October 30, 2008

"Phoenician" Y-chromosomes

It has been several years since the inception of the Genographic project, and to say that the quantity and quality of the work produced by it is underwhelming would be charitable.

The newest bit of Genographic wisdom is that haplogroup J2 in the Mediterranean is associated not with the Neolithic, Greek, or other population movements, but with the sea-faring Phoenicians. They achieve this feat by (allegedly) comparing areas of Phoenician with those of no (or low) such influence.

I have intentionally limited myself to five major weak points of the study: to cover more would be too time-consuming and unnecessary.



1. The Hellenistic age did not happen

A central assumption of this work is that the conquest and occupation of the Middle East by Alexander the Great does not count as Greek influence, despite centuries of Greek domination that followed, both during Hellenistic, and later in Roman times.

The authors write that their method could be further used to:
include systematic investigations of military expansions, such as the Greek signal, from the time of Alexander the Great in central and south Asia
Apparently they didn't think of applying it to West Asia itself, which was also conquered by Alexander the Great, and in which the Greek-speaking element persisted far longer than in "south Asia".

Thus, the population of Phoenicia and its "periphery" is implicitly assumed to be free of Greek influence. That is a bizarre contention, given that Greek was spoken in "Phoenicia" long after the Phoenician language became extinct.

2. Crete was influenced by the Phoenicians

This totally unsupported claim is necessary for the authors' thesis, since Crete has the world maximum of haplogroup J2. I have no doubt that Phoenicians traded with Cretans, just as Cretans traded with Phoenicians. But, that is no excuse to think of Crete as an area of Phoenician influence.

Indeed, settlement of the Levant by Aegean peoples is archaeologically supported, while Phoenician settlement of Crete is not.

But, speaking of Phoenician settlement, the only area of Greece where such settlement is believed to have taken place is in mainland Greece, in Thebes, where Cadmus and his Phoenicians founded Cadmeis. I doubt that this had any substantial effect, but if the authors wanted to be intellectually honest, they would list this as an area of Phoenician influence, rather than Crete.

3. West Asia Minor (or the Pontus) was not colonized by Greeks

The most laughable claim of the authors (see map) is the absence of blue (Greek) dots on West Asia Minor, and the Pontus (Northeast Turkey). Apparently the Greek colonies of the far West (such as Marseilles) count as areas of Greek influence, while the countless Greek cities on the Asian side of the Aegean, or in northeast Turkey do not.

The motivation of this is obvious, since Asia Minor is a J2-heavy area and asserting the Greek influence there would upset the paper's thesis. But, it is absurd to place blue dots in Paphlagonia and Caria and not in Ionia or the Pontus.

4. Modern Lebanese are descendants of Phoenicians

This central assumption of the paper has no actual support, except for a vague geographical congruence. Modern Lebanese are a hybrid people, divided into Christians and Muslims. Both are Arabs, with Muslims being more influenced by the original Arabians, and Christians more influenced by the pre-Arab (Greco-Syrian) and post-Arab (West European) migrations. Perhaps, there is a trace of Phoenician genes in them, but this is really not a self-evident claim.

5. R1b in Greece and Turkey is due to the Celts

R1b in Greece and Turkey belongs primarily into the "eastern" variety, and not the "western" variety. It is in Italy and north of Greece where the two varieties begin to blend with each other. No care to distinguish between these varieties is taken.

Certainly, some R1b in this region may be due to Western Europeans (e.g. from the period of the Frankokratia), but to assign its totality to this factor is nonsensical. Apparently, the geniuses of the Genographic project have decreed that the brief foray of the Celts into Greece introduced massive amounts of R1b, but a thousand years of Greco-Roman domination of the Levant did nothing of the kind.

6 (bonus). Haplogroup J2 is more frequent in East than in West Sicily

Sicily is an island which had well-documented and not insignificant settlements by both Greeks and Phoenicians. Moreover, these settlements were geographically divided: Greeks in the East, Phoenicians in the West. It is in the East that J2 has its highest frequency, and not in the Phoenician West.

Conclusion

Is there anything of value in this paper? Well, it's a good idea to try to correlate Y-chromosome distribution with historical rather than pre-historical events. Too bad the authors botched the job, but their paper can at least serve as a reference point for how not to go about doing it.

UPDATE: Take a look at the "haplotype groups" suggested by the authors as signals of Phoenician and Greek colonization.



Not only are haplotype groups not clades (they do not designate common ancestry), but 7-marker haplotypes don't even designate anything that can be remotely tied to the time period in question, given the huge confidence intervals associated with even larger numbers of markers. Feel free to plug these haplotypes to yhrd or ysearch to find plenty of long-lost "Phoenicians" all over the planet.

UPDATE II: The "evolutionary" mutation rate rears its ugly head

From the paper:
Because there is a significant chance that a haplotype existing 3000 years ago has accumulated a one-step difference in an STR (we expect 0.6 mutations per seven-STR haplotype when a rate of 6.9x10-4 per locus per 25 yr is used), these one-step neighbors have been included in each set, producing what we have labeled STR+s. STR-s can contain both haplotypes deriving from mutations, which should have been included, and independent haplotypes unconnected with the migrations that we are trying to detect.
UPDDATE III: What of the Arabs?

The modern Lebanese are Arabs, as are most modern North Africans where Phoenician colonies were founded. The Arabs also affected several Mediterranean islands, as well as Iberia. One would think that the most salient feature of modern Mediterranean populations would be mentioned in a paper which attempted to trace patterns of Y-chromosome variation in the Mediterranean.

Certainly, the Neolithic, Greek, and Phoenician migrations, as well as the Jewish Diaspora moved people around. But the Phoenicians have been extinct for 2,000 years. The Jews had (and have) communities around the Mediterranean, but did not amount to a significant population element anywhere. It is the Arabs who are the elephant in the room, and yet they are ignored. Are similarities between the Levant, North Africa and Spain due to Phoenicians or due to this later Arab movement? By failing to trace the distribution of their "Phoenician colonization signals" among Arabians, the authors have overstated their case.


American Journal of Human Genetics doi: :10.1016/j.ajhg.2008.10.012

Identifying Genetic Traces of Historical Expansions: Phoenician Footprints in the Mediterranean

Pierre A. Zalloua et al.

Abstract

The Phoenicians were the dominant traders in the Mediterranean Sea two thousand to three thousand years ago and expanded from their homeland in the Levant to establish colonies and trading posts throughout the Mediterranean, but then they disappeared from history. We wished to identify their male genetic traces in modern populations. Therefore, we chose Phoenician-influenced sites on the basis of well-documented historical records and collected new Y-chromosomal data from 1330 men from six such sites, as well as comparative data from the literature. We then developed an analytical strategy to distinguish between lineages specifically associated with the Phoenicians and those spread by geographically similar but historically distinct events, such as the Neolithic, Greek, and Jewish expansions. This involved comparing historically documented Phoenician sites with neighboring non-Phoenician sites for the identification of weak but systematic signatures shared by the Phoenician sites that could not readily be explained by chance or by other expansions. From these comparisons, we found that haplogroup J2, in general, and six Y-STR haplotypes, in particular, exhibited a Phoenician signature that contributed > 6% to the modern Phoenician-influenced populations examined. Our methodology can be applied to any historically documented expansion in which contact and noncontact sites can be identified.

Link

59 comments:

cacio said...

Seems like an interesting paper, I hope I'll get hold of it. The Crete assumption seem really fundamental. If I remember from other papers, they claimed that a lot of the J2 in Italy and the Mediterranean came from Anatolia in the Neolithic, bypassing most of Greece, but stopping in Crete. It seems natural that some of this neolithic Anatolian/Middle Eastern J2 persists also in modern Lebanon. This doesn't make it phoenician, especially if it is present also in Anatolia or other parts of the fertile crescent.

The pace of papers originating from the genographic project is really slow - there were some notable annoucements (Chad, Tajikistan, negritos), but I haven't seen any of them.

cacio

Dienekes Pontikos said...

J2 is found all over the Middle East as well, in regions far from the coast, and into Iran and Central Asia.

The authors should have examined the distribution of their putative J2 Y-STR haplotypes more fully, although 7-marker haplotypes as evidence of historical descent are pretty useless -- as the Cohen Modal Haplotype debacle has illustrated.

It's as if time has stopped in the early 2000s and all the downstream markers in J2 don't exist. I can understand this from budget-constrained researchers, but what's the Genographic Project's excuse?

Crimson Guard said...

I think the usage of Y-Chromosomes and mtDNA is largely like Bloodgroups were of the 1980's and 90's, highly misleading and with scientists blindly playing historians and archeologists. Most of what they come up with is guess work and conjecture at best, and later on dates and stuff are changed but hardly ever fully corrected as you find decades old errors popping up again every often which is beyond sloppy. Autosomal DNA I beleive is more accurate in the long run and the end.

Anyway, the land which comprises of "Phoenicia" was dominated by the Western Mediterranean peoples(mainly Italians) and Greeks during the heyday of the Sea People and their migrations. Even if one is to dimiss the "Sea People", you can dismiss the Roman and Alexanderian Greek military colonies and complete domination of that area for nearly a continuous 2,000 years, plus German, Frenh and Italian Crusaders.

Although the "Sea People" can be reinforced with the Philistines being a prime example of foreign occupation of that area at an early time. And they were no doubt mostly Greeks, I dont think there is much argument with that assessment. Besides, Phoenicians only come about after them and the Achaean contact/influence on the island of Cyprus.

In Sicily, Motya, the best-known and archetypal Phoenician settlement probably evolved due to an early collaboration with local indigenous populations on the mainland for its success. Moreover its important to indicate that the majority of indigenous people in western Sicily adopted Greek, not Phoenician, cultural traditions during the 6th century BC.

You can find Sicilian made weapons and pottery in Cretan burials. As they imported a lotve Sicilian made swords which the archeological record shows. And Sicilian traders and sailors were noted in Homer's works. So of course they undertaking the their sea voyages and didnt need Greeks or Phoenicians to do so as often erroneously made out to be.



Its important to note that historical Greek and Phoenician colonies in the western Mediterranean were primarily administrative not military such as the case of Alexandrian and Roman. So I dont see waves of tens of thousands of people repopulating areas which were already populated and thriving beforehand.

Hellenism was popular and a sign of being civilized and cultured at that point in time , it would rather be like Roman/Latin and currently English is used today. So in any case I think it was more cultural and linguistic rather than genetic for the most part for the (so-called) Phoenicians and even Greeks.

Crimson Guard said...

* Edit, was to be: "you can't dismiss the Roman and Alexanderian Greek military colonies and complete domination of that area for nearly a continuous 2,000 years, plus German, French and Italian Crusaders."

Damn f'n typos,LOL

Judith Weingarten said...

Your Greek point is good, but what about 'Arab' genes: how did they split the 'Arab' from 'Phoenician' genes in Lebanon? [Testing Christians only? Tricky] And note that the places they studied were all affected by the Arab migrations too:

They looked at the genetic signatures carried on the Y chromosomes of men from former Phoenician colonies across the Mediterranean. The sites included coastal Lebanon, Cyprus, Crete, Malta, eastern Sicily, southern Sardinia, Ibiza, southern Spain, coastal Tunisia and the city of Tingris in Morocco.

Dienekes said...

but what about 'Arab' genes: how did they split the 'Arab' from 'Phoenician' genes in Lebanon?

The short answer is: they didn't. Strangely enough, according to the authors, the Phoenician expansion and even the Jewish Diaspora are worthy of consideration, while the Arab expansion, which is both more recent, and affected all of the Middle East, North Africa, and parts of Southern Europe, isn't.

terryt said...

Dienekes. In the post you wrote, "Modern Lebanese are a hybrid people, divided into Christians and Muslims. Both are Arabs".

Firstly, it seems to me that all people are hybrids, especially if you're prepared to go go back more than a couple of thousand years. Can you give us an example of a purebred group?

Secondly, even though the people now living in ancient Phoenicia speak Arabic this in no way means that Arab genes have replaced the earlier genes. Languages can, and often do, spread independently of genes to a surprising extent.

As for the argument as to whether the Y-chromosomes in question were spread by Phoenicians, Greeks or Jews. As Crimson Guard said there was a huge amount of movement around the eastern end (at least) of the Mediterranean from (and probably before) the time of the Sea People. Therefore all three groups would have much the same distribution of said Y-chromosomes.

Therefore I agree with you that it would be near impossible to isolate the genetic effect of any one ancient migration in the Mediterranean.

Dienekes said...

Firstly, it seems to me that all people are hybrids, especially if you're prepared to go go back more than a couple of thousand years. Can you give us an example of a purebred group?

No doubt, which is why it's important to study the hybridity of the "homeland" of any supposed migration.

Secondly, even though the people now living in ancient Phoenicia speak Arabic this in no way means that Arab genes have replaced the earlier genes. Languages can, and often do, spread independently of genes to a surprising extent.

To what extent they have preserved Phoenician genes is a matter of investigation, and not something that should be a priori assumed. There certainly is no reason to assume that the genetic heritage of the Phoenicians would be more important than that of the people that preceded them in the region, or that of the people that succeeded them.

Alain said...

"Modern Lebanese are a hybrid people, divided into Christians and Muslims. Both are Arabs".

The last phrase is incorrect IMHO, cause in a recent TV appearance Dr. Pierre Zalloua, a Lebanese member of the Genographic project showed that statistically the Lebanese people have about 20% J2 haplogroup and about 18%-20% J1 haplogroup (which he related to the Arab genes). So doing such an assumption is not accurate.

"Are similarities between the Levant, North Africa and Spain due to Phoenicians or due to this later Arab movement?"

This issue cannot be compressed to either this or that, either the Phoenicians or the Arabs. By the way if you have a source that shows the percentage of J1 in northern Mediterranean countries that would be surely informative.

Dienekes said...

"Modern Lebanese are a hybrid people, divided into Christians and Muslims. Both are Arabs".

The last phrase is incorrect IMHO, cause in a recent TV appearance Dr. Pierre Zalloua, a Lebanese member of the Genographic project showed that statistically the Lebanese people have about 20% J2 haplogroup and about 18%-20% J1 haplogroup (which he related to the Arab genes). So doing such an assumption is not accurate.


Christian and Muslim Lebanese differ from each other.

Dienekes said...

Also, J1 may be related to an _Arabian_ (not Arab) component in the Lebanese. That they are different from Arabians (who have very high frequency of J1) is clear, that they (especialy the Muslims) are particularly different (especially due to a "Phoenician heritage") from other Levantine Arabs is not.

terryt said...

It would be most interesting if anyone (perhaps Ebizur?) could supply us with the varying proportions of haplogroups in Lebanese, Palestinians, Jews, Greeks, Cretans, Sicilians and Southern Italians. It seems certain that genetic interchange between these regions has been going on for a long time. As also with Egyptians, Libyans, Tunisians and Southern and Western Turks but these would be greatly complicated by the huge hinterland involved.

Perhaps we might then get some idea of how the various haplogroups reached their present distribution. Untangling it would still be a major operatioin of course.

"no reason to assume that the genetic heritage of the Phoenicians would be more important than that of the people that preceded them in the region". Surely the 'Phoenicians' were simply the product at the time "of the people that preceded them in the region".

Dienekes said...

Surely the 'Phoenicians' were simply the product at the time "of the people that preceded them in the region".

Would you say then that Americans are simply the product of Native Americans?

terryt said...

No. But Americans today are a simply the product of the people that have preceded them in the region, including the Native Americans.

The 'Phoenicians' are simply the people who happened to be living in the region when the term was used. Who, obviously, were "the product at the time of the people that preceded them in the region". Nothing complicated in that.

Howard said...

Dienekes,

I think your rejection of this study is correct. People have a natural desire to see patterns in randomness. That's why "technical analysts" of the stock market are sought out and quoted.

What the authors of this paper should be forced to do is state a hypothesis based on the data they have collected and then go out and collect 1330 new DNA samples to repeat the study. When they do so, they will probably find out that the symphony they heard was merely white noise.

There is a difference however in the settlement and administrative patterns of the Phoenicians and the Greeks. The Greeks maintained a greater cultural separation from the eastern peoples they lived amongst than did the Phoenicians, and to some extent they have maintained it even to modern times. In Roman times, the Hellenistic East continued to speak Greek, and the urban Greek cities like Alexandria and Antioch maintained a linguistic and cultural separateness from the countryside.

Howard

Judith Weingarten said...

Howard
Despite the barriers erected by the Greeks, intermarriage with Egyptians became fairly common -- especially up-country. It was almost always between a Greek or Macedonian man and an Egyptian woman, not vv, since the children inherited the status of the father.

Naphtali Lewis (Greeks in Ptolemaic Egypt) writes of "an almost mathematical progression: the farther a settlement was from an urban centre, most particularly ... Alexandria, the more quickly and more readily...." intermarriage took place. Children of such mixed marriages bore both names, one Greek, the other Egyptian. Papyri record many with such dual names.
Under Roman rule, of course, Greeks lost their privileged status: they were all Egyptians to their conquerors. By this time, anyway, intermarriage had become more common, now in both directions. Some of the regulations set by Augustus and continued for 200 years specifically refer to 'Those born of an urban Greek mother and an Egyptian...' and 'If an urban Greek marries an Egyptian woman ...."
And then, of course, there was prostitution.

I am not as sure about Syria since the records are much poorer but I imagine that, there too, intermarriage began early, probably in rural areas and among the more humble, and increased both in frequency and among 'middle classes' over the centuries.

Sam said...

First thing first Phoenicia is only 10% Lebanon (coastal Lebanon):

The marker that distinguishes Lebanese people from the rest from the rest of the Arabians is E1b1b not J2!

E1b1b Frequencies

Coastal Lebanon = 40-50%
Lebanon = 32%-25%
Syria = 11%
Iraq = 12%
Yemen = 12%

J2 Frequencies
Iraq 30%
Syria 30%
Crete-Cyprus 30%
Inland Lebanon 25-30%
Kurds 28%
Turkey 27%
Georgia 26%
Peloponnesse 25%
Israel 25%
Coastal Lebanon 25%
Palestinians 24%
Azerbaijan 24%
Greeks 23%



Phoenician influence in Sicily:
The Greeks had 14 ancient pre-Punic wars colonies vs 2 Punic colonies

Pre-Punic wars Greek colonies:

Greek East Sicily
Messana
Katane
Leontenoi
Syrakousai
Gela
Naxos
Himera

Greek West Sicily
Akragas
Silenous
Segesta
Thermai

Phoenician Colonies -NW Sicily
Panormos
Motya

Both the Phoencian colonies became the heartland of the Aghlabids in Sicily, they build Palermo and populated it by far more than the Phoenicians did (adding near Eastern genes -that were volunteered as Phoenician markers in the study!-)

Add to that The Byzantine empire brought Sicily under Greek contol again.

NorthWest sicily was settled by various NearEasterners the Jews, Aghlabids (Saudi-Iraq), Kalbids (Yemen-Syria), followed by the Aragonese (Ancient Greek Iberia & at that time they had Mudejer troops)! All who had Near Eastern genes!

How are they picking out the so called Phoenician DNA!!

Who are they kidding?

I am glad I am not the only who smelled the bias.

terryt said...

"The marker that distinguishes Lebanese people from the rest from the rest of the Arabians is E1b1b not J2". Thanks for that information Comf.

Do you have information on E1b1b for the regions you list for J2? The frequencies you provide certainly hint at extensive intermixing in the Eastern Mediterranean, presumably starting long before the Phoenicians developed and continuing since their time.

Sam said...

"Do you have information on E1b1b for the regions you list for J2?"

J2/E1b1b Frequencies
Azerbaijan 24/10
Syria 30/11
Iraq 30/12
Kurds 28/12
Georgia 26/12
Turkey 27/14
Inland Lebanon 25-30/20
Crete-Cyprus 30/30
Peloponnesse 25/47
Israel 25/28
Coastal Lebanon 25/45
Palestinians 24/25
Greeks 23/23






"The frequencies you provide certainly hint at extensive intermixing in the Eastern Mediterranean, presumably starting long before the Phoenicians developed and continuing since their time."

Thats what I also think, the East Med (COASTAL) population shared a common E1b1b/J2/T that predates the Semitic & Hellenic cultures in the region.

J2 = Neolithic Farmers
E1b1b = Coastal People

Unknown said...

Influx of Post-Phoenician J into Lebanon.

Mesopotamians
7th BC Assyrians (35%J2/15%J1)

Levantine-Mesopotamians
6th BC Arameans (30%J2/20%J1)

Hellenics
4th BC Greeks (25%J2/5%J1)

Southern Arabians
1st BC Amela (80% J1/10%J2)

Southern Europeans
1st BC Romans (15% J2)

Southern Arabians
3rd AD Ghassanids (80%J1/10%J2)

Levantines + Mesopotamians
5th AD Syriacs(25%J2,25%J1)

Peninsulars
7th AD Muslims (50%J1,10%J2)

Other Notable groups:
Malhami Muslims 30% J2, 15% J1 (Tripoli Sunnis original from Southern Anatolia)

Ottoman armies? (25%~J2 10% J1)


& Thats the story of the Phoenician J2/J1 forgive me if I forgot someone here & there

terryt said...

Thanks again Comf.

terryt said...

Comf. There's some interesting stuff in your numbers. E1b1b certainly reaches its highest frequency on the coastal regions around the Eastern Mediterranean, more than 20%. Perhaps spread by the Sea People, whoever they were?

And Israelis share half their Y-chromosomes, in the same ratio, with their immediate neighbours, including those just across the sea. Therefore how different from their neighbours are they in fact? Or are their claimed differences simply tribal? And finally, is one quarter of their ancestry from the Sea People?

Sam said...

"E1b1b certainly reaches its highest frequency on the coastal regions around the Eastern Mediterranean, more than 20%. Perhaps spread by the Sea People, whoever they were?"

I read other comments associating E-M78/T with Sea people, which make sense, but I disagree with the part where the person claimed their ancestory is Somalian. Somalians are one big group that has E-M78 & T distributed amongst its clans so they had to be already one group when they entered Somalia (still a scarcely populated region till this day which explains the high %), if their was no other groups that had segregated E-M78 & T in high % I would have accepted a Somalian ultimate origin of the Sea People E-M78/T cluster)

For the cluster to be mathmatically perfect the stats has to look like this:

Somalia 78% E-M78 10.4%T
Lebanon-Greece 21% E-M78 3% T
Siciliy 17% E-M78 2.5 T%


Which is near identical to the actual ratio of E-M78/T

Balkans
Kosova has 45% E-M78 with 0*% T

While Serbia has 7% T & a much lower E-M78(older stats use to show Serbia high because of Kosovars), Kosova is the second M78 nation right after Somalia & Serbia is the highest T % nation in Europe, add the proximity between Serbia & Kosova & it will make alot of sense.

Since E-M78 & T people are usually fast moving populations we can't be certain where they started expanding as one (Sea People -Assumed one people-) The expansion of the one unified E-M78/T group could have came from anywhere in the Mediteranean & we should revisit Herods record of the Phoenicians origin, it makes more sense now!

Sam said...

And Israelis share half their Y-chromosomes, in the same ratio, with their immediate neighbours, including those just across the sea

I think Israeli Y-DNA is a special case where you have to take each group aside, then get % to the total population, so we don't get alot of Confusion. Many of these groups added some Y-DNA here & there that made them genetically similar to groups that they didn't necessarily interact with directly! (its just a case of common genotype combos)

Ashkenazi Y-DNA ~ Kurds-Turks
Sepharadi Y-DNA ~ Sicilians-Maltese
Mizrahi Y-DNA ~ Iraqis-Syrians
J-Yemenite Y-DNA ~ Saudis
J-Ethiopian Y-DNA ~ Amhara

All in all Israel as a nation (indluding Israeli Arabs) adds up to:

J > 50%
E > 25%
R ~ 20%


Non-Israeli jews, usually have higher R1% & lower J & E%, mainly because Jewish lineage is passed Maternally. However, MTDNA is another story its isolated & somewhat exclusive to each jewish sect with most sects sharing the same MTDNA with another sect.

*Portion of the Ashkenazis MTDNA ~ % with Morroccan MTDNA
*Portion of the J-Moroccan MTDNA shares ~ % with Sepharadi MTDNA
*Portion of the Sepharadi MTDNA shares ~ % with Mizrahi MTDNA
*Portion of the Mizrahi MTDNA shares ~ % with J-Yemenite MTDNA
*Portion of the J-Yemenite MTDNA shares ~ % with J-Ethiopian MTDNA

MTDNA amongst the Israeli diff sects is geographic more than anything, because many of these sects shared their MTDNA with the locals & soem MTDNA was already shared before.

terryt said...

A couple of comments Comf. "I would have accepted a Somalian ultimate origin of the Sea People". If the 'Sea People' could travel easily around the Eastern Mediterranean surely they could travel easily around the Red Sea. In other words the "Somalian ultimate origin" may be the Sea People.

"Many of these groups added some Y-DNA here & there that made them genetically similar to groups that they didn't necessarily interact with directly" Isn't it more likely that it's the opposite? The groups were originally simply members of a geographic continuum. They are fundamentally the same people. Genetic difference between Israelis and their neighbours is mainly a result of gene flow into the geographically isolated Jewish groups before they returned to Israel.

Sam said...

If the 'Sea People' could travel easily around the Eastern Mediterranean surely they could travel easily around the Red Sea. In other words the "Somalian ultimate origin" may be the Sea People.

Yes I accept that theory, but I don't like to wholesale a theory or even try to back it up, so I don't end up like these guys saying J2 is the Phoenician marker!

I rather seperate
(origin of T)?
(Origin of E1b1b & E ?)

The E1b1b & T cluster origin imo is the Balkans, but not all E-M78/T clusters are 8/1 (in Oman & Egypt diff ratio & subclades)

E1b1b1 people in general are coastal dwellers so interaction with T are well travelled & distributed all over Eurasia, so we can't be very sure. However, it still by far less outlandish than claiming J2 is the Phoenician marker in the Med Sea!

If E1b1b were not Coastal dwellers their will not be no Sea people.



Genetic difference between Israelis and their neighbours is mainly a result of gene flow into the geographically isolated Jewish groups before they returned to Israel.

At this point Israel (all in all) is closer to the Levant than any other region, but when they start doing tests by selecting certain groups that changes a bit.

Same with Levantines, for Example the (Lebanese Druze) DNA alone will make them seem as a Hybrid Anatolian-South Asian people, then it could be published as Lebanese data. Ignoring the fact that Lebanese Druze are only 5% of the population of Lebanon & they should have 5% sample representation.

Ofcourse many of these studies serve short term political interests...

Ebizur said...

Here are the data from this study's Table S3 for haplogroups E1b1b, T, G, and I. Beside the name of each haplogroup, I have listed, in order from greater to lesser frequency, the five regions in which this haplogroup was most frequently found. Each island is counted as a region unto itself; the ranking for a country, such as "Italy" or "Greece," is calculated [i]without[/i] the data for any islands that belong to the respective country.

Haplogroup E1b1b: Tunisia, Egypt, Greece, Sicily, Serbia
N. Egypt: 23/43 = 53.5%
Non‐contact Tunisia: 46/94 = 48.9%
S. Greece: 20/46 = 43.5%
Coastal Tunisia: 14/36 = 38.9%
N. Greece: 34/96 = 35.4%
S. Egypt: 9/29 = 31.0%
E. Sicily: 25/87 = 28.7%
S. Italy: 18/68 = 26.5%
Italy NEL*: 14/55 = 25.5%
Serbia: 18/81 = 22.2%
Greece: 17/78 = 21.8%
Albania: 11/51 = 21.6%
Italy WCP: 15/71 = 21.1%
N. Portugal: 21/101 = 20.8%
Greece: 16/77 = 20.8%
Non-Phoenician Levant: 58/286 = 20.3%
Cyprus: 13/65 = 20.0%
W. Sicily: 24/125 = 19.2%
Minorca: 7/37 = 18.9%
Phoenician Periphery: 112/700 = 16.0%
Italy WCL: 9/57 = 15.8%
Phoenician Heartland: 88/558 = 15.8%
Turkey #1: 8/52 = 15.4%
Turkey #4: 12/82 = 14.6%
N. Sardinia: 11/86 = 12.8%
Turkey #9: 10/81 = 12.3%
Non‐contact Iberia: 47/383 = 12.3%
S. Turkey: 4/33 = 12.1%
Turkey #6: 4/33 = 12.1%
Italy TLB*: 9/79 = 11.4%
Lowland Crete: 14/127 = 11.0%
Valencia Spain: 8/73 = 11.0%
S. Sardinia: 20/187 = 10.7%
Turkey #5: 4/43 = 9.3%
S. Spain: 15/167 = 9.0%
Turkey #7: 8/90 = 8.9%
Croatia: 8/90 = 8.9%
Turkey #3: 7/83 = 8.4%
S. Portugal: 8/100 = 8.0%
Ibiza: 4/54 = 7.4%
Italy EBL*: 7/95 = 7.4%
N. Turkey: 8/112 = 7.1%
Turkey #8: 2/30 = 6.7%
Mallorca: 4/62 = 6.5%
Malta: 12/187 = 6.4%
Italy SLA*: 3/51 = 5.9%
Italy CMA: 3/59 = 5.1%
Lasithi Plateau Crete: 2/41 = 4.9%
Italy NWA: 2/46 = 4.3%
Turkey #2: 1/29 = 3.4%

Haplogroup T: Ibiza, Egypt, Crete, Tunisia, Cyprus
Ibiza: 9/54 = 16.7%
S. Egypt: 3/29 = 10.3%
Lasithi Plateau Crete: 3/41 = 7.3% (This must be based on the data of Martinez et al. (2007), but Zalloua et al. have copied it as 3/40 = 7.5%)
Non‐contact Tunisia: 5/94 = 5.3%
Turkey #5: 2/43 = 4.7%
Cyprus: 3/65 = 4.6%
E. Sicily: 4/87 = 4.6%
Phoenician Heartland: 25/558 = 4.5%
Lowland Crete: 5/127 = 3.9%
Turkey #9: 3/81 = 3.7%
Turkey #7: 3/90 = 3.3%
Turkey #8: 1/30 = 3.3%
W. Sicily: 4/125 = 3.2%
Phoenician Periphery: 22/700 = 3.1%
S. Turkey: 1/33 = 3.0%
Turkey #6: 1/33 = 3.0%
S. Italy: 2/68 = 2.9%
Coastal Tunisia: 1/36 = 2.8%
Greece: 2/77 = 2.6%
Non‐Phoenician Levant: 7/286 = 2.4%
Turkey #3: 2/83 = 2.4%
N. Egypt: 1/43 = 2.3%
S. Greece: 1/46 = 2.2%
N. Greece: 2/96 = 2.1%
N. Turkey: 2/112 = 1.8%
Mallorca: 1/62 = 1.6%
Valencia Spain: 1/73 = 1.4%
Greece: 1/78 = 1.3%
Turkey #4: 1/82 = 1.2%
Turkey #2: 0/29 = 0.0%
Minorca: 0/37 = 0.0%
Italy NWA: 0/46
Albania: 0/51
Italy SLA*: 0/51
Turkey #1: 0/52
Italy NEL*: 0/55
Italy WCL: 0/57
Italy CMA: 0/59
Italy WCP: 0/71
Italy TLB*: 0/79
Serbia: 0/81
N. Sardinia: 0/86 = 0.0%
Croatia: 0/90
Italy EBL*: 0/95
S. Portugal: 0/100
N. Portugal: 0/101
S. Spain: 0/167 = 0.0%
Malta: 0/187 = 0.0%
S. Sardinia: 0/187 = 0.0%
Non‐contact Iberia: 0/383 = 0.0%

Haplogroup G: Sardinia, Ibiza, Italy, Egypt, Turkey
N. Sardinia: 18/86 = 20.9%
Turkey #8: 6/30 = 20.0%
S. Egypt: 5/29 = 17.2%
Turkey #3: 13/83 = 15.7%
Italy TLB*: 12/79 = 15.2%
S. Italy: 10/68 = 14.7%
N. Turkey: 16/112 = 14.3%
Turkey #5: 6/43 = 14.0%
S. Sardinia: 26/187 = 13.9%
Italy NWA: 6/46 = 13.0%
Ibiza: 7/54 = 13.0%
W. Sicily: 16/125 = 12.8%
Italy NEL*: 7/55 = 12.7%
Turkey #7: 11/90 = 12.2%
Italy SLA*: 6/51 = 11.8%
Italy WCP: 8/71 = 11.3%
Italy EBL*: 10/95 = 10.5%
Italy WCL: 6/57 = 10.5%
Turkey #2: 3/29 = 10.3%
Malta: 17/187 = 9.1%
Greece: 7/77 = 9.1%
S. Turkey: 3/33 = 9.1%
Turkey #6: 3/33 = 9.1%
Turkey #4: 7/82 = 8.5%
Turkey #9: 6/81 = 7.4%
Lasithi Plateau Crete: 3/41 = 7.3%
Lowland Crete: 9/127 = 7.1%
S. Portugal: 7/100 = 7.0%
N. Egypt: 3/43 = 7.0%
Italy CMA: 4/59 = 6.8%
Non‐Phoenician Levant: 19/286 = 6.6%
Phoenician Periphery: 46/700 = 6.6%
Non‐contact Iberia: 25/383 = 6.5%
S. Greece: 3/46 = 6.5%
Mallorca: 4/62 = 6.5%
N. Greece: 5/96 = 5.2%
N. Portugal: 5/101 = 5.0%
E. Sicily: 4/87 = 4.6%
Phoenician Heartland: 22/558 = 3.9%
Turkey #1: 2/52 = 3.8%
S. Spain: 6/167 = 3.6%
Greece: 2/78 = 2.6%
Albania: 1/51 = 2.0%
Valencia Spain: 1/73 = 1.4%
Serbia: 1/81 = 1.2%
Croatia: 1/90 = 1.1%
Coastal Tunisia: 0/36
Minorca: 0/37
Cyprus: 0/65
Non‐contact Tunisia: 0/94

Haplogroup I: Croatia, Serbia, Sardinia, Albania, Greece
Croatia: 66/90 = 73.3%
Serbia: 29/81 = 35.8%
S. Sardinia: 65/187 = 34.8%
N. Sardinia: 24/86 = 27.9%
S. Greece: 11/46 = 23.9%
Italy EBL*: 19/95 = 20.0%
Albania: 10/51 = 19.6%
Greece: 15/77 = 19.5%
Turkey #1: 8?/52 = 15.4%
Italy WCP: 9/71 = 12.7%
N. Greece: 12/96 = 12.5%
W. Sicily: 14/125 = 11.2%
Turkey #2: 3/29 = 10.3%
Lowland Crete: 13/127 = 10.2%
Turkey #9: 8/81 = 9.9%
Valencia Spain: 7/73 = 9.6%
Malta: 17/187 = 9.1%
Italy CMA: 5/59 = 8.5%
Mallorca: 5/62 = 8.1%
Italy SLA*: 4/51 = 7.8%
Greece: 6/78 = 7.7%
Cyprus: 5/65 = 7.7%
Lasithi Plateau Crete: 3/41 = 7.3%
S. Portugal: 7/100 = 7.0%
N. Portugal: 7/101 = 6.9%
Turkey #8: 2/30 = 6.7%
S. Spain: 10/167 = 6.0%
S. Italy: 4/68 = 5.9%
Phoenician Heartland: 32/558 = 5.7%
Non‐contact Iberia: 21/383 = 5.5%
Italy NEL*: 3/55 = 5.5%
Italy TLB*: 4/79 = 5.1%
Turkey #5: 2/43 = 4.7%
E. Sicily: 4/87 = 4.6%
N. Turkey: 5/112 = 4.5%
Turkey #7: 4/90 = 4.4%
S. Egypt: 1/29 = 3.4%
Non-Phoenician Levant: 9/286 = 3.1%
S. Turkey: 1/33 = 3.0%
Turkey #6: 1/33 = 3.0%
Phoenician Periphery: 21/700 = 3.0%
Minorca: 1/37 = 2.7%
Turkey #3: 2/83 = 2.4%
Ibiza: 1/54 = 1.9%
Italy WCL: 1/57 = 1.8%
Coastal Tunisia: 0/36
N. Egypt: 0/43
Italy NWA: 0/46
Turkey #4: 0/82
Non‐contact Tunisia: 0/94

Ebizur said...

Also, please note that the data for the "S. Turkey" and "Turkey #6" entries in this study's Table S3 appear to be derived from one and the same sample. It seems that the authors have failed to point out this duplication.

Ebizur said...

ERRATUM: Turkey had slightly more haplogroup G (73/635 = 11.5%) than Egypt (8/72 = 11.1%). Sorry for mixing them up. The order should have been Sardinia (44/273 = 16.1% G), Ibiza (7/54 = 13.0% G), Italy (69/581 = 11.9% G), Turkey, Egypt.

Haplogroup J(xJ2): "Non-Phoenician Levant," "Phoenician Homeland," Egypt, Turkey, Cyprus

Non-Phoenician Levant: 83/286 = 29.0%
Phoenician Periphery: 191/700 = 27.3%
S. Egypt: 6/29 = 20.7%
Phoenician Heartland: 96/558 = 17.2%
Turkey #1: 8/52 = 15.4%
S. Turkey: 5/33 = 15.2%
Turkey #6: 5/33 = 15.2%
Turkey #4: 10/82 = 12.2%
Turkey #3: 8/83 = 9.6%
N. Egypt: 4/43 = 9.3%
N. Turkey: 10/112 = 8.9%
N. Greece: 8/96 = 8.3%
Turkey #9: 6/81 = 7.4%
S. Portugal: 7/100 = 7.0%
Turkey #5: 3/43 = 7.0%
E. Sicily: 6/87 = 6.9%
Turkey #2: 2/29 = 6.9%
Cyprus: 4/65 = 6.2%
Italy WCP: 4/71 = 5.6%
Turkey #7: 5/90 = 5.6%
Italy CMA: 3/59 = 5.1%
S. Italy: 3/68 = 4.4%
W. Sicily: 5/125 = 4.0%
Albania: 2/51 = 3.9%
Italy WCL: 2/57 = 3.5%
Non-contact Iberia: 13/383 = 3.4%
Turkey #8: 1/30 = 3.3%
Valencia Spain: 2/73 = 2.7%
S. Sardinia: 5/187 = 2.7%
Lasithi Plateau Crete: 1/41 = 2.4%
S. Spain: 4/167 = 2.4%
Italy NWA: 1/46 = 2.2%
S. Greece: 1/46 = 2.2%
Mallorca: 1/62 = 1.6%
Lowland Crete: 2/127 = 1.6%
Greece: 1/77 = 1.3%
Greece: 1/78 = 1.3%
N. Portugal: 1/101 = 1.0%
Coastal Tunisia: 0/36
Minorca: 0/37
Italy SLA*: 0/51
Ibiza: 0/54
Italy NEL*: 0/55
Italy TLB*: 0/79
Serbia: 0/81
N. Sardinia: 0/86
Croatia: 0/90
Non-contact Tunisia: 0/94
Italy EBL*: 0/95
Malta: 0/187

I wonder how the Maltese acquired their Arabic language. This study's data show zero J(xJ2) and T in Malta.

Ebizur said...

Haplogroup J2: Crete, Cyprus, Malta, "Phoenician Homeland," Turkey
Lowland Crete: 59/127 = 46.5%
Cyprus: 24/65 = 36.9%
Italy CMA: 21/59 = 35.6%
Italy WCL: 20/57 = 35.1%
Malta: 59/187 = 31.6%
Turkey #1: 16/52 = 30.8%
Turkey #7: 26/90 = 28.9%
E. Sicily: 25/87 = 28.7%
Turkey #4: 23/82 = 28.0%
Phoenician Heartland: 156/558 = 28.0%
Turkey #2: 8/29 = 27.6%
Turkey #5: 11/43 = 25.6%
Italy SLA*: 13/51 = 25.5%
S. Turkey: 8/33 = 24.2%
Turkey #6: 8/33 = 24.2%
Albania: 12/51 = 23.5%
Phoenician Periphery: 155/700 = 22.1%
N. Turkey: 24/112 = 21.4%
Greece: 16/78 = 20.5%
Turkey #9: 16/81 = 19.8%
Italy WCP: 14/71 = 19.7%
S. Greece: 9/46 = 19.6%
Turkey #3: 16/83 = 19.3%
Italy TLB*: 15/79 = 19.0%
W. Sicily: 23/125 = 18.4%
Italy NWA: 8/46 = 17.4%
N. Greece: 16/96 = 16.7%
S. Italy: 11/68 = 16.2%
Greece: 12/77 = 15.6%
Italy NEL*: 8/55 = 14.5%
Non-Phoenician Levant: 41/286 = 14.3%
Coastal Tunisia: 5/36 = 13.9%
Turkey #8: 4/30 = 13.3%
S. Portugal: 11/100 = 11.0%
S. Spain: 18/167 = 10.8%
S. Sardinia: 20/187 = 10.7%
Lasithi Plateau Crete: 4/41 = 9.8%
N. Egypt: 4/43 = 9.3%
Serbia: 7/81 = 8.6%
Italy EBL*: 8/95 = 8.4%
Mallorca: 5/62 = 8.1%
N. Sardinia: 6/86 = 7.0%
N. Portugal: 7/101 = 6.9%
Non-contact Iberia: 26/383 = 6.8%
Valencia Spain: 4/73 = 5.5%
Ibiza: 2/54 = 3.7%
S. Egypt: 1/29 = 3.4%
Minorca: 1/37 = 2.7%
Croatia: 1/90 = 1.1%
Non‐contact Tunisia: 0/94


Haplogroup L: Turkey, "Phoenician Homeland," "Non-Phoenician Levant," Greece
Turkey #3: 10/83 = 12.0%
N. Turkey: 10/112 = 8.9%
Phoenician Heartland: 41/558 = 7.3%
Turkey #9: 4/81 = 4.9%
Turkey #4: 3/82 = 3.7%
Turkey #7: 3/90 = 3.3%
Turkey #8: 1/30 = 3.3%
Phoenician Periphery: 20/700 = 2.9%
Turkey #1: 1/52 = 1.9%
Greece: 1/78 = 1.3%
N. Greece: 1/96 = 1.0%
Non-Phoenician Levant: 2/286 = 0.7%
Turkey #2: 0/29
S. Egypt: 0/29
S. Turkey: 0/33
Turkey #6: 0/33
Coastal Tunisia: 0/36
Minorca: 0/37
Lasithi Plateau Crete: 0/41
Turkey #5: 0/43
N. Egypt: 0/43
S. Greece: 0/46
Italy NWA: 0/46
Albania: 0/51
Italy SLA*: 0/51
Ibiza: 0/54
Italy NEL*: 0/55
Italy WCL: 0/57
Italy CMA: 0/59
Mallorca: 0/62
Cyprus: 0/65
S. Italy: 0/68
Italy WCP: 0/71
Valencia Spain: 0/73
Greece: 0/77
Italy TLB*: 0/79
Serbia: 0/81
N. Sardinia: 0/86
E. Sicily: 0/87
Croatia: 0/90
Non-contact Tunisia: 0/94
Italy EBL*: 0/95
S. Portugal: 0/100
N. Portugal: 0/101
W. Sicily: 0/125
Lowland Crete: 0/127
S. Spain: 0/167
Malta: 0/187
S. Sardinia: 0/187
Non-contact Iberia: 0/383


Haplogroup R1a: Serbia, Greece, Croatia, Crete, Albania
Lasithi Plateau Crete: 8/41 = 19.5%
N. Greece: 18/96 = 18.8%
Greece: 12/77 = 15.6%
Serbia: 11/81 = 13.6%
Croatia: 11/90 = 12.2%
Turkey #5: 5/43 = 11.6%
Greece: 9/78 = 11.5%
Turkey #4: 9/82 = 11.0%
Albania: 5/51 = 9.8%
S. Turkey: 3/33 = 9.1%
Turkey #6: 3/33 = 9.1%
Turkey #9: 7/81 = 8.6%
Lowland Crete: 10/127 = 7.9%
Italy NWA: 3/46 = 6.5%
Turkey #7: 5/90 = 5.6%
Malta: 10/187 = 5.3%
Italy TLB*: 4/79 = 5.1%
Turkey #3: 4/83 = 4.8%
Italy SLA*: 2/51 = 3.9%
Turkey #1: 2/52 = 3.8%
Phoenician Periphery: 25/700 = 3.6%
N. Turkey: 4/112 = 3.6%
Turkey #8: 1/30 = 3.3%
Cyprus: 2/65 = 3.1%
S. Italy: 2/68 = 2.9%
Italy WCP: 2/71 = 2.8%
Valencia Spain: 2/73 = 2.7%
Minorca: 1/37 = 2.7%
Phoenician Heartland: 14/558 = 2.5%
W. Sicily: 3/125 = 2.4%
S. Spain: 4/167 = 2.4%
N. Egypt: 1/43 = 2.3%
E. Sicily: 2/87 = 2.3%
S. Greece: 1/46 = 2.2%
S. Portugal: 2/100 = 2.0%
Italy NEL*: 1/55 = 1.8%
Italy WCL: 1/57 = 1.8%
Italy CMA: 1/59 = 1.7%
Non-contact Tunisia: 1/94 = 1.1%
Italy EBL*: 1/95 = 1.1%
Non-Phoenician Levant: 3/286 = 1.0%
Non-contact Iberia: 4/383 = 1.0%
S. Sardinia: 1/187 = 0.5%
Turkey #2: 0/29
S. Egypt: 0/29
Coastal Tunisia: 0/36
Ibiza: 0/54
Mallorca: 0/62
N. Sardinia: 0/86
N. Portugal: 0/101


Haplogroup R1b: Balearic Islands, Iberian Peninsula, Italy, Sicily, Malta
Minorca: 27/37 = 73.0%
Mallorca: 41/62 = 66.1%
S. Spain: 108/167 = 64.7%
Valencia Spain: 47/73 = 64.4%
Non-contact Iberia: 233/383 = 60.8%
Ibiza: 31/54 = 57.4%
Italy EBL*: 50/95 = 52.6%
N. Portugal: 53/101 = 52.5%
Italy NWA: 24/46 = 52.2%
S. Portugal: 52/100 = 52.0%
Italy TLB*: 32/79 = 40.5%
Italy NEL*: 21/55 = 38.2%
Italy CMA: 22/59 = 37.3%
Italy SLA*: 19/51 = 37.3%
Lasithi Plateau Crete: 15/41 = 36.6%
Italy WCP: 24/71 = 33.8%
Italy WCL: 18/57 = 31.6%
Turkey #2: 9/29 = 31.0%
W. Sicily: 34/125 = 27.2%
Greece: 21/78 = 26.9%
S. Italy: 17/68 = 25.0%
Malta: 41/187 = 21.9%
S. Turkey: 7/33 = 21.2%
Turkey #6: 7/33 = 21.2%
Turkey #8: 6/30 = 20.0%
N. Sardinia: 17/86 = 19.8%
S. Greece: 9/46 = 19.6%
E. Sicily: 17/87 = 19.5%
S. Sardinia: 36/187 = 19.3%
N. Turkey: 20/112 = 17.9%
Albania: 9/51 = 17.6%
Turkey #7: 14/90 = 15.6%
Turkey #9: 12/81 = 14.8%
N. Greece: 14/96 = 14.6%
Turkey #5: 6/43 = 14.0%
S. Egypt: 4/29 = 13.8%
Turkey #1: 7/52 = 13.5%
Turkey #4: 11/82 = 13.4%
Turkey #3: 11/83 = 13.3%
Lowland Crete: 15/127 = 11.8%
Greece: 9/77 = 11.7%
N. Egypt: 4/43 = 9.3%
Cyprus: 6/65 = 9.2%
Phoenician Heartland: 49/558 = 8.8%
Non-contact Tunisia: 8/94 = 8.5%
Serbia: 5/81 = 6.2%
Phoenician Periphery: 33/700 = 4.7%
Coastal Tunisia: 1/36 = 2.8%
Croatia: 2/90 = 2.2%
Non-Phoenician Levant: 6/286 = 2.1%

terryt said...

Thanks for all that Ebizur. It'll take weeks to work my way through it.

But I noticed "Haplogroup E1b1b: Tunisia, Egypt, Greece, Sicily, Serbia". I remember asking here some time ago, in relation to E1b1b in Albania, if it's possible that the haplogroup originated in North Africa. Seems it may have, Tunisia. We know from Egyptian records that the Sea People were allied at times with 'Libyans', probably refering to a wider stretch of North African coast than bears the name today. If this is so the Y-chromosome would have been introduced to Phoenicia from the sea rather than originating there or from inland.

Judith Weingarten said...

Regarding the Sea Peoples, archaeologically speaking, they were never numerous enough to spread about their Chromosomes as some of you are suggesting. Even the main group in the Levant, which came into history as the inhabitants of the Philistine cities, had merged into the Canaanite population by Biblical times: they are the only Greek-speakers (as far as I know) who ever entirely lost their language. They can not have been a great multitude, and certainly too few to have influenced the 'Libyans' as well.

Ebizur said...

The most obvious pattern in these data is the positive correlation between haplogroup I and haplogroup R1a. In this Mediterranean context, both these haplogroups are associated with the Balkans (including Greece), and, to a somewhat lesser degree, Anatolia.

By the way, I should add that my attempted correction to the number of haplogroup I individuals in the "Turkey #1" sample seems to have missed the mark. This sample appears to correspond to the "Region 1" sample of Cinnioglu et al. ("Excavating Y-chromosome haplotype strata in Anatolia," Human Genetics, 2004), which contains 1/52 I-P37(xM359), 1/52 I-M223, 2/52 I-M170(xP37, M223, M253), and 2/52 I-M253 for a total of 6/52 = 11.5% haplogroup I. Cinnioglu's "Region 1" represents the Marmara region of Turkey, which includes Thrace/Rumelia. Zalloua et al. (2008) reproduce the haplogroup I data from this sample as "Turkey #1," n(Total)=52, n(HgI)=6, n(HgI)/n(Total) = 0.155, and I had corrected this to "8?/52 = 15.4%," but it looks like Zalloua et al. have recorded incorrectly the frequency of haplogroup I in this sample, rather than the number of haplogroup I individuals in the sample.

Dienekes Pontikos said...

The most obvious pattern in these data is the positive correlation between haplogroup I and haplogroup R1a.

I would warn against the belief that these two are historically related. Within the Balkans they are not correlated. Also, R1a is concentrated in the eastern provinces of Turkey rather than the ones closest to the Balkans.

But, certainly the proposition that the Balkans have more I and R1a than West Asia is correct.

Ebizur said...

dienekes said,

"I would warn against the belief that these two are historically related. Within the Balkans they are not correlated. Also, R1a is concentrated in the eastern provinces of Turkey rather than the ones closest to the Balkans."

Well, the fact is that four out of five in the lists for haplogroup I and haplogroup R1a are the same: (1) Croatia, (2) Serbia, (4) Albania, (5) Greece for haplogroup I and (1) Serbia, (2) Greece, (3) Croatia, (5) Albania for haplogroup R1a. You are correct that within the Balkans taken alone, there is not a clearly positive correlation between the frequency of hg I and the frequency of hg R1a, but that would make some sense, because if hg I is larger in one population then there is that much less of the pie for hg R1a to occupy, and vice versa. My point is only that, in the greater Mediterranean region, haplogroup I and haplogroup R1a are clearly associated with the Balkans as a whole.


dienekes said,

"But, certainly the proposition that the Balkans have more I and R1a than West Asia is correct."

Not only do the Balkans have more I and R1a than West Asia, but they also have more of these haplogroups than Italy, Iberia, the Mediterranean islands (except Crete, which has substantial amounts of both I and R1a, and Sardinia, which has a very high frequency of haplogroup I but an extremely low frequency of haplogroup R1a), and North Africa as well.

terryt said...

Ebizur, no mention of Israel, or any Jewish groups, in any of your lists for comparison. Any additions you'd like to make?

Way back Comf wrote: "I think Israeli Y-DNA is a special case".

Why is it any different to anywhere else? Surely "you have to take each group aside, then get % to the total population, so we don't get a lot of confusion" for groups within any modern political boundaries.

Judith Weingarten wrote: "[Sea People] were never numerous enough to spread about their Chromosomes as some of you are suggesting. Even the main group in the Levant, which came into history as the inhabitants of the Philistine cities, had merged into the Canaanite population by Biblical times".

For a start we could argue as to whether the Philistines were in fact actually the main group. I'd argue that others survived long into Biblical times as well. But the fact they "had merged into the Canaanite population by Biblical times" in no way, on its own, negates a substantial movement. Suppose the migration was mostly of men with improved boating technology. Substantial numbers could have arrived, but had children with the local women. These would have been bilingual, speaking their mothers' language as well as any incoming one. Once the migration flow had ceased the original Semitic language would readily become re-established, just as English was after the Norman conquest.

Judith Weingarten said...

Terryt,

The Sea Peoples are archaeologically visible (mainly through pottery) only in the Philistine cities on the southern coastal plain of Canaan. These are built on the ruins of Canaanite cities and clearly reflect the Aegean background of the new inhabitants. Where else would you like to put the Sea Peoples?

My point about the loss of Greek is that there is no trace of this tongue in the languages subsequently spoken in the area(unlike post-Norman English which most certainly shows large dollops of Norman French).

terryt said...

Judy. As I understand it 'Philistine' pottery ( a hybrid between Mycenaean and Levantine) is quite widely distributed, right across to the Jordan River. Possibly through trade I'll admit.

"Where else would you like to put the Sea Peoples?"

Apart from the obvious connection between Peleset and Philistines. From being in the same place at the same time: how about Tjekker and Manasseh. From a similarity of names at about the same place and time: Denyen and Dan along with Weshesh and Asher. There is also often a connection made between Shekelesh and Issachar. The 'Sea People' could have been quite widespread.

Sam said...

Coastal Tunisia: 0/36
Non-contact Tunisia: 0/94
Malta: 0/187

I wonder how the Maltese acquired their Arabic language. This study's data show zero J(xJ2) and T in Malta.


Tunisia also shows a zero...although older studies had 6% & 10een% J1 in Tunisia, so it could be the same in Malta

I think most of the Maltese descend from Siculo-Arabic speaking Sicilian-Calabrian refugees who were cornered in Noto (the last Siclian muslim town). Noto is the closest part of Sicily to Malta...they also share the same genes

Ebizur said...

comf said,

"Tunisia also shows a zero...although older studies had 6% & 10een% J1 in Tunisia, so it could be the same in Malta

I think most of the Maltese descend from Siculo-Arabic speaking Sicilian-Calabrian refugees who were cornered in Noto (the last Siclian muslim town). Noto is the closest part of Sicily to Malta...they also share the same genes"

I singled out the Maltese in my previous post because I was grouping J(xJ2) and T together as examples of Y-DNA haplogroups that might typically be found among Arabic-speaking peoples, and the Tunisian samples of Zalloua et al. actually did contain a fair number of haplogroup T individuals, so I could not have justly described the Tunisian samples as lacking both J(xJ2) and T.

However, you have made some very good points here. I have also been thinking about the possibility of linking the Arabic language of the Maltese to the haplogroup J(xJ2), T, and E1b1b found in Sicily. Of these three haplogroups, the Maltese sample of Zalloua et al. contains only 12/187 = 6.4% E1b1b and no examples of J(xJ2) or T, but all three of these haplogroups are found quite frequently in Sicily. I predict that J(xJ2) and T should also show up with at least low frequency in a larger sample of Maltese, but this is mere speculation.

Furthermore, the Maltese sample of Zalloua et al. does contain a very high percentage of haplogroup J2 individuals, and at least some of this haplogroup J2 may have been brought to Malta by Semitic-speaking ancestors from Southwest Asia.

Here are some data on the Y-DNA of Jews and a few other Middle Eastern populations, courtesy of Nebel et al. (2001):

Muslim Kurds:
16/95 = 16.8% P(xR1a)
16/95 = 16.8% Y(xA, DE, J, K)
11/95 = 11.6% R1a
27/95 = 28.4% J2
11/95 = 11.6% J(xJ2)
7/95 = 7.4% E
4/95 = 4.2% K(xO2b-SRY465, N1c-Tat, L-M20, P-92R7)
3/95 = 3.2% L

Kurdish Jews:
20/99 = 20.2% P(xR1a)
6/99 = 6.1% Y(xA, DE, J, K)
4/99 = 4.0% R1a
15/99 = 15.2% J2
22/99 = 22.2% J(xJ2)
12/99 = 12.1% E
19/99 = 19.2% K(xO2b-SRY465, N1c-Tat, L-M20, P-92R7)
1/99 = 1.0% L

Sephardic Jews:
23/78 = 29.5% P(xR1a)
9/78 = 11.5% Y(xA, DE, J, K)
3/78 = 3.8% R1a
12/78 = 15.4% J2
10/78 = 12.8% J(xJ2)
15/78 = 19.2% E
6/78 = 7.7% K(xO2b-SRY465, N1c-Tat, L-M20, P-92R7)

Ashkenazi Jews:
9/79 = 11.4% P(xR1a)
5/79 = 6.3% Y(xA, DE, J, K)
10/79 = 12.7% R1a
19/79 = 24.1% J2
15/79 = 19.0% J(xJ2)
18/79 = 22.8% E
3/79 = 3.8% K(xO2b-SRY465, N1c-Tat, L-M20, P-92R7)

Palestinian Arabs:
12/143 = 8.4% P(xR1a)
9/143 = 6.3% Y(xA, DE, J, K)
2/143 = 1.4% R1a
2/143 = 1.4% A3b2-M13
24/143 = 16.8% J2
55/143 = 38.5% J(xJ2)
29/143 = 20.3% E
10/143 = 7.0% K(xO2b-SRY465, N1c-Tat, L-M20, P-92R7)

Bedouin:
2/32 = 6.25% Y(xA, DE, J, K)
3/32 = 9.4% R1a
1/32 = 3.1% J2
20/32 = 62.5% J(xJ2)
6/32 = 18.75% E

If all the haplogroup E individuals in these samples of Nebel et al. (2001) are assumed to belong to haplogroup E1b1b, then the Ashkenazim would bump Serbia out of the list for haplogroup E1b1b. As for the Sephardim, they appear to have notable amounts of haplogroup R1b (represented by P(xR1a) in the data of Nebel et al.) and haplogroup T (represented by K(xO2b, N1c, L, P)).

terryt said...

Ebizur. Thanks once more.

Ebizur said...

In addition, the haplogroup R1a ranking would be amended as follows:

1. Serbia (11/81 = 13.6%)
2. Greece (40/297 = 13.5%)
3. Ashkenazim (10/79 = 12.7%)
4. Croatia (11/90 = 12.2%)
5. Muslim Kurds (11/95 = 11.6%)

Crete (18/168 = 10.7%) and Albania (5/51 = 9.8%) got bumped off the list.

Judith Weingarten said...

Terryt,

It's not because of a similarity of names (Peleset, Philistine) that archaeologists put the Sea Peoples in Philistia from the early 12th C on -- but because of the archaeological assemblage which appears there and not all the way to the Jordan. Large quantities of pottery, of course, but also settlement patterns, cult, metallurgy and glyptic: what Trude Dothan called "the Philistine/Sea Peoples" culture which was rooted in Aegean traditions.

Lots of up-to-date information in the studies In Honor of Trude Dothan, Mediterranean Peoples in Transition (1998). And please don't call me 'Judy'. My name is Judith.

Sam said...

"Sephardic Jews:
23/78 = 29.5% P(xR1a)
9/78 = 11.5% Y(xA, DE, J, K)
3/78 = 3.8% R1a
12/78 = 15.4% J2
10/78 = 12.8% J(xJ2)
15/78 = 19.2% E
6/78 = 7.7% K(xO2b-SRY465, N1c-Tat, L-M20, P-92R7)"

Those are quite diff than Semino (its obvious now that samples just give us a general idea!)


Semino et al:
Sepheradic Jews
E = 30%
J1= 12%
J2= 29%

Semino et al:
Tunisians
J1 = 30% (compared to 0% in Zalloua!)
J2 = 4%
E = 55.2%

Ebizur said...

I should add that the Jews (Ashkenazi, Sephardi, Kurdish) and Arabs (Palestinian and Bedouin) tested by Nebel et al. (2001) are all residents of the state of Israel or the Palestinian Authority Area. I am not sure about the location in which the sample of Muslim Kurds has been obtained.

Sam said...

Phoenician Footprints in the Mediterranean

Biggest "footprint" = Tunisia

So could this explain?

Semino:
Tunisia J1 = 30%

Zalloua:
Tunisia J1 = 0%


This 0% was very necessary to prove that J2 sea faring farmer theory!

J2 in Tunisia in earlier studies accounts ~ 4%, should be a bit higher in Modern Carthage & most the fertile regions of Tunisia because they had an influx of Iberian moriscos that moved into the region 400 years ago. They are known in Tunisia as the Andalusians!

Dienekes said...

J2 in Tunisia in earlier studies accounts ~ 4%, should be a bit higher in Modern Carthage & most the fertile regions of Tunisia because they had an influx of Iberian moriscos that moved into the region 400 years ago. They are known in Tunisia as the Andalusians!

If you search yhrd for the "Phoenician Colonization Signals" you will see that at least some of them are found in "Andalusian Arabs" from Tunisia.

Zalloua:
Tunisia J1 = 0%


I hadn't noticed that. It is pure nonsense of course. Substantial frequency of F*(xH, I, J2, K) which should include J1 by and large was also found by Arredi et al. (Am. J. Hum. Genet. 75:338–345, 2004)

Sam said...

Substantial frequency of F*(xH, I, J2, K) which should include J1 by and large was also found by Arredi et al.

I didn't know about Arredi et al. it said (J* = 48% which is usually J1) OK so the Zalloua guys were sampling tourists!

J1 Tunisia
Arredi = 48%
Semino = 30%
Zalloua = 0%

How about this?

E1b1b Lebanon
Wells = 32%
Semino = 26%
Zalloua = 15.8%


So there is a trend of polishing J2 = Phoenician seafaring farmers when the older stats show E1b1b/T are stronger candidates

J2
Iraqis 30% (sanchez et al.)

Iraq is probablly the highest nation with J2 (but the study ignored Mesopotamia & made it seem that J2 fades in the Syrian deserts (obviously), but then it peaks in Mesopotamia where its naturally connected to Anatolia.

The study ignores:
- The Mesopotamian J2 Assyrian recorded invasions & the Aramization of Lebanon followed by the Greco-Roman occupation.

They ignored:
(1500years of post-Phoenician J2 in the -heartland-)

terryt said...

Sorry Judith. "It's not because of a similarity of names (Peleset, Philistine) that archaeologists put the Sea Peoples in Philistia from the early 12th C on".

I'm pretty sure that for the Egyptians the 'Sea People' included groups present in the Eastern Mediterranean long before the Philistines appeared there. Groups such as the Lukka, Sherden and Denyen, for example, were roaming around the region as early as the time of Akhenaten, so plenty of time to distribute any Y-chromosomes they had quite widely.

Abate Vella said...

I would like to point out another interesting fact that may add something to undermine the credibility of this study.

As you may now, Sicily and Malta and strongly connected. Actually Malta was until very recently just one of the islands belonging to Sicily.

Now, the study claims basically that Lebanese people and Maltese people show a very high genetic similarity.

There is a problem though: to my knowledge none of the inhabitant of Malta at the time of the Phoenician could have transmitted that genetic heritage to this day, because it appears that the island lost all of its inhabitant in later times.

Malta has been completely repopulated with people coming from Sicily, and if I remember correctly this happened when Sicily was a Muslim kingdom.

So the similarity between Maltese and Lebanese does not go back to Phoenicians times.

Abate Vella said...

Here there is something related to what I just said:

http://en.wikipedia.org/wiki/Harbi_al-Himyari

All the arguments against this idea are self-serving, as you can see.

The last point is at best elusive.

Abate Vella said...

Sorry to put so many comments,

but I found the article that I wanted:

http://forum.stirpes.net/genetics-human-microbiology/13332-genetic-origin-contemporary-maltese.html

Sam said...

Here there is something related to what I just said:

http://en.wikipedia.org/wiki/Harbi_al-Himyari


That article is wrong Harbi Al-Himyari was the mentor of Jabir Bin Hayyan (8th century AD)

Muhammad Al-Himyari is the geographic traveller with the famous book:

http://en.wikipedia.org/wiki/Kitab_al-Rawd_al-Mitar

He described the repopulation of Malta, it corresponds with the later link you posted linking the Maltese to 11th century Sicily-Calabria.

The muslims were pushed out of sicily North-West to South-East it was a slow process that took about 50 years that ended with the Muslims cornered in Noto which is the closest Sicilian port to Malta. Conversions must have toke place in the 12th & 13th century along with Sicily

AndyB1580 said...

I understand your misgivings about this study, but regarding the Cretian-Phoenician connection I have some additional data that seems to suggest just such a connection. I belong to a very rare haplotype within the J2a4d y-DNA haplogroup. I'm even an outlier in the J2 y-DNA groups, and even more an outlier in the English midlands where my family was recorded in the oldest surviving parish records. The scarcity of my haplotype suggested to me that my ancestors may have been the survivors of some conquered culture that was driven to near-extinction in ancient times.

1)The town where my family was recorded was originally a Roman port about 40km from a Roman fort.

2)The Roman legions that were stationed there were previously all in Cantabria, Spain which had alliances with Carthage and had a Carthagenean/Phoenician trade presence for exporting lead.

3)The Roman navy is said to have conscripted sailors from the conquered Carthageneans.

4)Other than a few unexplained outliers, the only other places where I have any reasonably close genetic matches are Cantabria, Sicily, Tunisia, Algiers, Turkey and Lebanon - all places with potential Phoenican histories.

5)I match the "Phoenican" haplotypes cited in the Genographics project very closely.

6)My J2a4d y-DNA haplogroup is said to have originated in Crete.

7)Phoenicia only appeared in the historical record after the decline of the Minoan culture on Crete.

Circumstantial evidence, but there seems to be a consistent thread in it. Actually, I'm implying that Crete may have contributed a geneflow to Phoenicia (sea peoples?) not the reverse.

Rob said...

(very late comment)

The Poenecians cannot be define 'genetically' . They were an anceint maritime trading culture , now long gone. There is no way to isolate them by deductions from (alleged) modern day proxies, nor is there reason to assume they bore a distnict, single lineage

I would also be careful not to overstate the Greek case - Greek domination (if we can call it that) in the Mediterranena was predominantly cultural and economic, not military or demographic

Regards

YeomanDroid said...

Excellent catch on the apparent missing Greek colonists in Asia Minor. I was a bit perplexed why they were not even a consideration in this paper, especially when the history of Greece is fully evident of colonization in Anatolia.

Unknown said...

Has anyone taken into consideration that scholars say the Phoenicians originated from Bahrain? Everyone is quick to dismiss "Arab" Gulf DNA. What if the key to finding "Phoenician" (Semitic language) DNA is actually in the Arabian peninsula. Zalloua (and others) appears to want to distance whatever is Phoenician from the Arabian peninsula.

Let's look at the pattern of migration of man out of Africa into Arabia and into the Levant, as well as out of Africa directly into the Levant, and the settlements.

Greek historians place the origin of the Phoenicians as a people from what is now known as the Arab Gulf.

Zalloua is obviously very anti-Arab. Who is to say the Phoenician DNA marker doesn't originate from Arabia considering the migrational patterns?

Unknown said...

In ancient times "Libya" consisted of the majority of North Africa (including Egypt). Libya to the Egyptians and Greeks meant all of North Africa.

pamelaj said...

AndyB1580 I am similiar. My family is very well documented at least 26 gens. I am J2a2. I am Welsh and Scots/French. Its a well known that the Phoenicians traded with the Welsh for tin to make bronze. The most DNA matches I had in Wales were on Anglesey. In England, Staffordshire, 12th c. Two interesting traditions in those 26 generations: shipbuilders and metal workers. Genealogically, all my ancestors besides the Welsh came from the Cotentin, Normandy.
There's a common seafaring thread throughout this ancient story. The same people built ships and came to Boston in the 1630's already wealthy and carried on building ships. That's how J2a2 came to America in my story.
So are they Phoenicians-Carthaginians- Hispanic Romans-Vikings-or the original Brythonic people?