search this blog

Wednesday, April 19, 2017

Zarathushtra and his steppe posse

Part of the introduction to the new Lopez et al. preprint on the genetics of Zoroastrians says this:

The Zoroastrian religion developed from an ancient religion that was once shared by the ancestors of tribes that settled in Iran and northern India. It is thought to have been founded by the prophet priest Zarathushtra (Greek, Zoroaster). Most scholars now believe he lived around 1200 BCE, at a time when the ancient Iranians inhabited the areas of the Inner Asian Steppes prior to the great migrations south to modern Iran, Afghanistan, Northern Iraq and parts of Central Asia.

Disappointingly, in the rest of the preprint we hear nothing about these great migrations from the Eurasian Steppe and if perhaps they brought at least some of the ancestors of modern-day Zoroastrians to what is now Iran.

The preprint's title, The genetic legacy of Zoroastrianism in Iran and India: Insights into population structure, gene flow and selection, makes it clear that the authors are focusing on the genetic legacy of the Zoroastrians. OK, but why not also expand the focus to a detailed analysis of their genetic origin?

Possibly there's another paper on the way on the genetic origin of the Zoroastrians and other Indo-Iranians? Perhaps, but I'd say the issue here is that the authors have decided to make their main points with haplotypes, rather than unlinked SNPs, probably because, in principle, haplotypes are more powerful than unlinked SNPs.

Thus, they've chosen to limit themselves to using only a few relatively high quality, ancient genomes as reference samples. However, none of these ancient genomes are from the Eurasian Steppe.

As a result, the preprint includes a set of technically powerful haplotype analyses that, unfortunately, say nothing about the potential steppe origin of the Zoroastrians and are generally very difficult to interpret.

To fix this problem they can either sequence a couple of relevant ancient samples from the steppe at a high enough coverage to be useful as reference samples in haplotype tests, and/or expand their use of formal statistics to model Zoroastrians with the already available pseudo-haploid ancients from the steppe (see here).

Actually, since the Iranian Zoroastrians from this study are available online courtesy of Broushaki et al. 2016, I can try some formal statistics models now, using the latest qpAdm and the updated qpAdm methods from Lazaridis & Reich 2017. The results are sorted by statistical fit, best to worst:

Yamnaya_Kalmykia + Anatolia_ChL + Han + Iran_N

Sarmatian + Anatolia_ChL + Han + Iran_N

Afanasievo + Anatolia_ChL + Han + Iran_N

Poltavka + Anatolia_ChL + Han + Iran_N

Scythian + Anatolia_ChL + Han + Iran_N

Potapovka + Anatolia_ChL + Han + Iran_N

Yamnaya_Samara + Anatolia_ChL + Han + Iran_N

Andronovo + Anatolia_ChL + Han + Iran_N

Sintashta + Anatolia_ChL + Han + Iran_N

Srubnaya + Anatolia_ChL + Han + Iran_N

Not a huge difference there in terms of the fits. The best model is with Yamnaya_Kalmykia, probably because of its highest ratio of southern ancestry amongst these ancient steppe herder and warrior groups. Interestingly, the next best model is with the early Sarmatians from Pokrovka, Russia, who were, in all likelihood, Iranian-speakers.

I've also tested many other models using ancient Near Eastern reference samples other than Anatolia_ChL (Anatolia Chalcolithic), and can say with some confidence that the Zoroastrians have, one way or another, ~20% ancient steppe-related ancestry.

But how do other Iranian groups compare? It's an interesting and important question, because if modern-day Zoroastrians harbor elevated ancient steppe-related ancestry compared to other Iranians, this would strengthen the case for the steppe origin of Zarathushtra and his early followers. Let's test this using the same Sarmatian model as above (except with Yoruba added for the Bandaris to account for their minor African admixture):

Iranian_Bandari 27.7±0.6.5%

Iranian_Fars 22.2±4.6%

Iranian_Jew 2.9±0.5.9%

Iranian_Lor 15.3±5.5%

Iranian_Mazandarani 19.9±5%

Iranian_Persian 20±5.1%

And just to see what happens:

Iran_Chalcolithic 0%

So the Iranian Jews and Chalcolithic farmers from Iran basically show 0% Sarmatian-related ancestry. On the other hand, non-Jewish and non-Zoroastrian Iranians harbor, on average, 21.02±5.34% Sarmatian-related ancestry. That's actually not significantly different from the Zoroastrian result of 25.7±4.7%.

But importantly, modern-day Zoroastrians certainly don't appear to fall short in this regard compared to other ethnic and/or regional Iranian groups, despite being a relatively strong genetic isolate for many generations. What this suggests is that the Sarmatian-related ancestry mostly arrived south of the Caspian sometime between the Chalcolithic and the rise of Islam in Iran, quite possibly with the early followers of Zarathushtra during the Iron Age.


Lopez et al., The genetic legacy of Zoroastrianism in Iran and India: Insights into population structure, gene flow and selection, bioRxiv, Posted April 18, 2017, doi:

Update 30/04/2018: Lots of new ancient samples have arrived, so it's time to revisit this issue. See here.


Davidski said...

I always get confused with this: am I supposed to say pseudo-haploid or pseudo-diploid?

When_in_Rome said...

I had a question concerning a previous post (paper by Lazaridis,, is there a way I can contact you?

Davidski said...

eurogenesblog [at] gmail [dot] com

Nirjhar007 said...

Thanks Dave for your detailed input .

Unknown said...

Would the other Iranians containing as much, if not more, Sarmatian ancestry than the Iranian Zoroastrians indicate that Zoroastrians are the true descendants of the ancient Persians, since the Persians and Sarmatians were distinct cultures.

Nirjhar007 said...

Dear Daran,
Zoroastrianism possibly developed around Bactria and Pamir , it likely predates Sarmatians, though there are Suggestions that Zoroaster lived as late as 6th Century BC ! . Zoroastrianism also has iconographic parallels from ancient civilizations like Jiroft and Indus, see here :

Davidski said...

@daran b

In this context Sarmatian or Sarmatian-related refers to all other early Iranians, like Persians.

This analysis is not designed to differentiate early Iranian input, just estimate its overall impact, more or less.

MaxT said...

steppe-input is highest in Iranian_Bandaris ~27% compared to Persians at ~20%? I would have expected Persians to be highest for some reason.

"can say with some confidence that the Zoroastrians, one way or another, have ~20% ancient steppe-related ancestry."

"On the other hand, non-Jewish and non-Zoroastrian Iranians harbor, on average, 21.02±5.34% Sarmatian-related ancestry. That's actually not significantly different from the Zoroastrian result of 25.7±0.4.7%."

These two sentences are rather confusing, are Zoroastrians 20% or 25% steppe?

Davidski said...

But 25.7±4.7% = ~20%, right?

The Bandaris are from Hormozgan Province, which is fairly close to Pakistan, and Pakistanis generally have more steppe input than Iranians, so it's probably a post-Neolithic geographic thing.

Matt said...

Might be interesting:

Relative Fsts of Iranian_Zoroastrian against:

Ancient Iranians:

Other Modern Iranians:

(Column A - names, Column B - Fst to Iranian_Zoroastrian, Column C - Other, Column D - Difference between the two, Column E - Difference adjusted assuming Mbuti as 0, which should be adjusting for overall genetic drift, assuming Mbuti is an outgroup and any differences in how Eurasians relate to them are due to drift).

a) Shifts between modern Iranian_Zoroastrian and Iranian_Neolithic are biggest for relatedness to SW_Europeans and EEF. The most major shift between that time is an influx of Anatolia_Neolithic related ancestry.

b) Shifts between modern Iranian_Zoroastrian and Iranian_Chalcolithic are biggest for relatedness to Northern Europeans. This seems like a composite of continuing Anatolian influence and steppe related shift.

c) The shifts within modern Iranians suggest the Bandaris are more Iran_Neolithic like and less Anatolia_Neolithic like compared to other Iranians, less close net of drift to Mbuti to all other modern populations.

Shift with Mbuti normalised to zero, suggest minor Levantine, North Asians and South Central Asian shifts in all other Iranians relative to the Zoroastrians. Most are North Asian shifted, but Iranian_Lor also shows some Levant shift, and Iranian_Mazandarani shows a clear South Asian shift.

This agrees pretty well with these fits, with perhaps some slightly different fine detail on intra-moderns.

Aram said...


Maybe I misinterpret something. But why some Caucasian groups and even Levantines have lower fst distance to Iran_Neo than Zoroastrians?

Anonymous said...

"Sarmatian + Anatolia_ChL + Han + Iran_N"

From the percentages in terms of Steppe/Samartian the Zoroastrians are the same as other typical Iranians.

This suggests they this is NOT a migration of people from the steppe but either a local culture or a mass conversion of local people.

I would still like to see if Zoroastrians are higher in one of the other components (han etc) but I doubt it.

Matts fsts imply:
(1) Zoroastrians have less European (Barcin etc) than Bandari.
(2) Zoroastrians have less east asian (Miao etc) than Fars.
(3) Zoroastrians have less Bedouin and east asian (Miao etc) than Lors.
(4) Zoroastrians have less central and east asian (Miao etc) than Persians.

Basically there does not seem to be any external group that they have at a disproportionately higher level. Which further points to the Zoroastrians being genetically local and perhaps protected against incoming people by their social isolation.


Davidski said...

This suggests they this is NOT a migration of people from the steppe but either a local culture or a mass conversion of local people.

So did you miss this?

Iran_Chalcolithic 0%

I guess you forgot that most of Iran was Zoroastrian until the rise of Islam.

Anonymous said...

From Mats ancient DNA

Zoroastrians are more closely related to chalcolithic Iranians than neolithic Iranians (as expected).

For most external populations there is no significant difference between Zoroastrians and chalcolithic Iranians. Chalcolithics are a bit closer to ancient samples than the Zoroastrians as is reasonable.

I am going with Zoroastrians being in Iran since the chalcolithic, in genes if not in culture.

For the king said...

Not surprising results. Modern Iranians are also remarkably similar to those from the Copper age period. Those Sarmatians should be a good representative of Pre BMAC Iranics.

Few points from the study:

- Y-DNA wise, Zoroastrian priests are dominated by non Steppe related haplogroups such as L1b, the West-central Asian/European subclade of L (founder effects, similar to R1a1a in many south Asian castes)

- In Page 50, the graph shows that Muslim Iranians are less Homogenous (inbred?) than most Indians.

- Some Indian Zoroastrians are extremely shifted towards modern Iranians, while one sample clusters with Indians. Indian Zoroastrians seem to be shifted towards Muslim Iranians, rather than Zoroastrian Iranians. Keep in mind that most Iranian Zoroastrians come from the provinces of Yazd and Kerman in SE and central Iran, while Indian Zoroastrians(Parsis) probably came from multiple provinces.

@Arame probably because they're extremely inbred. Not sure.


" This suggests they this is NOT a migration of people from the steppe but either a local culture or a mass conversion of local people. "

It's totally possible that Sarmatian like people mixed with BMAC like people then migrated to modern Iran. This scenario is accepted by many archaeologists, historians, and anthropologists. Similar to what happened in Europe (Yamnaya derived groups mixing with local groups created Corded Ware, who probably spread IE languages/Admix into large parts of Europe). But BMAC seemed more complex and advanced than Pre-Corded Ware Europe, so the local BMACian influence on early Iranics might've been quite large.

@David, Iran was also majority Zoroastrian or Manichean till the 10-11th centuries according to some sources.

Davidski said...

I am going with Zoroastrians being in Iran since the chalcolithic, in genes if not in culture.

Obviously that's impossible considering that they have 20-25% admixture from the steppe that the Chalcolithic Iranians lack.

Nirjhar007 said...

Need nuclear aDNA from BMAC area , hard to suggest anything without it .

Zoroastrians show great affinity with Near East , both gene wise and also for example symbolism wise too :

Actually the name Asura/Ashura which also has PIE origins( See here : is very related to the Near Eastern Terms and can even be cognates! .

Anonymous said...

"So did you miss this?

Iran_Chalcolithic 0%"

Nope. I never said they were identical. I did say chalcolithic samples had changed between ancient and modern times. FST 0.022.

What I am saying is that from the fsts Matt used, compared to modern population groups, the chalcolithics and Zoroastrians are essentially identical. Same people. Zoroastrians are closer to modern Iranians in general but in terms of external distortions (on the fst list) they are the same as the chalcolithics.

Samartian Pokrovka is 0.028 versus 0.025 (insignificant) by the way. More of this kind of Samartian in the chalcolithic Iranians than modern, but insignificantly so. You are probably using a different Samartian proxy.

For the king said...

Elamites had a shit load of contact with Assyrians and other Mesopotamian cultures. Iranian Zoroastrians(Even those from the Teheran) are mainly from Yazd(Central Iran) and Kerman(SE Iran). If anything Lors seem like the most native group (Least steppe and most near eastern).

Keep in mind that Iranian Zoroastrians are not a special ethnicity, they're regular inhabitants of their respective regions that didn't convert to Islam. The ones in Iran are inbred/isolated due to their low numbers and religious persecution, but the ones in India are strict towards mixing with other ethnic groups. As far as I know, Iranian Zoroastrians accept converts while Indian Zoroastrians refuse them.

Also, Persians from Yazd are almost Identical to Zoroastrians from Yazd Y-DNA wise. The Zoroastrians also speak Persian and consider themselves Persian. The ones in India call themselves Parsi(AKA Persian)

Nirjhar007 said...

New Zoroastrian Paper with some new aDNA of ancient Zoroastrians .

Alberto said...

Well, I think that between Iranian Chalcolithic samples and modern Iranians there is a clear shift (some 20%) towards Europeoid populations (EHG, mostly). But the reason why we need better sampling is because one cannot so easily draw conclusions from such comparison.

The trend is probably mirrored whichever samples from other areas we used. If we compare Chalcolithic Iberians to modern Spaniards, there's going to be a shift towards ANE rich populations (usually something like Pathan gives a strong signal). If we compare Chalcolithic steppe populations (Samara_Eneolithic) to modern Ukrainians we're going to see a shift towards Anatolia_Neolithic and Near Easterners in general (maybe Iraqi_Jews work well).

Basically gene flow has been going in all directions in West Eurasia since the Mesolithic. But at specific points in time, there have been specific migrations. That's the interesting part, but it requires more complete sampling for most of the cases.

Nirjhar007 said...


Background: The Parsis, one of the smallest religious community in the world, reside in South Asia. Previous genetic studies on them, although based on low resolution markers, reported both Iranian and Indian ancestries. To understand the population structure and demographic history of this group in more detail, we analyzed Indian and Pakistani Parsi populations using high-resolution autosomal and uniparental (Y-chromosomal and mitochondrial DNA) markers. Additionally, we also assayed 108 mitochondrial DNA markers among 21 ancient Parsi DNA samples excavated from Sanjan, in present day Gujarat, the place of their original settlement in India. Results: Our extensive analyses indicated that among present-day populations, the Parsis are genetically closest to Middle Eastern (Iranian and the Caucasus) populations rather than their South Asian neighbors. They also share the highest number of haplotypes with present-day Iranians and we estimate that the admixture of the Parsis with Indian populations occurred ~1,200 years ago. Enriched homozygosity in the Parsi reflects their recent isolation and inbreeding. We also observed 48% South-Asian-specific mitochondrial lineages among the ancient samples, which might have resulted from the assimilation of local females during the initial settlement. Conclusions: We show that the Parsis are genetically closest to the Neolithic Iranians, followed by present-day Middle Eastern populations rather than those in South Asia and provide evidence of sex-specific admixture from South Asians to the Parsis. Our results are consistent with the historically-recorded migration of the Parsi populations to South Asia in the 7th century and in agreement with their assimilation into the Indian subcontinent population and cultural milieu like sugar in milk. Moreover, in a wider context, our results suggest a major demographic transition in West Asia due to Islamic-conquest.

Davidski said...

I don't find that paper very useful, because they didn't consider gene flow from the steppe, which is pretty dumb at this stage of the game. Not only because of what we've seen recently, but also because a lot of new data is coming any day, and guess what...

Ancient DNA Suggests Steppe Migrations Spread Indo-European Languages
David Reich

Nirjhar007 said...

Nonsense . You are again denying reality . Anyway, I know about that talk also what its on, you will not find it very amusing ..... unfortunately .

Davidski said...

Not sure what you're trying to communicate, but I can assure you that South Asians and Iranians have Bronze Age steppe ancestry.

There's no way around that fact. You'll have to accept it and include it in your calculations no matter what theory you prefer.

Olympus Mons said...

how would you better describe "Bronze Age steppe ancestry"?

Davidski said...


how would you better describe "Bronze Age steppe ancestry"?

Basically identical to Yamnaya and early Corded Ware and originating from what is now Ukraine and southern Russia.

Davidski said...


If we compare Chalcolithic Iberians to modern Spaniards, there's going to be a shift towards ANE rich populations (usually something like Pathan gives a strong signal).

That ship sailed a long time ago. We can be a lot more precise now. South Central Asia is out.

Nirjhar007 said...

Basically identical to Yamnaya and early Corded Ware and originating from what is now Ukraine and southern Russia.

That's not possible . Let come the data , all we need is the data .

Dilawer (Eurasian DNA) said...

Gently put, the sources used in tables S8-S10 to model don't seem relevant for Zoroastrians and other SC Asians such as Baloch and Brahui, and are not consistent with historical accounts or other evidence.

For example, in table S8 in the supplement, they model Parsis as LBK + Vishwabrahmin. Why even use 2 unknown sources, when source 1 is already known, ie Iranian Zoroastrians. So Iranian Zoroastrtian + NW Indian pop would be more informative and make more sense.

In S9, they model Pathan as IranianA + Bengali, and Balochi as Pakistani + Indian_Pakistani, and Brahui as Balochi + Pakistani. Seriously....why value do these have...

Based on historical accounts, linguistics, and my work, formally, Baloch and Brahui, at least some of their tribes, originated in the NW Iran region and surrounds and form a near clade with Kurds. Thus models consisting of Kurd (or NW Iranian/Caucasian) + Indian (or perhaps even Pathan) would make much more sense for them.

For Pathan/Pashtuns an additional steppe source, or for some Pashtuns an additional Siberian source may produces tighter fits, because for some additional steppe or Siberian is needed above and beyond what Kurds/Caucasians have.

Al Bundy said...

@Davidski Any other info on the Reich presentation? Saying steppe migrations spread IE languages seems to be a done deal.

Olympus Mons said...

Give me the components! wtf.
You start talking like a priest.

Al Bundy said...

Andrew Garrett from Berkeley is also giving a talk there.Should be interesting.

Davidski said...

@Al Bundy

The talk might be online, either live or uploaded later. If not, doesn't matter. Expect the paper soon after.

Al Bundy said...

Thanks great news

jv said...

Super interesting! Thanks Davidski!

Anonymous said...

Davidski, hear me out on this, and see if it changes your model.

You seem to be assuming that Iranian speakers were the first steppe arrivals in the Iranian Plateau, but I think there was a definite Indo-Aryan presence beforehand. I am not just talking about the Mitanni, but the fact that there is a substratum of Indo-Aryan in old Iranian languages. As an example, certain words have 's' sounds that should not exist in Iranian, due to a sound shift to 'h', but follow an Indo-aryan pattern.

Would it therefore make more sense to look at a two pulse model? You pointed out that some steppe samples look closer to Indo-Aryan, so might these work as a proxy?

Ric Hern said...

Some Ancient DNA from Khwarezm will be nice.

TruthPrevails said...

Do you guys know what steppe component is, here check this out.

So that is why you see steppe component in the neighboring regions of steppe's, when in reality those have been the contributors.

And there is no way to know how much isolation, drift and later admixture has affected the modern genetic profiles.

Untill the steppe fanboy scientists , can take group of samples
(not 1 or 2 as has been done to characterize other regions apart from steppe) they will not see the reality, or maybe they do not want to see.

Ideal way is to collect 10 to 15 random samples in a given region 1000 years apart, beginning from chalcolithic to the beginning of the era.
For example:

4000BCE - RegionA(n>=15) RegionB(n>=15) RegionC(n>=15) Steppe(n>=15)
3000BCE - RegionA(n>=15) RegionB(n>=15) RegionC(n>=15) Steppe(n>=15)
2000BCE - RegionA(n>=15) RegionB(n>=15) RegionC(n>=15) Steppe(n>=15)
1000BCE - RegionA(n>=15) RegionB(n>=15) RegionC(n>=15) Steppe(n>=15)

And then assess how admixture drift and isolation has affected the current modern population, rather than speculate.

But the steppe scientists are just happy to collect numerous steppe samples from one period and color everyone else with the admixed color.

That is just not only silly but shows a directed effort to prove the weak IE expansion hypothesis.

PS: (when I say scientists i do not mean davidski, he is just a blogger affected by confirmation bias :) )

Arch Hades said...

These people were the world's first Monotheists..when the ancient Hebrews were still worshiping statues.

At some point in time seems that Iranian Monists added duelism to their worldview which formed the basis of Zoroastrianism while the Aryans in India stayed fully Monistic.

Samuel Andrews said...

"Zarathushtra and his steppe posse"

Maybe a little romanticization of the Steppe horse riding raider.

Gaspar said...

Zorastrians, true West-asians, Balochi/Gedrosian group

Bactrians, Sogdrians, Pathans, Zaza, Tajik group of peoples

Fire worshippers some say of the Yazidi today.

IIRC........modern kurds are suppose to have originated form these lands before moving to there present areas

Matt said...


Fst from Iranian_Zoroastrian-Iran Neolithic: 0.046; Fst from Georgian-Iran Neolithic: 0.042

It's because of higher drift in Iranian_Zoroastrian.

Still, that's a different thing from what I'm trying to show in the columns though, which is the axis from Iranian_Zoroastrian->Iran Neolithic, and where Iranian_Zoroastrian is going to be the furthest from Iran_Neolithic on that axis by definition anyway.

So you can see which populations most similar to Iranian_Zoroastrian relative to their distance to Iran_Neolithic (mostly high EEF populations) or closest to Iranian_Zoroastrian relative to Iran_Chalcolithic (mostly Northern Europeans).

Matt said...

@ Annie, Davidski ran the Fsts for us so actually they're his, not mine! But also, I would say look at the shifts relative to the Mbuti outgroup. If you are just looking at the raw distance to EHG / Samara_Eneolithic, that is confounded by the effect of drift.

Like you note, the Zoroastrians are about exactly as separated in Fst from various modern European and Steppe populations as the Iran_Chalcolithic... but they are also more separated from all other populations (e.g. Fst to Mbuti - Iranian_Z - 0.210, Iran_Chalcolithic- 0.202, or Fst to Assyrian - Iranian_Z - 0.019, Iran_Chalcolithic 0.010). So relatively they are closer to the European and Steppe populations. I think this is probably because admixture brought them closer to European and the Steppe, then genetic drift pushed them further away.

I agree the difference between Iranian_Zoroastrian and Iran_Chalcolithic looks like a lesser and smaller shift comparing the difference between Iran_Chalcolithic to Iran_Neolithic though. Davidski's ancestry proportions suggests that Iranian_Zoroastrian group would be ancestrally about 70% Iran_Chalcolithic.

MaxT said...


"Zorastrians, true West-asians, Balochi/Gedrosian group
Bactrians, Sogdrians, Pathans, Zaza, Tajik group of peoples"

Not really, Ancient Zoroastrians would be closer to steppe-people than to Gedrosian or modern Iran ppl.

Closest living population to ancient Zoroastrians in that region would be Tajik, they have high steppe ancestry.

For the king said...


Zoroaster was probably a Post BMAC Central Asian Iranic, so his local admixture was pretty high. He was rejected in his homeland and was accepted by kingdoms in modern-day Iran. So Ancient Zoroastrians were most likely pretty west Asian like. Zoroastrian was always more popular in modern day Iran than central/south central Asia. There was a lot of conflict between early western and east Iranic regarding Zoroastrianism(The east usually practiced other religions, or mixed other elements with Zoroastrian like elements).

For the king said...

Blogger Samuel Andrews said...
"Zarathushtra and his steppe posse"

Maybe a little romanticization of the Steppe horse riding raider.

Hahahaha. Ironically there is a lot of evidence claiming that Zoroaster(And early Zoroastrians) disliked Nomads and promoted a settled life + agriculture. They often viewed nomads as evil barbarians with no morals and culture(BMAC influence most likely).

"On the side of the Wise Lord are the settled herdsmen or farmers, caring for their cattle and living in a definite social order. The follower of the Lie (Druj) is a thieving nomad, an enemy of orderly agriculture and animal husbandry. "

" Zoroaster assists the transition from nomadism to tillage with new moral standards, fortified by a book of account and a last judgment. "

MaxT said...

@For the king

Modern population living around BMAC culture today have high steppe ancestry, take Tajiks and Pashtuns for example, part of their ancestry is steppe-derived compared to modern Iran or Zoroastrians who are mostly of local admixture.

Ancient* Zoroastrians in Central Asia homeland probably had even higher steppe-ancestry than modern population living around BMAC culture.

Kurti said...

Zoroaster was neither a Central Asian Iranic nor was Zoroastrianism archaic religion directly coming from the Steppes. That is one of the issues when Genetists/linguists try to do the work of historians/archeologists and vica versa without having enough background knowledge. As I wrote already a few articles early in the comment section. The Persians are the result of a tribe that diverged from the ancestors of the Medes from Northwest Iran and merged with the local Elamites. The Zoroastrian religion itself emerged from the proto West Iranic sun cult of Mithraism that was born when Proto_Iranic relgiion came into contact with Mesopotamian planet_cult. Ahura Mazda (highest deity among Zoroastrians) was actually only one of many deities among the Medes with Mithra (deity of order/treaty and personification of the sun) being the highest one. Later when Zoroastrianism emerged out of this, Ahura Mazda took the place of the highest deity from Mithra among the Persians. Zoroaster himself is believed to have been actually a Mithraistic priest himself who was banished and created his own religion with most elements deriving from Mithraism. And he found believers among the Persians. But up until the Sassanid era there was still some conflicts between the Mithraistic Parthians (Descends of the Medes) and the Persians. This is why Ardashir is depicted together with Ahura Mazda and Mithras on the stone inscriptions to demonstrate him as the king of all West Iranic groups.

left Mithra with the sun rays around his head. right is Ahura Mazda

In Northwest Iran all the way into Anatolia there are still sculptures of all the typical proto Iranic deities such as Mithra, Anahita, Vahagn, Ahura Mazda.

The Monolithic heads on mount nemrut which were cut off from their bodies represent actually those Old Iranic deities and go back to the Medes period.

For more informations on how Zoroastrianism emerged out of the sun cult of the Medes see this presentation of Iranologist Kreyenbroek

Matt said...

@ Alberto If we compare Chalcolithic steppe populations (Samara_Eneolithic) to modern Ukrainians we're going to see a shift towards Anatolia_Neolithic and Near Easterners in general (maybe Iraqi_Jews work well).

Against EHG / Samara_Eneolithic / Yamnaya, the most Ukraine-like is Sardinian / EEF. Against WHG, the most Ukraine-like is Iraqi_Jew / Georgian / Abkhasian. Basically like the most negative WHG admixture f3 stats, like you rightly guess:

For South Asia:

(with expected stuff - Dravidian South vs Brahmin North is most pronounced for Sintashta and North European, Dravidian South vs Munda more Iran_Chal, Brahmin North vs Munda more Yamnaya).

For Levant:

Interestingly, biggest shift between Neolithic and Bronze Age to present day for the Levant are not Iran_N (or Arabian, or African) but present day North Indian populations.

We will see if the Indus Valley Culture populations take their place! (Seems like the logical extrapolation...).

MaxT said...


Sounds like local non-IE West Asians and Elamite played role in shaping early Zoroastrianism. High local non-I.E admixture in West Iranics can be explained by their vast multicultural empire, admixing with various non-IE West Asian population would have been a norm.

I don't think monotheism is a steppe/I.E concept considering Scythian religion in steppes was pretty barbaric and they were certainly not monotheistic. But, various concepts in Zoroastrianism does derive from the steppe Indo-Europeans, including fire worship. I wouldn't be surprised if Zoroastrian monotheism was derived from local West Asian population rather than steppe Indo-Europeans.

Anthro Survey said...

@Kurti and Max:

I strongly agree that Zoroastrianism represents a syncretism of ideas of steppe newcomers and the local(more numerous) West Asian population belonging to the Babylonian cultural sphere.

Some other things to keep in mind---

Zoroastrianism was also practiced in ancient Khorasan/Bactro-Sogdia but it was of a different character and it competed with Buddhism there.
It should also be pointed out that most fire temples(atashkada) are found in present-day west Iran: esp Fars and Azerbaijan. Not so much in Bactria or Sogdia.

An example readily coming to mind are the Barmakids, an important Bactrian family from Khorasan who attained a high standing in the Abbasid caliphate. They were of Buddhist heritage, not Zoroastrian.

Additionally, Zoroastrian communities did not survive in Khorasan as they had in Iran. Bactrians and Sogdians enthusiastically converted to Islam and greatly augmented it. Bukhara is to Sunni Islam what Nicea is to most forms of Christianity. Zoroastrianism was clearly not as rooted in the population there as it was in Western Iran.

Arza said...

@ Matt
(...) present day North Indian populations.
We will see if the Indus Valley Culture populations take their place!

If you're expecting that IVC samples will swap with modern NI populations than it means to me that you're also expecting that modern NI populations are similar to IVC.

Let's say that they migrated North. To the steppe.

They encountered some unknown ANE-rich population that we know it existed (Srub. out., Q) and mixed with them.

Finally they moved West and came into contact with local people (EHG,UkraineHG + CHG). They passed further the language, culture, but also the genes:

Yamnaya_AVG (Global 10)

Latvia_LN1:ZVEJ28 59.9
Srubnaya_outlier:I0354 17.05
Balochi 13
AfontovaGora3:I9050.damage 6.5
MA1:MA1 3.55

distance % = 0.7659 / distance = 0.007659

(Seems like the logical extrapolation...)


The correct answer is no, IVC won't take their place. We already know how ASI, ANI, IVC and Indo-Europeanized BMAC looked like.

Davidski said...


It's painfully obvious that you don't have a clue what you're talking about.

Your posts aren't adding anything to the debate, let alone being informative. They're simply taking up bandwidth.

Please make yourself familiar with the ancient steppe samples currently available, and what the consensus is about their ancestry. And then try to put together sensible, logical and coherent arguments.

Otherwise you need to change your nick to something more appropriate, like, for instance, TotalBullshit.

Kurti said...

Anthro said

"I strongly agree that Zoroastrianism represents a syncretism of ideas of steppe newcomers and the local(more numerous) West Asian population belonging to the Babylonian cultural sphere. "

Correct as I wrote above, just that I wouldn't call it the Babylonian sphere because they predate it and lean back to the Sumerian, Hurrian-Urartaen and Elamite period. Like the planet cult of the Sumerian. This has definitely influenced the proto Mithraistic believes of the West Iranics as mentioned by Prof. Kreyenbroek. West Iranic religions (Mithraism and it's derivate Zoroastrianism) are a synthesis of Old Mesopotamian and proto Indo_Iranian elements. Like the fact that Mithra was originally only meant to be the Deity of Order and contract and by the time of the Medes/Parthians) suddenly became the Deity and personification of the sun (which was already holy among the proto Indo_Iranians but merged into the Sumerian planet cult).

And example for that is the symbol of the Yezidis which according to Kreyenbroek (and I agree a 100% with him on that) are simply a sect of Mithraistic religion which have adopted some Islamic elements as proven by their shrines.

The symbol took my attention because it is identical to the Sumerian symbol of the sky/heaven.

Sumerian symbol

The Yezidi symbol look at some of the signs on it, especially down in the left corner. See it?

Among Kurds 4 Animals are symbolic. First is the Eagle/Falcon which represents the Blue Sky (typically Proto Iranic) 2. The Lion which represents royality/power and bravery (typically Mesopotamian, Anatolian, Iranian_Plateau). 3. Is the Snake which is especially depicted around Mithra (Mithras being the personification of the sun and the Snake representing the cosmos around it. The Yezidis believe that everything was born when a small round sphere exploded, yes I know sounds allot like the Big Bang theory. Another influence of Sumerian planet cult into the proto Iranic religion which gave birth to Mithraism.

Mithra with a snake around him

Snake on Yezidi shrine

Here more about the Yezidis and their believe of how the Universe emerged.

the last animal is the peacock which however is almost exclusive to Yezidis the reason for that could be that the peacock represents their highest angel which is according to some Muslim communities the Satan. And this could explain why the Peacock today only has a high status among the Yazidis.

Kurti said...

Something additional to Zoroaster

He had an obvious dislike for Mithras and put Ahura Mazda as the highest deity.

"The lack of Mithra's presence in the texts was once a cause of some consternation amongst Iranists. An often repeated speculation of the first half of the 20th century was that the lack of any mention (i.e., Zoroaster's silence) of Mithra in these texts implied that Zoroaster had rejected Mithra. This ex silentio speculation is no longer followed. Building on that speculation was another series of speculations, which postulated that the reason why Zoroaster did not mention Mithra was because the latter was the supreme god of a bloodthirsty group of daeva-worshipers that Zoroaster condemned. However, "no satisfactory evidence has yet been adduced to show that, before Zoroaster, the concept of a supreme god existed among the Iranians, or that among them Mithra – or any other divinity – ever enjoyed a separate cult of his or her own outside either their ancient or their Zoroastrian pantheons."[2]"

Zoroaster considered Mithra worshippers as infidels and according to Kreyenbroek in the youtube video I posted above, his followers oppressed the Mithra worshippers.

What could be the reason for the hate he had for Mithra and his followers? I mean at the end of the day Mithra was a deity among the local Iranic nomads. Maybe the fact that he himself was a former Mithra priest banished from to exile? This is seen by many historians as a very likely scenario.

Another mistake many amateurs make is to connect the emergence of West Iranics with the book of Avesta, while in reality Avesta is the point of emergence for the Zoroastrian religion and West Iranic religion/culture and languages actually predates it.

Coldmountains said...

Zoraosterianism represented an innovative religion with many BMAC elements. Vedic Hinduism was far more conservative. Zoraoster disliked nomadic Iranians. He was from Central Asia but monotheistic Zoraosterianism did not exist there prior to Sassanids.

Matt said...

@Azra, the general pattern, of this has been that where you compare Fst stats to a modern from a region (A) relative to an ancient from the same region (B), another Fst ancient (C) tends to show the greatest difference between A and B. C almost always tends to be an ancient from a region where moderns also show relatively great differentiation from A and B.

This holds despite the fact that Yamnaya, for instance, is not really very like modern Northern and Eastern Europeans (they have approx 50% ancestry from another population), and that there is huge population discontinuity.

So to me seems there's a good chance it will hold for comparing ancient North Indian populations where A and B are modern and ancient Levant. When we have that dna. (It may not, but there's enough of a hint for me to potentially waste my time putting it out there.)

Kurti said...


There is no reason to believe Zoroaster was from Central Asia. Most and oldest Zoroastrian temples are actually in West Iran. There are no records of Zoroastrians in Western Central Asia and South Asia prior to the expansion of the Achaemenids and following Empires.

Zoroastrianism is a synthesis of Old Mesopotamian and Proto Iranic elements.

Coldmountains said...

Avesta mentions no place in West Iran . Most of the mentioned places are in modern day Afghanistan and Tajikistan. Avestan language is early East Iranian and was spoken somewhere between modern day Turkmenistan, Tajikistan and Afghanistan. Zoraosterianism was in the beginning a sect in the eastern Iranian world which later spread in the West Iranic and became more popular there

MfA said...

Nations of the Vendidad, Avesta

Kurti said...

That is incorrect Coldmountain. And can only be explained with the fact that you interpret some of the mention region as in Central Asia. Avesta barely to never talks about explicit regions which can be pointed to a place. It speaks metaphoric. And I said Zoroastrianim emerged in Western Asia I didn't say in West Iran. I said it's parent culture is from West Iran just like Zoroaster but he himself moved out into exile (likely Northeast Iran) to find followers. And many scientists believe the reason why he wrote Avesta in an archaic language like Avesta (which according to them isn't actually the real language of any known tribe) was to reach as much people as possible and this could explain why there are maybe some regions in Central Asia mentioned in it too. the Zoroastrian cult and all it's elements are typical Mithraistic and old Mesopotamian.

Vara said...

Long time reader, first time posting.

Vendidad is not a reliable book for studying Zoroastrianism since it was written during the very late Sassanian period. The book mentions crossbows, stirrups..etc. Also, taking a look at the 16 perfect lands, geographically they don't make sense (all the way up to Rhaga but not Kerman?). What actually makes sense is that they were the lands that were in constant war/rebellion against the Rashidun and Ummayad Caliphates.

The Yashts and Yasnas are more reliable but they suffer from the same problem since they were written during the Achaemenid period and finished during the Sassanid period. The Yasht mentions three salt lakes and IMO identifying these lakes can tell us where the Zoroastrian homeland is. Now,, it's known from Middle Persian texts that Kavi Vishtaspa celebrated Zoroastrianism around Lake Frazdan. Middle Persian texts say that it is in Sakastan. However, there is Lake Hrazdan in Armenia, a place close to the Aras river where the Middle Persians thought Zoroaster came from.

Finally, the Gathas don't say much about where and when Zoroaster lived. However, it does mention a few things like: Horsemen, and metal (not Iron for sure). Weirdly, the warrior caste in the Avesta is not related to charioteers and there is no mention of chariots in the Old Avesta. We know that cavalry existed in Jiroft since 4.6kya, while chariots roughly 4.2kya. The Gathas mention another thing, the farmer-nomad conflict. People take it for guaranteed that this is a conflict between Andronovo and BMAC but not necessarily. Farmer-Nomad conflicts were all over West Asia, like Gutian-Sumerian, Lullubi-Babylonian...etc.

And of course some people above mentioned that CA can't be the homeland of Zoroastrianism, since pre-Achaemenids it was the homeland of the Dahae and the Scythians. In fact, the first main Zoroastrian city there was built by Cyrus. So in conclusion the Zoroastrian homeland is somewhere around Carmania and Drangiana.

Sorry for the long post.

Ric Hern said...

It looks like the same case as the "Historical Version" about the Peopling of Ireland. Political motivations formed the product of most Historical accounts. Actual Ancient History (The true version) gots mixed into the then current happenings and Political needs.

A simple Breogans Tower in Spain from where the Gaels could see Ireland and then migrated to it. This is of cause total Bullcrap since you need a Tower that is higher than Mount Everest to be able to see Ireland from Spain with the help of a powerful telescope.

On the other hand we have the DNA evidence that basically shows that the Irish Male Line migrated from Germany.

So the Spain connection was propagated by Political needs much later.

Ric Hern said...

So if Zoroastrianism had any connection to the Steppe is would have been through the Tazabagyab Culture in Khwarezm near the Aral Sea or the Chus Culture in Fergana.

Karl_K said...


"On the other hand we have the DNA evidence that basically shows that the Irish Male Line migrated from Germany."

Where did you get this data? How do you know where they came from? Couldn't the R1b males in both Germany and Ireland have come directly from a third location, such as Iberia? (but not originating there of course)

Ric Hern said...


Karl_K said...


I am familiar with the studies. However, we don't have Bell Beaker Y haplogroups from Iberia yet, so I think your comment is premature, and could be incorrect.

Ric Hern said...


They have a higher affinity to Central European Middle Neolithic than to Spain Middle Neolithic. The Irish Neolithic sample shows closer affinity to Spain Middle Neolithic.

Don't you think that the Bronze Age Males would have shown more affinity to Spain Middle Neolithic if their forefathers had to basically migrate through the whole of Spain to get to the Tagus Estuary and from there spread to Ireland ?

Karl_K said...


I could think many things, but I will hold off on making statements in this case. They ckearly had an undetermined migration route, as R1b men came from the steppe but did not have much of an effect in Corded Ware territory.

No data has been released to make a clear determination at this time.

It could be that the first Irish Indo-Europeans came from Germany, but that is very far from clear at this time.

It could be that the Neolithic admixture occurred far from the direct origin of the R1b migrants to both Germany and to Ireland.

Ric Hern said...


I am specifically talking about the Irish Males (Gaels) compared to Irish Myths of origin.

The Myth about migrating straight from Spain across the Sea to Ireland because they could "see" Ireland from Breogans Tower in Spain is Bullshit.

If they migrated from Spain there would have been more affinity to the Spanish Middle Neolithic.....

Ric Hern said...

This Myth was most probably written by someone that had a Map of Europe and he most probably drew a straight line from Spain to Ireland and made the Usumption that a person could see Ireland from Spain without taking into account the Curvature of the Earth and the Actual distance between them.

What is the oldest Map that you know about found in Ireland ?

Karl_K said...


Obviously the myth is bullshit. But, that does not mean that the first of the "genomically modern Irish" didn't immigrate (almost) directly from Iberia. We just do not know yet. Perhaps by this summer we will.

Anthro Survey said...


I meant to say Messopotamian cultural sphere instead of "Babylonian" but due to the strong connotation of the word to the geographical region, my brain took a shortcut. lol

Thank you for the lengthy presentation. That Sumerian cuneiform sky symbol appearing in the Melek Tavuš emblem(as wel as the other graphics) are really telling and demonstrate a clear continuity from pre-Iranic times.

I had always suspected this strong substratum/continuity, albeit on a much more intuitive level. It's nice to see my intuitions confirmed by an expert(from that region?).

Feel free to inbox my youtube account, if you like. I may have questions on specifics of the matter in the future.

Anthro Survey said...

@Vara and Kurti:

Could the mentions of the holy lands in the Vendidad simply REFLECT an increased importance of the Greater Khorasan area in the late Sassanid era? This importance manifested itseld in a few ways it seems and obv. carried over into early Islamic period.

The top 4 lands mentioned are all located there: Khorasm, Sogdia, Balkh and Marv. Well, the number 1 land is actually assumed to mean Khorasm and I belive this is a valid assumption considering it fits the overall trend. It should be stressed that all 4 of these areas "struck it rich" after the sassanid period----becoming great centers of power, trade, theology and secular studies.

Anthro Survey said...


"But, that does not mean that the first of the 'genomically modern Irish' didn't immigrate (almost) directly from Iberia. We just do not know yet"

We kinda do, actually!
Bronze age (2000-1500BC) genomes from Ireland taken from samples at Rathlin island are heavily influenced by the Steppe---on a level comparable to modern Irish and Brits. They appear to have a very similar balance in terms of overall balance in ancestral components: WHG, Anatolia and Steppe.

Moreover, the Rathlin samples are considerably more steppic than MODERN Iberians, let alone contemporaneous Iberians who were mostly Sardinian-like at that time.

Then there is the ample Y-dna evidence: completely diff clades of R1b in Iberia and Ireland.

Now, if we're talking about NEOLITHIC Ireland, that's a different story and one can potentially make a case for an Iberian connection. But even here it's not highly likely due to the primitive nature of Neolithic sailing craft. The first Anatolian-like arrivals to Ireland probably trecked the Doggerland land bridge, not boats.

If you're wondering, it is the Neolithic Irish genome from Balynahatty that indicates Neolithic Irelanders were comparable to other Neolithic Europeans and were quite Sardinian-like.

On a side note, I don't know why Davidski never used either Rathlin or Balynahatty to model Scots in a post a couple months back. Would probably be a stellar fit!

Anthro Survey said...

*the ATP9 Bronze Age sample from Iberia does show some steppic influence, to be fair, but it's still less steppic than modern Iberians/Basques.

Karl_K said...


What are you talking about? Completely different clades? The Basque and the Irish are both ~85% R1b-M269. Perhaps you mean sub-clades, but the dates of those splits are very recent (in this context) and most of the populations involved have had massive founder effects and drift since then.

Whenever this big Bell Beaker paper arrives, then I will probably be convinced one way or another. Hopefully it will be soon.