23andMe On My Y-DNA With Conjectures Regarding Indo-European Linguistics

Overview

I learned a bit about my Y-DNA from the 23andMe summary. But, I read their sources and researched more. This gave rise to some insights and conjectures.

* My Y-DNA haplogroup E-V13 spread across Europe as part of the Indo-European expansion in the Bronze Age, not as part of the Neolithic expansion of farmers from Anatolia (or at some other time).

* E-V13 is found at low frequencies everywhere that multiple distinct waves of Indo-European expansion reached. It is found in European Corded Ware derived populations where the expansion was dominated by Corded Ware haplogroups of Y-DNA R1a-M458. It is found in similar percentages in European Bell Beaker derived populations where the expansion was dominated by Bell Beaker haplogroups of Y-DNA R1b-L51. It is found at quite high percentages among the Greeks. It is found among Indo-European language speakers in the Caucuses. It is found in similar percentages in Indo-Iranian (like Iranians and the Kurds) peoples dominated by R1a-Z93 and R1b-Z2103. E-V13 has been found as far away as central Siberia, near the Altai, a region also known to have been settled by Bronze Age Indo-Europeans, which could have sources from either Indo-Iranians or Tocharians. But, it is not found among linguistically Indo-Aryan populations, even though these languages are often considered a sub-family of Indo-Iranian languages with otherwise similar Y-DNA profiles - a fact that can, in theory, shed light on the paternal ancestry diversity, and likely size within wide error bars, of the male founding population of Indo-Aryans in absolute terms and relative to the entire Indo-Iranian population.

* The two of the defining mutations of E-V13 date to ca. 3500 BCE and 2100 BCE (the date for TMRCA) and those mutations are shared by all living members of E-V13 in all of these places. The expansions that E-V13 was a part of probably took place from 2100 BCE to 1800 BCE. This means that E-V13 men must have assimilated into each of these expanding Indo-European populations (to an extent greater than other Indo-European populations did with each other), rather than just one of them. 

* E-V13 differentiated from E-M78 its parent haplogroup ,which was common in the pre-Neolithic and Neolithic Levant, no earlier than the beginning of the Holocene era (8000 BCE), and probably quite a bit later.

* The distribution of E-V13 in Anatolia and the vicinity, interpreted in light of the larger historical context, can help us evaluate and compare different proposals regarding the timing and character of the Anatolian language family's origins and its relationship to the larger Indo-European language family. E-V13 is found at much higher frequencies in Western Anatolia than in Central Anatolia. But, it is found among Kurds, in the Caucasus region, and in Iran. 

* Critically, if Anatolia has some form of Y-DNA R that is associated with Indo-European demic migration, but not E-V13, this suggests that Anatolian languages could have broken off pre-3500 BCE, could have been excluded due to founder effects in a manner similar to that of the Indo-Aryans, or could (especially if Y-DNA R is rare) signal that the Anatolian languages were a product of elite dominance with only slight demic impacts. 

* The analysis is complicated by the fact that we have limited information about the presence of Anatolian languages (or Indo-Europeans more generally based upon archaeology without written language evidence) in Anatolia prior to 1800 BCE. Also, after the Hittite Empire fell, Anatolia spend long periods of time in which it was ruled by Greeks (and to a lesser extent, some other Indo-Europeans), especially in Western Anatolia. Even modern Turkish history, like the Armenian genocide and the expulsion and killing of many Greek speaking Anatolians in the 20th century, complicate trying to make inferences about the prehistory and ancient history from the genetic character of modern Turkish people. Our knowledge of the history of the linguistically Armenian people, whose language doesn't fit neatly into categories for classifying Indo-Europeans languages, which adjacent to the historical area where Anatolian languages were spoken, is even more patchy. It is also complicated by the fact that many of the Turkic people who caused the country of Turkey to speak Turkish also have Indo-Iranian admixture.

The Y-DNA mix in Turkey as a whole is as follows:

• R1b=15.9%
• R1a=6.9%
• E-M35=10.7% (E-M78 and E-M123 accounted for all E representatives in the sample, except a single E-M81 individual). E-V13 is a sub-haplogroup of E-M78.
• G=10.9%
• J2=24% - J2 (M172)
• J1=9%
• I=5.3%
• K=4.5%
• L=4.2%
• N=3.8%
• T=2.5%
• Q=1.9%
• C=1.3%
• R2=0.96% 

Others markers than occurs in less than 1% are H, A, E3a , O , R1*.

The first Neolithic farmer populations of Anatolia (at least in Western Anatolia but probably more broadly) were dominated by Y-DNA G. Y-DNA R1b, R1a and E-M78 were almost surely all intrusive to Anatolia in the post-Neolithic era, and are associated with Indo-Europeans.

My Y-DNA

The ancestry narratives at 23 and Me evolve over time, in part due to new research and in part due to more effort on the company's part.  Generally speaking, the information provided by the company is accurate, albeit simplified. 

Their article also clarifies an issue I had pondered and had not clarified for myself in the literature, which is whether Y-DNA E-V13 spread from the Balkans to the rest of Europe with the first farmers (apparently, it mostly did not) or around the time of the demographic transition that took place around the time of the Indo-European expansion in the Eneolithic Age (i.e. Copper Age) and early to Middle Bronze Age (apparently, it mostly did), or at some later time such as the Iron Age (apparently, it mostly did not). 

This conclusion not too surprising given that my paternal ancestors as of the 1700s to 1847 were from central Germany (near the former West German and East German border and not too far to the North or South along that former border). 

This is an area which experienced in the Bronze Age significant, although not complete, population replacement from the original Linear Pottery Culture derived first farmers of most Anatolian origins who in turn largely replaced the European hunter-gathers that preceded them in the Mesolithic era as Europe was repopulated after the Last Glacial Maximum, while admixing somewhat with them. 

The defining Y-DNA haplogroups of those mass migration and replacement episodes that followed in this area through the Bronze Age were clades within Y-DNA R1a and Y-DNA R1b, but Y-DNA E-V13 was apparently a minor component of that mix that came along for the ride in that wave of migration.

Here is what mine says right now (in part, to provide a baseline and in part, out of general interest, my commentary is in italics and brackets):

Haplogroup A 275,000 Years Ago 

The stories of all of our paternal lines can be traced back over 275,000 years to just one man: the common ancestor of haplogroup A. Current evidence suggests he was one of thousands of men who lived in eastern Africa at the time. However, while his male-line descendants passed down their Y chromosomes generation after generation, the lineages from the other men died out. Over time his lineage alone gave rise to all other haplogroups that exist today.

[Ed. The narrative omits, for simplicity's sake, the emergence of Y-DNA haplogroup B from Y-DNA haplogroup A, and of Y-DNA haplogroup DE from Y-DNA haplogroup B, which gives rise to Y-DNA haplogroup DE (and also to Y-DNA haplogroup CF). It also dodges the issue of the precise location of the B to CF and DE split, and the DE split into Y-DNA haplogroup D, found mostly in the Andaman Islands, Tibet, North Asia, at low frequencies in South Asia, and Japan, and Y-DNA haplogroup E, found mostly in Africa with probably later migrations mostly to Europe via both Western and Eastern and central maritime migrations takes place, and the related controversy of the minority position that Y-DNA haplogroup E back migrated from the Middle East (or beyond) to Africa. 

The map used to illustrate the article suggests, without definitively saying so or being very specific, that the B to DE split took place in the vicinity of Ethiopia and that the DE to E split took place in the vicinity of South Sudan. The map also suggests that the first modern human man emerged in East Africa and that the A to B split also took place in that vicinity.] 

leads to [Ed. I use the words "leads to" in lieu of the line running from top to bottom in the margin of the original.] 

Haplogroup DE-M145 76,000 Years Ago 

leads to 

Haplogroup E-M96 73,000 Years Ago 

leads to 

Haplogroup E-M78 23,000 Years Ago.  

Origin and Migrations of Haplogroup E-M78 

Your paternal line stems from the common ancestor of haplogroup E-M78, a branch of E that dates back approximately 24,000 years. The earliest carriers of the E-M78 lineage likely lived in a population that moved from eastern Africa into northeastern Africa about 14,000 years ago, during the final days of the Ice Age. From northeastern Africa, their descendants expanded to the west between the Sahara and the Mediterranean coastline, and to the east out of Africa into the Middle East, where E-M78 men remain common. 

Today, men bearing this haplogroup are also common in southern Europe, including in the Balkans, Iberia, and Italy. In Greece, Bulgaria, and Albania, between 15% and 30% of men bear haplogroup E-M78. Their ancestors were likely relatively late arrivals to the region. While some branches of haplogroup E were carried into Europe nearly 8,000 years ago, recent research suggests that the major spread of E-M78 occurred in the last 5,000 years or so during the Bronze Age. Bronze Age cultures learned to smelt tin and copper to create beautiful and complex bronze items like hardier tools and weapons. They journeyed along river waterways in the Balkans and spread into east-central Europe. Today, men from Ukraine, Hungary, Romania, and Slovakia all carry E-M78 at levels of nearly 10%. 

While the majority of E-M78 European males trace their recent ancestry to Turkey and the Middle East, some men carrying E-M78 from Spain, Italy and Greece trace their ancestry directly from North African populations, probably within the last 4,000 years. The ancestors of these men must have sailed across the Mediterranean Sea and settled in communities along the European coast. 

leads to 

Haplogroup E-V13 11,000 Years Ago 

Your paternal haplogroup, E-V13, traces back to a man who lived approximately 11,000 years ago. 

[Ed. The 11,000 years ago date is shortly before the advent of the Holocene era which is usually considered to begin with the Fertile Crescent Neolithic Revolution about 10,000 years ago.] 

That's nearly 440.0 generations ago! E-V13 is relatively common among 23andMe customers. Today, you share your haplogroup with all the men who are paternal-line descendants of the common ancestor of E-V13, including other 23andMe customers. 1 in 43 23andMe customers share your haplogroup assignment. 

[Ed. The 23 and Me logo for E-V13 below, is a picture of a small castle on an island in a lake in Slovenia, which is in the Balkans within Southeast Europe.]

References

• Al-Zahery N et al. (2003). "Y-chromosome and mtDNA polymorphisms in Iraq, a crossroad of the early human dispersal and of post-Neolithic migrations." Mol Phylogenet Evol. 28(3):458-72.

• Alonso S et al. (2005). "The place of the Basques in the European Y-chromosome diversity landscape." Eur J Hum Genet. 13(12):1293-302.

• Arredi B et al. (2004). "A predominantly neolithic origin for Y-chromosomal DNA variation in North Africa." Am J Hum Genet. 75(2):338-45.

• Balanovsky O et al. (2008). "Two sources of the Russian patrilineal heritage in their Eurasian context." Am J Hum Genet. 82(1):236-50.

• Barać L et al. (2003). "Y chromosomal heritage of Croatian population and its island isolates." Eur J Hum Genet. 11(7):535-42.

• Beleza S et al. (2005). "The genetic legacy of western Bantu migrations." Hum Genet. 117(4):366-75.

• Bereir RE et al. (2007). "Co-introgression of Y-chromosome haplogroups and the sickle cell gene across Africa's Sahel." Eur J Hum Genet. 15(11):1183-5.

• Bosch E et al. (2001). "High-resolution analysis of human Y-chromosome variation shows a sharp discontinuity and limited gene flow between northwestern Africa and the Iberian Peninsula." Am J Hum Genet. 68(4):1019-29.

• Bowden GR et al. (2008). "Excavating past population structures by surname-based sampling: the genetic legacy of the Vikings in northwest England." Mol Biol Evol. 25(2):301-9.

• Capelli C et al. (2006). "Population structure in the Mediterranean basin: a Y chromosome perspective." Ann Hum Genet. 70(Pt 2):207-25.

• Cinnioğlu C et al. (2004). "Excavating Y-chromosome haplotype strata in Anatolia." Hum Genet. 114(2):127-48.

• Coia V et al. (2004). "Binary and microsatellite polymorphisms of the Y-chromosome in the Mbenzele pygmies from the Central African Republic." Am J Hum Biol. 16(1):57-67.

• Cruciani F et al. (2002). "A back migration from Asia to sub-Saharan Africa is supported by high-resolution analysis of human Y-chromosome haplotypes." Am J Hum Genet. 70(5):1197-214.

• Firasat S et al. (2007). "Y-chromosomal evidence for a limited Greek contribution to the Pathan population of Pakistan." Eur J Hum Genet. 15(1):121-6.

• Francalacci P et al. (2003). "Peopling of three Mediterranean islands (Corsica, Sardinia, and Sicily) inferred by Y-chromosome biallelic variability." Am J Phys Anthropol. 121(3):270-9.

• Hurles ME et al. (2005). "The dual origin of the Malagasy in Island Southeast Asia and East Africa: evidence from maternal and paternal lineages." Am J Hum Genet. 76(5):894-901.

• Karafet TM et al. (2002). "High levels of Y-chromosome differentiation among native Siberian populations and the genetic signature of a boreal hunter-gatherer way of life." Hum Biol. 74(6):761-89.

• Kayser M et al. (2005). "Significant genetic differentiation between Poland and Germany follows present-day political borders, as revealed by Y-chromosome analysis." Hum Genet. 117(5):428-43.

• Luis JR et al. (2004). "The Levant versus the Horn of Africa: evidence for bidirectional corridors of human migrations." Am J Hum Genet. 74(3):532-44.

• Marjanovic D et al. (2005). "The peopling of modern Bosnia-Herzegovina: Y-chromosome haplogroups in the three main ethnic groups." Ann Hum Genet. 69(Pt 6):757-63.

• Nasidze I et al. (2004). "Mitochondrial DNA and Y-chromosome variation in the caucasus." Ann Hum Genet. 68(Pt 3):205-21.

• Onofri V et al. (2007). "Y-chromosome genetic structure in sub-Apennine populations of Central Italy by SNP and STR analysis." Int J Legal Med. 121(3):234-7.

• Pericić M et al. (2005). "High-resolution phylogenetic analysis of southeastern Europe traces major episodes of paternal gene flow among Slavic populations." Mol Biol Evol. 22(10):1964-75.

• Qamar R et al. (2002). "Y-chromosomal DNA variation in Pakistan." Am J Hum Genet. 70(5):1107-24.

• Regueiro M et al. (2006). "Iran: tricontinental nexus for Y-chromosome driven migration." Hum Hered. 61(3):132-43.

• Rosa A et al. (2007). "Y-chromosomal diversity in the population of Guinea-Bissau: a multiethnic perspective." BMC Evol Biol. 7:124.

• Rosser ZH et al. (2000). "Y-chromosomal diversity in Europe is clinal and influenced primarily by geography, rather than by language." Am J Hum Genet. 67(6):1526-43.

• Semino O et al. (2002). "Ethiopians and Khoisan share the deepest clades of the human Y-chromosome phylogeny." Am J Hum Genet. 70(1):265-8.

• Sengupta S et al. (2006). "Polarity and temporality of high-resolution y-chromosome distributions in India identify both indigenous and exogenous expansions and reveal minor genetic influence of Central Asian pastoralists." Am J Hum Genet. 78(2):202-21.

• Shen P et al. (2004). "Reconstruction of patrilineages and matrilineages of Samaritans and other Israeli populations from Y-chromosome and mitochondrial DNA sequence variation." Hum Mutat. 24(3):248-60.

• Tambets K et al. (2004). "The western and eastern roots of the Saami--the story of genetic "outliers" told by mitochondrial DNA and Y chromosomes." Am J Hum Genet. 74(4):661-82.

• Tishkoff SA et al. (2007). "History of click-speaking populations of Africa inferred from mtDNA and Y chromosome genetic variation." Mol Biol Evol. 24(10):2180-95.

• Wells RS et al. (2001). "The Eurasian heartland: a continental perspective on Y-chromosome diversity." Proc Natl Acad Sci U S A. 98(18):10244-9.

• Wood ET et al. (2005). "Contrasting patterns of Y chromosome and mtDNA variation in Africa: evidence for sex-biased demographic processes." Eur J Hum Genet. 13(7):867-76.

Your haplogroup migrated in large numbers from the Balkans into Europe about 4,500 years ago, triggered by the beginning of the Balkan Bronze Age. During this migration, members of your haplogroup mainly followed rivers connecting the southern Balkans to northern-central Europe. Technological leaps often cause lineages to grow dramatically in numbers and in geographic range. The development of Bronze technology may have given men in your lineage a competitive advantage over other men, causing your lineage to proliferate and become widespread.

[Ed. This implies that my patrilineal ancestors were Southeastern European hunter-gathers in smallish semi-nomadic tribes and bands for about 2,000 to 3,000 years. But, Trobetta (2015) which it cites, favors a date that would start the E-V13 split at 8.1 kya which would be at the start of the Balkan Neolithic without a period as Southeastern European hunter-gathers that an 11 kya data would imply, which I am inclined to favor as a better motivated narrative. The key sentence of Trobetta (2015) states that: "The TMRCA of E-V13 chromosomes (8.1 ka; 95% CI: 5.6–10.8 ka) is consistent with a previous hypothesis about a post-Neolithic expansion of this haplogroup in Europe (Cruciani et al. 2004, 2007)." Then, they would have been Neolithic farmers or herders somewhere in the Balkans using the Fertile Crescent Neolithic package of crops, probably linked to the Cardial Pottery Neolithic wave or to earlier culture diffusion of food production technology from Anatolia and the Levant, while living in smallish villages for about 3,500 to 4,500 years. Then 4,500 years ago, they would have joined to Bronze Age wave of migration to the rest of Europe led by Indo-Europeans with ancestry derived from the Pontic-Caspian steppe. This is based mostly on Cruciani (2007) which, making inferences from 517 modern Y-DNA samples states in the abstract that: "A single clade within E-M78 (E-V13) highlights a range expansion in the Bronze Age of southeastern Europe, which is also detected by haplogroup J-M12. Phylogeography pattern of molecular radiation and coalescence estimates for both haplogroups are similar and reveal that the genetic landscape of this region is, to a large extent, the consequence of a recent population growth in situ rather than the result of a mere flow of western Asian migrants in the early Neolithic. Our results not only provide a refinement of previous evolutionary hypotheses but also well-defined time frames for past human movements both in northern/eastern Africa and western Eurasia." Eventually, they would have ended up in what is now central Germany sometime after that. The localization of my French-German ancestry in another part of the 23andMe ancestry report suggests that those ancestors may have started out in Switzerland or nearby in what is now Southern Germany and then made their way north to the point were our family's genealogy records begin in Central Germany, probably sometime in the last 800 years or so, with the more recent 300 years or so of that time frame being favored. For example, they might have been Swiss Protestants who migrated to Protestant Germany for economic opportunity or to flee some uncomfortable scandal in Switzerland. Poznik (2016) adds almost nothing to the analysis.]

References

• Cruciani F et al. (2007). "Tracing past human male movements in northern/eastern Africa and western Eurasia: new clues from Y-chromosomal haplogroups E-M78 and J-M12." Mol Biol Evol. 24(6):1300-11.

• Poznik GD et al. (2016). "Punctuated bursts in human male demography inferred from 1,244 worldwide Y-chromosome sequences." Nat Genet. 48(6):593-9.

• Trombetta B et al. (2015). "Phylogeographic Refinement and Large Scale Genotyping of Human Y Chromosome Haplogroup E Provide New Insights into the Dispersal of Early Pastoralists in the African Continent." Genome Biol Evol. 7(7):1940-50.

The 23 and Me materials and the references that they cite don't discuss ancient Y-DNA E-V13 but there are some examples of it.

A lengthy and well annotated analysis of Y-DNA E-V13 can be found at Eupedia (see below). The narrative portion of that entry which is better reasoned and better sourced suggests a somewhat different history of E-V13.

E-V13 And Indo-European Historical Linguistics

This narrative corroborated by ancient DNA in places, draws from the fact that E-V13 was present a minor subcomponent of at least, the Y-DNA gene pool of the Corded Ware, the Bell Beaker and the Indo-Iranian branches to Indo-European expansion, although apparently not the Indo-Aryan branch, divisions of the Indo-European expansion that each have their own dominant Y-DNA R haplogroups. 

Thus, the assimilation of E-V13 into the Indo-Europeans must have happened in or near the shared homeland of these three Indo-European cultures immediately prior to their respective and roughly contemporaneous expansions, in or near the Pontic Caspian steppe. Indeed E-V13 is good evidence of such a shared pre-expansion homeland. 

The time depth of that is well bracketed by the genetic time depth of E-V13 diversification, since "all the modern members of E-V13 . . . descend from a later common ancestor who carried the CTS5856 mutation. That ancestor would have lived about 4,100 years ago, during the Bronze Age. Almost immediately afterwards, CTS5856 split into six subclades, then branched off into even more subclades in the space of a few generations." So, you are looking at the phase of Indo-European expansion that included E-V13 taking place over a few centuries starting ca. 2100 BCE and mostly having run its course by ca. 1800 BCE.

The Eupedia states with regard to E-V13 that:

Origins & History

Assimilation of Neolithic European E-V13 by the Indo-Europeans 

For many years the vast majority of academics have assumed that E-V13 and other E1b1b lineages came to the Balkans from the southern Levant via Anatolia during the Neolithic, and that the high frequency of E-V13 was caused by a founder effect among the colonisers. This theory has it that E1b1b people were associated with the development of Neolithic lifestyle and the advent of agriculture in the Fertile Crescent and its earliest diffusion to Southeast Europe (Thessalian Neolithic) and Mediterranean Europe (Cardium Pottery culture). The testing of ancient DNA from the Natufian culture (Mesolithic Levant) and Pre-Pottery Neolithic Levant confirmed a high incidence of haplogroup E1b1b in that region. However, out of 69 Y-DNA samples tested from Neolithic Europe, only two belonged to that haplogroup: one E-M78 from the Sopot culture in Hungary (5000-4800 BCE), another E-M78 (c. 5000 BCE), possibly E-V13, from north-east Spain, and a E-L618 from Zemunica cave near Split in Croatia from 5500 BCE (Fernandes et al., 2016). Whether these E-M78 samples came with Neolithic farmers from the Near East or were already present among Mesolithic Europeans is unclear at present. But in any case E-V13 was definitely not the major Neolithic European lineage it was once alleged to be.

Nowadays E-V13 is the only Mediterranean haplogroup consistently found throughout Europe, even in Norway, Sweden, Finland and Baltic countries, which are conspicuous by the absence of other Neolithic haplogroups like G2a (bar the Indo-European G2a-Z1815), J1 and T (except in Estonia). However, since G2a is the only lineage that was consistently found in all Neolithic sites tested to date in Europe, the absence of Neolithic G2a lineages from Scandinavia and the Baltic implies that no Neolithic lineage survives there, and consequently E-V13 does not date from the Neolithic in the region.

In fact, it has been calculated that E-V13 emerged from E-M78 some 7,800 years ago, when Neolithic farmers were advancing into the Balkans and the Danubian basin. Furthermore, all the modern members of E-V13 descend from a common ancestor who lived approximately 5,500 years ago, and all of them also descend from a later common ancestor who carried the CTS5856 mutation. That ancestor would have lived about 4,100 years ago, during the Bronze Age. Almost immediately afterwards, CTS5856 split into six subclades, then branched off into even more subclades in the space of a few generations. In just a few centuries, that very minor E-V13 lineage had started an expansion process that would turn it into one of Europe's most widespread paternal lineages and reach far beyond the borders of Europe itself, also spreading to the eastern edge of the Mediterranean, the Caucasus, Kurdistan, Iran, and even Siberia.

This data suggests that the fate of E-V13 was linked to the elite dominance of Bronze Age society. The geographic distribution of the six main branches show that E-V13 quickly spread to all parts of Europe, but was especially common in Central Europe. The only Bronze Age migration that could account for such a fast and far-reaching dispersal is that of the Proto-Indo-Europeans. At present the most consistent explanation is that E-V13 developed from E-M78 in Central or Eastern Europe during the Neolithic period, and was assimilated by the R1a and R1b Proto-Indo-Europeans around the time that they were leaving the Pontic Steppe to invade the rest of Europe.

What is surprising with E-V13 is that it is as common in R1a-dominant as in R1b-dominant countries. R1a Indo-European tribes are associated with the Corded Ware culture, which spanned across Northeast Europe, Scandinavia and the northern half of Central Europe. R1b tribes invaded the Balkans, the southern half of Central Europe, and joined up with Corded Ware people in what is now Germany, the Czech Republic and western Poland. If E-V13 was found among both groups, it would have needed to be either assimilated in the Pontic Steppe or very near from it (say, in the Cucuteni-Trypillian culture, around western Ukraine, Moldova and Romania), or at the junction between the two groups in central Europe (e.g. around the Czech Republic). 

The distribution and age of E-V13 clades in central and western Europe are consistent with a dispersal by Hallstatt and La Tène Celts, Italic tribes (including a Roman redistribution) and the later influx of Germanic tribes, particularly the Goths, who may have assimilated additional Proto-Slavic E-V13 lineages in East Germany, Poland and Ukraine before entering the Roman Empire. (=> see also the discussions Was E-V13 a major lineage of Hallstatt Celts and Italics? and Ancient East, West and North Germanics had different Y-DNA lineages).

Amorim et al. (2018) tested the ancient DNA from 6th century Italy and Hungary and identified one E-V13 in Collegno (Turin) who was autosomally fully Italian (not a Lombard immigrant like many other samples tested).

The eastern advance of the Corded Ware culture eventually gave rise to the Sintashta culture in the Ural region, which is the ancestral culture of the Indo-Iranian branch of Indo-Europeans. E-V13's presence in this culture would explain why modern Iranians and Kurds possess E-V13, in addition to R1a-Z93 and R1b-Z2103. E-V13 has been found as far away as central Siberia, near the Altai, a region also known to have been settled by Bronze Age Indo-Europeans.

Due to the scarcity of full genomic sequences available from the Balkans, it is not yet clear when E-V13 expanded in that region. The Indo-European migrations would certainly have brought some E-V13 early on, from circa 2500 BCE. But the history of the region is so complex that there might be many separate branches of E-V13 that each came with a different invasion (e.g. Iranic tribes, La Tène Celts, Romans, Goths, Slavs). The first Indo-European migration to Greece was that of the Mycenaeans from c. 1650 BCE. The Dorians from Central Europe followed from c. 1200 BCE. Both could have brought different subclades of E-V13, and a founder effect or the phenomenon of elite dominance among the ruling invaders might have caused a fast growth of E-V13 lineage in Late Bronze Age and Iron Age Greece.

There are at least three distinct sources of E-V13 in Italy. The first would be the Bronze Age Italic tribes from Central Europe, who in all logic would have possessed at least some E-V13 lineages before they invaded the Italian peninsula. Proto-Italics would have been a predominantly R1b-U152 tribe, but also carried a minority of E-V13, G2a-L140 (L13, L1264 and Z1816 subclades) and J2a1-L70 (PF5456 and Z2177 subclades). The second would be the ancient Greeks, who heavily colonized southern Italy from the 9th century BCE until the Roman conquest in the 3rd century BCE. The third are the Goths. As a Germanic tribe they might have carried a small percentage of E-V13. But that percentage very certainly increased after spending several centuries in Central and Southeast Europe and assimilating Proto-Slavs and Balkanic people before invading Italy. The Goths settled over all the Italian peninsula. They would have brought typically Germanic lineages like I1 and R1b-U106, but also the Proto-Slavic R1a-CTS1211, which is now found uniformly in 1 to 2% of the population. Since R1a-CTS1211 is not originally Germanic, it is likely that the Goths also brought a small but noticeable percentage of assimilated lineages from the Balkans, including E-V13 and J2b1 (I2a1b-CTS10228 would have come later from the East Slavic migrations from Ukraine during the Early Middle Ages, hence its absence from Italy, apart from a few coastal areas facing the Adriatic Sea).

An Indo-European dispersal of V13 subclades would not only explain why E-V13 is present in places like Finland, northwest Russia or Siberia, where Neolithic farmers had a negligible impact, but also why E-V13 is so conspicuously lacking from the Basque country and (central) Sardinia, the two regions of Europe with the highest Neolithic ancestry. Sardinia is also the only part of Europe where Bronze Age Steppe ancestry is virtually absent. The low percentage of E-V13 is coastal Sardinia would be better explained by more recent settlements on the island by the Romans, or even the Goths, who also settled in Sardinia.

The small presence of E-V13 in the Near East could be better explained by the extremely long Greek presence in the eastern Mediterranean from the time of Alexander the Great until the end of the Byzantine domination over the region during the Middle Ages. It would be unthinkable that over 1,500 years of Hellenisation and Byzantine rule in Anatolia and the Levant didn't leave any genetic trace. In Anatolia, E-V13 is found mostly in the western third of the country, the region that used to belong to ancient Greece. The absence of E-V13 from Central Anatolia does not concord with a diffusion linked to Neolithic agriculture. There is clearly a radiation from the Greece (where E-V13 makes up approximately 30% of the paternal lineages) to the East Mediterranean (where the frequency drops to under 5%).

E-V13 and the Anatolian Languages.

The distribution of E-V13 also potentially provides some insight into the Anatolian language controversy, although it does not resolve it. This is because according to Eupedia as quoted at greater length above: 

In Anatolia, E-V13 is found mostly in the western third of the country, the region that used to belong to ancient Greece. The absence of E-V13 from Central Anatolia does not concord with a diffusion linked to Neolithic agriculture. 

While this doesn't correspond with Neolithic agriculture (for this and for other reasons), it could be linked to Indo-European Anatolian language speakers, in addition to the Greek that follows. 

Also, the claimed distribution of E-V13 also seems somewhat self-contradictory because the paper also states that Y-DNA E-V13 is found in "the Caucasus, Kurdistan, Iran," with Kurdistan including much of Southern Anatolia and the Northern Mesopotamian basin, that were previously part of the Hittite Empire or adjacent to it.

Kurdish populations are found, in particular, in parts of Anatolia that correspond geographically relatively closely with the range of the historical Luwian language in the Anatolian language family, and with a significant subsection of the Hittite empire, both of which are shown below (although there is some question regarding whether the modern Kurds may have been migrants to the places where they currently reside from someone else in the highlands of West Asia). So, while there might be a deficit of E-V13 in Central Anatolia, if the Kurds have E-V13 (who might be under sampled for political reasons), the case for attributing E-V13 in Anatolia almost entirely to the Greeks is greatly undermined.

Kurdish inhabited areas as of 1992 (from here).

The Anatolian branch of the Indo-European languages include the Hittite language, a group of Indo-Europeans who established a Bronze Age empire that controlled all of Anatolia and adjacent parts of modern Syria, Lebanon and Iraq, part of the Caucuses, and a small part of the far Northeast portion of modern Iran. 

The Hittite Empire at its greatest extent under Suppiluliuma I 

(c. 1350–1322 BC) and Mursili II (c. 1321–1295 BC)

The only other Anatolian languages attested prior to the waning days of the Hittite Empire immediately before or after Bronze Age collapse (when the Hittite Empire fell) were Luwian, spoken in mostly in Southeastern Anatolia, and Paliac spoken in North Central Anatolia.

Area where the 2nd millennium BC Luwian language was spoken.

After the Hittite Empire fell, many Anatolian languages are attested, all in Western Anatolia. These these may have had a relationship to the Hittite language similar to that of the Romance languages to Latin, and the Indo-Aryan languages to Sanskrit. But, they could also have been separate Anatolian languages from an earlier time period.

Anatolian languages attested in the mid-1st millennium BC.

On the other hand, if E-V13 really is found where there was Greek occupation, but is not found at meaningful percentages in places where Anatolian languages were spoken historically this could mean a number of things. At least, three scenarios could explain this:

* E-V13 was low enough in frequency that the same kind of Founder effects that kept it out of the Indo-Aryan founding population, even though it was in an Indo-Iranian founding population, could also have, by random chance, kept it out of the Anatolian founding population.

* The Anatolian languages were indeed, as many linguists assume (but I have argued against on mostly non-linguistic grounds), on the grounds that it is more distinct from other Indo-European languages linguistically, a branch that broke away from the Indo-European languages sooner than the Indo-European languages of Europe, West Asia and South Asia. Particularly if the Anatolian languages broke away prior to ca. 3500 BCE, the source population for E-V13 may not yet have arrived in the vicinity of the Pontic Caspian steppe to assimilate into the Indo-European community. 

There is some archaeological evidence that suggests that people who were culturally Egyptian and from the Levant, where the parent clade of haplogroup E-V13 (i.e. E-M78) was common made their way by boat to the Black Sea where they would have encountered populations ancestral to the Indo-Europeans (or the Proto-Indo-Europeans themselves) in about the right time frame. These people could have vanished by assimilating into the Proto-Indo-Europeans and could have been the source of Indo-European E-V13.

* The Indo-European people who were the source of the Anatolian languages may not have been very numerous and may have brought about language shift mostly through elite dominance (which may have been limited to members of those elites in the early stages) in a manner similar to the way that the Magyars brought language shift to Hungary. The Hattic culture that preceded the Hittite Empire was quite urban and advanced relative to that of many other areas conquered by the Indo-Europeans which were in a state of agricultural collapse and civilizational decay when the Indo-Europeans migrated into this power vacuum. In contrast, contemporaneous Akkadian and Hittite record establish that the expansion of the Hittite Empire was accomplished largely by forcefully seizing existing cities and communities of the Hattic culture, intact, and assimilating their citizens. It also also plausible, although not definitively clear, that there may have been long standing caste-like distinctions between conquered locals and the Hittites who had deeper foreign origins.

This scenario would also be consistent with a scenario in which Indo-European Hittite elites might have migrated en masse from other parts of their former empire to the Anatolian fringe neo-Hittite states of Western Anatolia where Anatolian languages were mostly spoken after the fall of the Hittite empire. This scenario is more plausible if the Anatolian languages were always the languages of a political and cultural elite that could migrate en masse more easily, than it is if language shift from the Hattic languages to the Anatolian languages was widespread even among illiterate peasants and city dwellers, regardless of their ancestry. There are historical examples of language shift following a conquest of a territory limited to elites and of language shift that penetrated more deeply into the general population.

The maps below help to disentangle other Y-DNA E-M35 haplogroups from E-V13 geographically. 

E-M81 has a Berber/Islamic empire conquest of Spain and colonial migration from Algeria to France profile.

The profile of E-M123 is harder to characterize succinctly.

And one more map for a context of the time of E-V13 expansion.

Progress In Calculating PDFs From First Principles

tl;dr 

Scientists have finally figured out how to do a key calculation about what comes out when you collide protons at close to the speed of light, using only select few of a couple dozen experimentally measured physical constants you can find on this blog, and a few of the equations of the single page worth of forty year old equations that make up the Standard Model of Particle Physics (which you can also find in a bare bones form without full explanations on this blog).

Previously the end product of these calculations, called a PDF (not to be confused with the computer file format) had to be figured out through the brute force collection of billions of data points, because the math involved in solving the equations was so hard. 

Figuring this out the hard way is a cottage industry in high energy physics because every other calculation that is done depends upon getting these right as one of the steps in the calculation. Progress on the brute force approach to figuring out PDFs is reported in probably at least one or two published physics journal articles a week, backed by large experimental collaborations, almost every week of the year.

Scientists realized that it was theoretically possible to do these calculations when they came up with the equations of the Standard Model about forty years ago. But, this month, for the first time, they managed to do those calculations. And, the Standard Model worked. The results of the calculations were a good fit to the billions of data points that had been collected in the meantime. 

You can look at the last chart in this post to see that while the prediction that the calculated wasn't a tight fit to the data (the data is the thick black lines) because the prediction had significant margins of error, that the data did fit perfectly in the predicted range (which is shaded) everywhere, and the margins of error weren't absurdly large.

A Less Abbreviated Description Of What They Did

A parton distribution function, also known as a "PDF" gives the probability to find a parton (a quark or gluon) in a hadron (for example, a proton) as a function of the fraction x of the hadron's momentum carried by the parton. 

A sample PDF from the Wikipedia link above. "The CTEQ6 parton distribution functions in the MS renormalization scheme and Q = 2 GeV for gluons (red), up (green), down (blue), and strange (violet) quarks. Plotted is the product of longitudinal momentum fraction x and the distribution functions f versus x." The Y axis corresponds to a probability from 0 to 1 (which is 100%). The X axis corresponds to the fraction of the proton's momentum carried by the parton from 0 to 1 (which is 100%).

Basically, in other words, you can used it to help determine the probability of particular kinds of particles spewing out from a collision in a particle collider of composite particles, each of which is bound by the strong force of quantum chromodynamics a.k.a. QCD.

This is not a trivial matter because a proton, for example, is really much more complicated than its two valence up quarks and one valence down quark, bound by gluons used to describe it on a simplified basis that is sufficient for many purposes, but not for describing the collisions of protons at relativistic speeds. A proton actually also includes a virtual sea of almost every other imaginable fundamental particles in the Standard Model, but at much lower probabilities of encountering them.

For example, even though neither a proton does not have a strange quark as one of its valence quarks, a proton-proton collision with the right amount of collision momentum can produce a kaon which does as a strange quark as one of its valence quarks.

Also, calculating PDFs require you to use what is called "non-perturbative QCD" regime which is much harder to do calculations with since non-linear effects overwhelm linear ones. In contrast, most kinds of QCD that we can make good calculations about , involve the usually higher energy perturbative QCD regime which is more tractable and linear mathematically.

In practice, PDFs are actually determined by looking at lots of collisions, seeing what falls out, and smoothing out that data from vast amounts of data collection, from billions of particle collisions at all sorts of different energies, into a continuous function.

But, the PDFs of particles in the Standard Model of Particle Physics aren't fundamental experimentally measured constants. In theory, you can calculate them using only the couple dozen experimentally measured fundamental constants of the Standard Model and its equations (which you could write readably by hand in chalk on a single normal classroom sized blackboard). It is just hard to do so. 

A new paper, the abstract of which is below, makes huge progress on this front with the first "from scratch" theoretical calculation of a PDF for protons and neutrons that is fully consistent with experimental results to within the relevant margins of error (which aren't small, as in all QCD calculations).

The details of the final results in the charts below are too technical to explain in this format. But, in the chart below, the thick black line is the experimental data, and the shaded areas are the theoretical predictions including the error due to the numerical approximations made in the calculations, and the inexactly measured physical constants used in the calculations. I reproduce it here to convey visually a sense of how accurate or inaccurate the calculations done in this paper were within the overall parameter space. The different colored shadings reflect different kinds of error sources that are included in each of the colored shadings.

For technical reasons, predictions when x (on the X-axis of the charts below) is very small, are harder to calculate.

There are at least a few reasons that this is significant.

First, a good theory is one that gives you information for free.[1] (Cf. "A good theory explains a lot but postulates little." - Richard Dawkins.) The case of deriving PDFs which previously took billions of data points to determine from a couple of dozen physical constants and a few equations in the Standard Model is an extreme example of a good theory, greatly improving the leverage we get from the Standard Model in terms of predicted results.

Second, the flow of information goes both ways. A theoretical determination of PDFs in terms of fundamental physical constants provides a new means of indirectly tuning our determinations of the values of some of those physical constants. This can be done by comparing the experimental data previously used to empirically determine PDFs to the theoretical predictions, and in the process finding which values of those physical constants provide the best fit to the PDF data, in a manner independent of the other hadron property measurements previously used for those purposes. At this point, the PDF determinations probably aren't precise enough to bound the values of many of the fundamental constants of the Standard Model very tightly, but even then, it provides independent corroboration of values for those constants measured by other means.

Third, as the abstract notes, the formula helps us to better understand the nature of the internal properties of protons, neutrons and other hadrons in a more compact and analytical way, as opposed to as the black box big data approaches by which PDFs are currently determined.

Fourth, the fact that this can be done is a good and robust non-parametric test of the conceptual soundness and completeness of the Standard Model. The fact that the same values of the same physical constants and the same equations can be applied in many diverse applications and produce consistent results is an excellent sign that the theory is sound and that any flaws in its are quantitatively small. There are lots of theories you can use to predict the hadron mass spectrum, for example. But, only really good ones can also accurately predict hadron PDFs and determine the mean lifetimes of hadrons from first principles. A PDF is highly sensitive to lots of beyond the Standard Model New Physics that one could imagine exists (particular new physics giving rise to additional particle content beyond the Standard Model) and the fact that it does not places empirical global constraints on the extent to which any true laws of Nature can deviate from the Standard Model.

[1] "The cognitive psychologist Pascal Boyer introduced me to the phrase “theory is information for free.” It’s a succinct way of saying that if you have a theoretical framework you can deduce and extrapolate a lot about the world without having to know everything. And, you can take new information and fit it quickly into your model and generate more propositions (you may not need to know everything, but you do need to know something)." - Razib Khan from here.

[Submitted on 5 May 2020]

Parton distribution functions from lattice QCD at physical quark masses via the pseudo-distribution approach

Manjunath Bhat, Krzysztof Cichy, Martha Constantinou, Aurora Scapellato

One of the great challenges of QCD is to determine the partonic structure of the nucleon from first principles. In this work, we provide such a determination of the unpolarized parton distribution function (PDF), utilizing the non-perturbative formulation of QCD on the lattice. We apply Radyushkin's pseudo-distribution approach to lattice results obtained using simulations with the light quark mass fixed to its physical value; this is the first ever attempt for this approach directly at the physical point. The extracted coordinate-space matrix elements are used to find the relevant physical Ioffe time distributions from a matching procedure. The full Bjorken-x dependence of PDFs is resolved using several reconstruction methods to tackle the ill-conditioned inverse problem encountered when using discrete lattice data. Another novelty of this calculation is the consideration of the combination with antiquarks qv+2q¯. The latter, together with the non-singlet valence quark PDF qv, provides information on the full distribution. Good agreement is found with PDFs from global fits already within statistical uncertainties and it is further improved by quantifying several systematic effects. The results presented here are the first ever ab initio determinations of PDFs fully consistent with global fits in the whole x-range. Thus, they pave the way to investigating a wider class of partonic distributions, such as e.g. singlet PDFs and generalized parton distributions. Therefore, essential and yet missing first-principle insights can be achieved, complementing the rich experimental programs dedicated to the structure of the nucleon.

Comments:	19 pages, 12 figures
Subjects:	High Energy Physics - Lattice (hep-lat); High Energy Physics - Experiment (hep-ex); High Energy Physics - Phenomenology (hep-ph); Nuclear Theory (nucl-th)
Cite as:	arXiv:2005.02102 [hep-lat]
	(or arXiv:2005.02102v1 [hep-lat] for this version)

The Latest Charm Quark Mass Measurement

A new analysis combining experimental data, and a combined lattice QCD and QED analysis, has produced the latest measurement of the charm quark mass adjusted to the energy scale of the charm quarks own mass (i.e. it's "pole" mass) from a measurement of its mass at a 3 GeV energy scale (at which it is more than 20% less massive than at its pole mass, because quark masses decline at higher energy scales).

It determined that charm quark mass to be 1,272.3 ± 7.8 MeV.

This compares to previous global estimates from multiple sources previously reported at this blog as follows:

2018 PDG charm quark mass 1,275 +25/-35 MeV

2019 PDG charm quark mass 1,270 ± 20 MeV

FLAG19 charm quark 1,282 ± 17 MeV

The new measurement is consistent at the one sigma level with the 2018 and 2019 estimates from the Particle Data Group and with the 2019 estimate from the FLAG group, but purports to be more than twice as precise as any of those estimates.

The FLAG2019 paper provides a chart that graphically shows some of its source data for its global average estimate of the charm quark mass including the margins of errors involved. Notably, the outliers (which tend to run heavy) have the largest margins of error and tend to be older.

Theoretical Bench Marks

Rivero's extended Koide's formula from 2011 predicts a charm quark mass of 1,359.56428 MeV which has long been known to be too high, and lacks of the precision and accuracy of Koide's rule for charged leptons. But, comes pretty close for a formula generalizing the Koide's charged lepton rule with only the electron and muon masses as inputs.

My first order adjustment of Rivero's formula based upon the CKM matrix and quark masses not considered in Rivero's formula predicts a charm quark mass of 1,293.21 MeV, which is also probably too high (although it is within one standard deviation of the 2018 PDG value and the FLAG19 value, and within 1.2 sigma of the 2019 PDG value), but is significantly closer to the experimentally measured value. Probably, even if my first order adjustment of Rivero's formula is correct, it also needs at least a second order adjustment as well that I haven't worked out.

Dark Matter Simulations Systemically Predict Too Few Stars Per Halo

One of the ways to try to understand the discrepancy between the apparent mass of systems of galaxy size or larger in astronomy based upon general relativity (usually in a weak field Newtonian gravity approximation), and the mass that can be observed from the stars in that galaxy, is to infer the presence of dark matter with a sufficiently small mean velocity that interacts only weakly via means other than gravity with ordinary matter or other dark matter.

Often, investigators attempt to determine what a particular kind of dark matter hypothesis implies in terms of predictions with simulations which they compare to what astronomers observe.

The paper below points out that the leading simulations get some key points wrong. Specifically, they systemically underestimate the amount of luminous matter in the form of stars (which where the lion's share of ordinary matter in the universe is found) relative to the inferred or simulated dark matter halos in which they are found, in a consistent and unbiased manner that can be quantified in several different ways.

But, because the simulations are wrong in some very predictable mays and magnitudes which the paper describes, it is also one of the more encouraging ones for the dark matter particle hypothesis in a long time. This is because the results suggest that only a small number of problems which a very specific character are wrong in the simulations, with shines a flashlight on how to look for a fix to the simulations and the dark matter and galaxy formation models that the simulations virtually test. It seems that the main problem with the simulations is probably related to stellar or active galactic nucleus feedback with the dark matter halos in the models. 

In contrast, many simulations have produced results that are wrong when compared to astronomy observations, while leaving astrophysicists to look under the metaphorical street light to look for possible solutions to explain why that might be the case.

Massive disc galaxies in cosmological hydrodynamical simulations are too dark matter-dominated

A. Marasco, L. Posti, K. Oman, B. Famaey, G. Cresci, F. Fraternali

We investigate the disc-halo connection in massive (Mstar/Msun>5e10) disc galaxies from the cosmological hydrodynamical simulations EAGLE and IllustrisTNG, and compare it with that inferred from the study of HI rotation curves in nearby massive spirals from the Spitzer Photometry and Accurate Rotation Curves (SPARC) dataset. We find that discrepancies between the the simulated and observed discs arise both on global and on local scales. Globally, the simulated discs inhabit halos that are a factor ~4 (in EAGLE) and ~2 (in IllustrisTNG) more massive than those derived from the rotation curve analysis of the observed dataset. We also use synthetic rotation curves of the simulated discs to demonstrate that the recovery of the halo masses from rotation curves are not systematically biased. We find that the simulations predict dark-matter dominated systems with stellar-to-total enclosed mass ratios that are a factor of 1.5-2 smaller than real galaxies at all radii. 

This is an alternative manifestation of the `failed feedback problem', since it indicates that simulated halos hosting massive discs have been too inefficient at converting their baryons into stars, possibly due to an overly efficient stellar and/or AGN feedback implementation.

Comments:	6 pages, 3 figures, submitted to A&A. Comments are welcome
Subjects:	Astrophysics of Galaxies (astro-ph.GA)
Cite as:	arXiv:2005.01724 [astro-ph.GA]
	(or arXiv:2005.01724v1 [astro-ph.GA] for this version)

From Christianity To Islam

When Muhammed established the Islamic faith, the time of Jesus on Earth (by religious convention), was more distant in the past (a century longer at Muhammed's death) from him, than Aztec King Montezuma II is from us today.