the heart of this process, while substrate population women are more often appropriate by men of incoming superstrate populations via polygyny of the equivalent (multiple women per man).
Both history and ethnography document mass population collapse in concert with an androcide of the Amerindians. By this, I mean that European males took Amerindian women as concubines, and engaged in de facto polygyny in the New World.
Polygyny remains common even today in much of Africa.

The evidence is sketchier in prehistoric populations like the Bell Beaker and Corded Ware peoples whose civilizations were superseded by later populations before it was possible for those civilizations to be documented in writing.

The alternative possibility that there was merely segregation of people into a rapidly expanding largely monogamous superstrate population and a stagnant largely monogamous substrate population, with some level of bride exchange between the two populations, is also possible.  In that scenario, bride exchange brings substrate women into the rapidly expanding population mix, while bride exchange in the other direction is hard to discern from genetic evidence in modern populations since it is hard to distinguish from superstrate women who remained in their home population and because exchanged brides would make up a small part of a stagnant population whose descendants are becoming an increasingly small proportion of the total population.

If the growth rate for the expanding population is much greater than for the segregated stagnant population, it is hard to distinguish a polygyny and androcide scenario from a scenario with monogamy and moderate levels of equitable bride exchange from each other on the basis of genetic evidence alone.

It ought, in principle, to be easier to distinguish a scenario in which migration and conquest was overwhelmingly by men (e.g. an invading army) from one that is a mixed gender migration - as there would be no non-indigenous mtDNA lines introduced in great frequencies while there was a great change involving a newly introduced group of Y-DNA lineages.  A male dominated conquering population could support a scenario of androcide (often literally with indigeneous men of reproductive age killed in war and their widows appropriated by the conquerors) without significant polygyny.

There is evidence of polygny and the equivalent in European cultures at the dawn of the historic era like the Celts (alluded to in several comments here and also inferred here and here).  This may have also been present in early Bell Beaker and Indo-European populations, and on the to do list is an effort to track down better authority on the extent to which it is likely that there was much polygyny in those prehistoric or ancient cultures.  There is also legendary history evidence of androcide in wars of conquest.

But, it is hard to tell.  Polygny leaves a far fainter trail of artifacts in archaeological digs than technologies like metal use or food production technologies.

A cursory search reveals that polygyny was practiced, at least among elites, in almost all of the cultures known at the dawn of the historic era in Eurasia, although sometimes secondary wives or concubines were permitted only if they held slave status.  The Mycenean Greeks and Trojans, the ancient Macedonians, the Hittites, the Celts, early Nordic pagans, the first century Merovingians and Carolingians, the early Jews, the Roman, the pre-Christian Slavs, Muslims and pre-Muslim pagans, the pre-Communist Chinese, the early Aryans (a controversial point to some), the pre-Zoroastrian religion of Iran, and the elites of ancient Egypt.  There is genetic evidence of prior polygamy in French Basques and many other modern populations.  Also, monogamous societies are a decided minority among anthropologically attested cultures (only about one in six fit that description).  The status of Avesta era Iran is unclear and may have been monogamous and polygamy was expressly forbidden by the legal code of the Visigoths as of the 1st century according to Tactius (although not by any other the other Germanic tribes of that time).  There is evidence of religious fertility rights involving annual divine marriages of Sumerian kings to temple priestesses in addition to their ordinary wives, but evidence regarding polygamy per se in ancient Sumeria is sketchy.

Still, as a Bayesian prior, we would strongly expect Chalcolithic and Bronze Age cultures of Europe to have been polygynous rather than monogamous.  Monogamy is a practice that seems to have been exceedingly rare as a universal rule of any culture prior to the Iron Age.  An academic case has been made for Proto-Indo-European monogamy, but I personally find it implausible given the prevalence of polygyny in the earliest attested Indo-European cultures.

However, if polygny was a practice largely restricted to kings and chiefs, rather than being widespread, the number of wives was not too immense, and the scale of chiefdoms was fairly large in this time period, the cumulative impact of this practice on the gene pool may have been fairly modest and this may not have been an important factor in explaining the expansion of Y-DNA lineages.  The considerable diversity of polygynous practices makes it hard to model.

Polygny isn't absolutely necessary to estimates of Y-DNA R1b expansion rates in Europe which, while high, were not so high that they couldn't be consistent with a sustained period of largely monogamous expansion by a population without the male selective androcide and without the polygyny that necessarily goes hand in hand with that pattern.

The expansion of mtDNA H in Europe apparently coincident to the R1b expansion (in both cases from a source in Iberia whatever prior sources those expansion had, perhaps the Bell Beaker population*) did not bring mtDNA H to quite as dominant a position a Y-DNA R1b in Western Europe, but without knowing what the mtDNA breakdown of the source population for the R1b expansion in Western Europe was and having detailed mtDNA haplogroup subtype data for non-mtDNA H haplogroups in Western Europe (both available to some extent but beyond the scope of this post), it is hard to know how much of the Y-DNA R1b expansion involved displacement of indigenous men and how much involved expansion of multiple mtDNA lineages in the source population complemented by bride exchange between growing and stagnant communities on an equitable basis.

UPDATED:

The 2008 analysis by Hammer et al. of sex differences in breeding ratio in six populations including the French Basque is most on point (internal references to figures and other papers of the paper omitted without acknowledgement of the omissions):
Our method uses a population genetic model (i.e., the coalescent) to account for the inherent uncertainty in estimating diversity and divergence rates from sequence data. For three out of six populations (Basque, Melanesians and Mandenka), the 95% confidence intervals for the ratio of X-linked and autosomal effective population sizes does not include 0.75 [Ed. the expected value in a monogamous population model.] (p = 0.001, 0.005 and 0.030 for the Basque, Melanesians and Mandenka, respectively). One interpretation of these results is that there is strong evidence for an unequal female and male Ne in at least three of our six populations, with estimates of the breeding sex ratio (i.e., the effective size of females to males) ranging from 2.1 in the San to 12.5 in the Basque. If the observed differences in nucleotide variability on the X chromosome and autosomes are caused by long-term (demographic) processes, then the estimates of Nx/Na presented will be highly correlated due to shared population history. When we use the intersection of all six confidence intervals (0.87–1.02) to estimate the range of Nx/Na values that are consistent with the data from all six populations, we estimate the range of the breeding sex ratio to be 2.4–8.7. We also note that even with a conservative Bonferroni correction, a 1:1 breeding sex ratio is rejected in two out of six populations.

We also employ a separate method for estimating the breeding sex ratio in each population that does not allow for intra-locus recombination but does permit independent mutation rates across loci. This method produces similar results to those described above, with estimates of the ratio of female to male effective population size ranging from 1.8 in the San to 14.0 in the Basque. We interpret this as additional evidence that the unusual patterns observed in our data are real and require explanation.
Naively, this suggests that the ancestors of the French Basque were, at some point, massively polygnyous.  There is an alternative explanation, however.

This is that while the pairings in some ancestral era of the French Basque were mostly of French Basque men and indigeneous women, that rather than being massively polygamous, the French Basque men were simply highly inbred and closely related. In other words, so many of the men were genetically identical or nearly so that they look in effective population size genetic statistics, like a single man.

This could happen, for example, if a Basque founding population were composed almost entirely of a group of migrating men who were all part of a patrilocal extended family group that had expanded fairly rapidly elsewhere in their insular community of origin, prior to migrating together to Iberia, while the indigeneous women of the pre-existing megalithic culture were highly diverse due to many generations of fairly long distance bride exchanges prior to the Basque male founding population arrived.

For example, suppose the early copper working in Central Europe or West Asia were organized on a patrilocal clan basis, so that only sons of existing (and prosperous and expanding) copper workers could become copper workers.  Then, suppose that word of rich unexploited copper resources in Southern Iberia reached them.  Large numbers of these men from the copper worker extended family might leave, en masse, in a Gold rush style, male dominated copper rush to Iberia that in this case proved to be not a myth, but a reality that led to multiple generations of migration to Iberia from the same patrilocal clan as those who arrived there prospered and married local women.

While the founding Bell Beaker population which I claim is proto-Basque probably was polygamous to some extent like other pre-Iron Age cultures, it seems highly unlikely to me that Basque men had, on average, for several hundred years, a dozen or so wives each.

After all, while the growth rate of the Western European specific clades of Y-DNA are phenomenal, they are merely at the very high end of what could be attained with a very fertile monogamous scenario.  Mildly polygamous family patterns (e.g. 1.5 or so wives per Basque man on average) fit the observed rates of R1b expansion much better than massively polygamous family patterns in which each man could conceivably have had as many as a dozen or more wives on average for ten or twelve generations in a row.

The patrilocal clan originating founding population scenario that I suggest seems like a more likely source for most of the disparity in effective male and female population sizes that is evident in French Basque DNA.

END UPDATE

* The old conventional wisdom was that the Bell Beaker people had predominantly a cultural impact on Western Europe and was largely a trading civilization with only modest demographic impact, while the subsequent Indo-Europeans may have had a major demographic impact.  Maju (who is blogging again after taking a hiatus for a while) has argued this case multiple times.  Ancient DNA evidence and modern population genetic studies seem to be favoring the reverse scenario - one with a major Bell Beaker demographic impact and a fairly shallow additional Indo-European wave demographic impact - although the ancient DNA data are still too fragmentary and inconclusive to definitively favor either scenario or some third scenario.  There is also some debate over the question of whether the Indo-Europeans may have themselves been Indo-European linguistically and culturally, a position that I consider a minority and disfavored view.  In my view, the Bell Beaker people were probably Vasconic (i.e. proto-Basque), and were themselves probably a second rather than an initial wave of farming people to expand in Western Europe.

10 comments:

Tienzen said...

Excellent article.

Maju said...

"Maju (...) has argued this case multiple times. Ancient DNA evidence and modern population genetic studies seem to be favoring the reverse scenario - one with a major Bell Beaker demographic impact and a fairly shallow additional Indo-European wave demographic impact (...)"

And I still argue it. What we are seeing in Central European Bell Beaker is probably an echo of previous Dolmenic-Megalithic flows (which are not yet studied in archaeo-genetic terms or almost). Bell Beaker almost certainly originate in Bohemia and not in Iberia and is clearly associated in form with Corded Ware and other IE elements, which again were not existent in Iberia (which is the only logical source of the excess H). Bell Beaker is almost invariably found as minority element in other cultural contexts, which are in the remaining Megalithic area West of the Rhine clearly much older, and East of the Rhine within Kurgan (IE) contexts. For all those reasons the data of Central European BB must be informing us of changes that happened before that phase, not in that phase.

In addition to that, R1b is not significant in many Bell Beaker affected areas, such as Italy or North Africa.

As for the polygyny issue, I find it more of an Orientalist-like fantasy than a real argument. Polygyny has been always privilege of few and many historical societies of Europe have been extremely refractory to it (Greeks or Romans for example had no polygyny, nor had Basques). Maybe you mean other type of de facto "polygyny", more akin to rape within context of slavery or serfdom - this can make better sense. But still if, as you claim (I wouldn't bet much for that), R1b (South clade?) expanded with mtDNA H, then a polygynistic explanation is total nonsense, because obviously those men and women must have traveled together and displaced others together as well.

andrew said...

@Maju

First, I am delighted to see that you prehistory/genetics blogging hiatus has ended. I really enjoy your posts.

Second, I've updated the post with a bit more thought about how the data can be fit to a realistic scenario. Like you, I find massive polygyny on the scale suggested by Hammer (2008) to be more of an Orientalist fantasy than a likely historical reality. (I would quibble with your statement that Greeks or Romans or Basques had no polygamy. Greeks certainly had it in spades in the Bronze Age and polygamy via having children with female slaves persisted among both the Romans and Greeks on a widespread basis well into the Iron Age. I am inclined to think that while Basques may have been monogamous or mostly so by the time of their first encounters with Roman historians, that they too were probably polygamous in a population genetic sense (even if there were not formally or legally acknowledged sister wives) in the Bronze Age and earlier although the extent to which this is the case is hard to measure and the genetic evidence certainly does nothing to rebut this possibility. I am indeed referring to de facto polygyny, whatever the social context.

I also think that your case for mtDNA H being local Iberian is pretty strong.

Massive Bell Beaker expansion clearly originated in Iberia even if the culture itself has antecedents in Central Europe.

As my updated and refined analysis suggests, I consider a likely scenario to involve stepwise folk migrations in pursuit of copper resource exploitation to be a particularly likely scenario. The last step was probably of male copper workers from a patrilocal Central European extended family of copper workers to Iberia Gold Rush style to probably provide the Basque male founding population. My guess is that the Central European clan, in turn, probably traced its origins to a copper working clan in West Asia, probably the Caucasus mountains copper workers.

I don't think that the scale of Bell Beaker impact on Italy or North Africa was anything like that of Western and Northern Europe.

As to your assertion that "Bell Beaker is almost invariably found as minority element in other cultural contexts, which are in the remaining Megalithic area West of the Rhine clearly much older, and East of the Rhine within Kurgan (IE) contexts."

I agree that Megalithic cultures in the West and Kurgan in the East were probably Bell Beaker predecessors.

But, I am not convinced that Bell Beaker is a mere minority in the cultural context - something that the hybrid nature of the Bell Beaker male-megalithic female ethnogenesis that would have taken place in Southern Iberian upon their arrival may have concealed. I think that this will increasingly be revealed to have been a misinterpretation of the evidence driven by their differing assumptions at the time, given the strong genetic signal of a post-first wave Neolithic/pre-Iron Age demographic explosion in a region where there is not other candidate source for the expansion since IE comes to late and there is continuity in the Megalithic culture pretty much back to first wave Neolithic farmers.

Maju said...

Guess that you're right about Bronze Age Greece. Searching around for a more authoritative source, I found a paper, which begins as follows:

"Greek and Roman men were not allowed to be married to more than one wife at a time and not meant to cohabit with concubines during marriage, and not even rulers were exempt from these norms".

Besides the dichotomy polygyny/monogyny, which are seldom realized extremes, he proposes an intermediate one "monogyny-polygyny", which is formally monogamous but actually tolerates and often implies a polygynous reality in actual mating.

In the polygynic extreme, monogyny is still the norm, because only the rich and powerful can afford several wives or concubines anyhow ("ecological monogamy"). Also the monogynic extreme is probably rare, at least in a generalized sense, being always exceptions, notably for the powerful (or maybe sexier) guys.

But notice that in any case monogyny rules by default, even if not socially enforced: most (married) men only have one wife at a time and their occasional infidelities are likely to be sterile (prostitution, which typically implies contraceptive practices, including abortion if need be).

While the paper confirms polygyny being somewhat common in the case of historical rulers (Middle East), it also states that under that top layer: "polygyny was normally much more limited in sub-elite circles" but concedes that master-slave relations were not restricted.

As for Ancient Greece:

"Moving on to the Greco-Roman world, elite polygyny looms large in the Homeric
tradition. By the historical period, by contrast, SIUM [monogyny] was firmly established as the only legitimate marriage system: polygamy was considered a barbarian custom or a mark of tyranny and monogamy was regarded as quintessentially “Greek.”"

So you seem to be right on the issue of Mycenaean Greece, although I guess it was limited to some elite types, even if only for "ecological" reasons (i.e. not enough women for every man if polygyny would be generalized).

Then he reckons that concubinage, as well as sex with slaves and prostitutes, remained part of that "monogynic" Classical Greece.

Further West however:

"There is no sign of an early polygamous tradition in Rome".

"As in Greece, sexual relations of married men with their own slave women were not unlawful, including relationships that resulted in offspring. Formal recognition of the latter was optional but not unknown".

As far as I know their Etruscan neighbors were also monogynic and way more gender egalitarian than Romans, who considered this (relative) gender equality and its expressions (such as men and women dining together in the same couch) as "immoral".

...

Maju said...

...

In real cases, it was much more likely that the "polygyny" of conquests, which I don't really question, at least as possibility, was much more a matter of unequal relations, including rape and quasi-rape, much more like slavery than marriage or other relatively equal forms of union.

This surely not just included democide of conquered men (in short and long term) but also import of captive women from surrounding areas of plunder. Pre-Christian Magyars for example are attested as doing exactly that in Germany: killing every man and all women too old or too young, and bringing the rest as captives to Hungary. This may be an extreme case but when you read about the Castilian conquest of America from non-whitewashing sources, you realize that it happened often. However most historians will agree that, compared with early Magyars, most "barbarian" invaders of Europe were true gentlemen, so to say, even the much feared Vikings.

But calling all that "polygyny" sounds a bit strange in any case. It's rather slavery for some of the defeated. On the other hand, the conquerors would still need plenty of men for work, either as slaves, serfs or other forms of submission, so these extreme practices cannot be considered normal. Even the Castilians tried to keep as many Natives as they could alive, especially after the first and most brutal period, succeeding in many cases, and in others failing because of the brutality of exploitation combined with the bacterian shock.

America anyhow got most of its European blood in the 19th century (and some of the 20th), so it's more like a second wave. I've heard the term for that period of "the whitening of Venezuela" or whatever other Latin American country, because earlier Europeans were quite few. This is true even for North America in fact.

Maju said...

"Massive Bell Beaker expansion clearly originated in Iberia even if the culture itself has antecedents in Central Europe".

There is no such expansion. The area of incidence of the "International" or "Maritime" Bell Beaker is not so large and rather responds to trade. The area of VNSP findings reaches to near Brittany and certainly not to Germany.

Not sure how good is this map but it's at least the less bad one I could find:

http://en.wikipedia.org/wiki/File:Beaker_culture_diffusion.svg

At the very least the arrows are roughly correct and you can see that the one stemming from Portugal only reaches to Britain, not further.

There was a counter-tide with the 2nd period of BB, centered in Portugal and more civilized, but the phenomenon is clearly of Central European origins. Just consider that the oldest beakers, even in Iberia, are all of the rather rustic "Corded" style, which clearly correlates with Corded Ware Culture.

There were three phases:

1. Corded type from Bohemia, expanding westwards and southwards.

2. Maritime (International) type from Portugal expanding (less extensively northwards).

3. Regionalization with plenty of local types (and some return of the centrality to Bohemia).

I see this phenomenon like a "guild", possibly with religious implications (a self-segregated, yet influential, religious community like Medieval Jews). The previous cultures continued existing almost as if these people would have not arrived at all. This applies to Germany, to Iberia, to Britain and every other place I know of.

The Rhine IE/Vasconic boundary created by Corded Ware (and Artenac) expansion was maybe weakened but not really erased yet. That would only happen with the late Bronze and Iron Age IE expansions (Urnfields-Hallstatt-La Tène) of Celts, Italics and maybe others.

"... migrations in pursuit of copper resource exploitation to be a particularly likely scenario".

Doesn't make sense:

1. Copper (and gold and silver) was exploited before BB: these guys did not open the Chalcolithic but actually close it.

2. Copper was quite abundant, particularly in Iberia, a mineral-rich region in antiquity. This made Iberia (the Hesperides) a colonial target but seemingly only since the Bronze Age, when tin (a much rarer resource, then concentrated in Galicia and Cornwall) became a strategical mineral.

3. At least in the Basque Country, the presence of brachicephalic types (believed to come from France or Central Europe) is not seen until the Bronze Age, when they do show up in mining contexts (Bidasoa: ~30%). These are indeed pre-Urnfields but also clearly and very much so post-BB. These prospectors or active miners did not bring any particular culture with them anyhow.

...

Maju said...

...

"The last step was probably of male copper workers from a patrilocal Central European extended family of copper workers to Iberia Gold Rush style to probably provide the Basque male founding population".

Basque genetics are stable since Neolithic. Basques are Neolithic remnants.

I don't understand why you insist on beating this dead horse, really.

"My guess is that the Central European clan, in turn, probably traced its origins to a copper working clan in West Asia, probably the Caucasus mountains copper workers."

And now the Caucasus dead horse... please!

"I don't think that the scale of Bell Beaker impact on Italy or North Africa was anything like that of Western and Northern Europe".

In this you may be right but the same ca be said of Megalithism, much more likely to be a demic force.

"I agree that Megalithic cultures in the West and Kurgan in the East were probably Bell Beaker predecessors".

Megalithism existed also in Central and Northern Europe but only before Kurgan expansion. They are opposed phenomena. BB may be a link between them but does not override either.

I wonder if Indoeruopeanized Megalithic peoples of Germany and the Netherlands adopted the new BB "religion" (as some sort of solution to their identity conflict) and that's why they had so easy to expand westwards across the otherwise very sharp ethno-cultural border of the Rhine. But this I reckon is speculative.

"I am not convinced that Bell Beaker is a mere minority in the cultural context"

They were, at least in most cases. In Iberia certainly they appear only in other cultural contexts and often the vessels and other material markers (such as the V-perforated buttons) show up isolated of any sort of burial, suggesting that they were being traded and used in non-religious (i.e. non-cultural) contexts.

Any serious BB study should answer these questions for each region or district:

1. BB burials, what percentage? If, as happens most often, they are just a minority, we can't talk of BB culture but just sub-culture or phenomenon.

2. BB materials, what types and what frequency in proportion to BB burials? Where materials are common but burials scarce or unknown, as often happens, BB sub-culture had only a very slim presence.

3. Where BB is not dominant (most places), which other culture it manifests within?

Only that way BB can be understood for what it really was.

[in Central Europe, I understand] "there is continuity in the Megalithic culture pretty much back to first wave Neolithic farmers".

That is indeed an issue but only in some areas. In Denmark or Britain or even Low Germany/Netherlands it should be not, because there are cultural changes associated with Megalithism (which may include even the very arrival of farming, at least in Britain and very likely in Denmark/Low Germany). The demographic explosion associated with Megalithism in these North Sea areas must be considered. In turn, in the Danubian areas, there was also a sharp decay in population.

In Poland this sharp decay was countered by Kurgan expansion but in Germany the signal of recovery is late and weaker (it'd be interesting to have regional data). BB does not correspond with any demographic expansion but rather decline, at least in those four areas.

andrew said...

So, bottom line, are you arguing for an Epipaleolithic expansion of the Western European specific clades of R1b, or a Megalithic one, or something else?

How does your theory produce a fit to the modern distribution of R1b?

andrew said...
This comment has been removed by the author.
Maju said...

Do you want to reach to the goal line even before starting? We can't say anything about R1b because we do not have enough archaeogenetic data. We only know something about the patterns of modern distribution and these, in Europe, suggest quite strongly a major "South" clade expanding from southern France most likely and a lesser "North" clade expanding from Netherlands or somewhere nearby. How does this fit with BB or whatever other Neolithic conjecture? It does not fit.

So I would say that the South clade looks Magdalenian, while the North clade is another story (Doggerland, Hamburgian and such). I'm arguing for it? Well, I do not know for sure: I'd like to have more direct archaeogenetic data to judge but I don't. But neither do you, nor Polako, nor Dienekes, nor Razib, nor Jean Manco, nor Hammer... so it's all speculative blah-blah.

What I know a bit more is about mtDNA H, which sometimes is related to R1b, or so say some (I'm not so obsessed about patrilineages, so I prefer to call things by its proper name: mtDNA H). And (1): this lineage existed in Europe before Neolithic, at least in SW and NE Europe, (2) it's star-like signature has nothing to do with Neolithic: it's much older, (3) it expanded to Germany (which is not all Europe but just a crossroads over there) first with Neolithic (from Balcans or Danube, less investigated areas) and later, more intensely, from somewhere else (probably the SW and probably within Megalithic flows).

That's what I can say about mtDNA H, which I am not certain of being able to relate with R1b. If anything I'd relate the SW-originated expansion (Megalithic) of H with some of the spread of R1b-S116 and not all kinds of R1b, just that one. In fact I'd dare say that only some subclades of S116 were involved and none of them looks Iberian by origin, but "French".

While R1b-S116 is the largest R1b subclade, it is relatively minor in Germany, Denmark or Netherlands anyhow: less than 50% (of R1b, i.e. less than 25% of total) in the South (including Czech Rep. but not Austria: ~20% per Myres 2010, all of it the "Irish clade") and even quite less in the North. It's more important in Britain and, of course, in France, Iberia and Ireland, where it is overly dominant.

Whether the R1b-U106 (North clade) expansion was related to that of the South in some way, I really do not know. They were surely rather unrelated processes.