October 13, 2012

An estimate of the admixture time for Finns

Using a similar procedure as in my recent post on the Baltic (Update II), I used 15 FIN individuals from the 1000 Genomes together with 12 Nganasans from Rasmussen et al. (2010) as reference populations, and 15 other FIN individuals to estimate admixture LD in a rolloff analysis. Three outlier Nganasan individuals (GSM558800, GSM558802, GSM558807) were removed.
The estimated time of admixture is 86.095 +/- 10.187 generations, or 2500 +/- 300 years. It corresponds rather well to the beginning of the Iron Age in northern Europe.

As I mention in my previous post, there is evidence for intrusive cultures (Battle Axe and Seima Turbino) converging on the area from different directions during the preceding Bronze Age. If the above date is accurate, it will suggest a rather late admixture event between the Europeoid and Siberian elements of Finns. The former may have included both the descendants of Mesolithic European hunter-gatherers and intruders from Central Europe (Corded Ware/Battle Axe); the latter may have included both Comb Ceramic and the descendants of the Seima Turbino metallurgists.

34 comments:

hamarfox said...

Seems I'm a little late to the party when it comes to Reich et al's paper, but I wonder how we should recast the question of Finnish-Siberian relations in light of that paper's findings. Finns obviously stand out from other N. Europeans (bar Russians) in ADMIXTURE, and phys. anthropology always noted their relative excess of Uralic-influenced racial types. Yet Reich suggests a more general distibution of East Eurasian genes throughout Europe that doesn't gel too neatly with anthropology or ADMIXTURE.

If Finns experienced two waves of E. Eurasian-like admixture (as opposed to Western Europe's one), it seems likely that it was the second that delivered the Uralic traits that are absent in, for example, Britain and France, and that it's this second wave that is picked up by ADMIXTURE.

Strangely, then, exactly what was the nature of the first wave? Whatever E. Eurasian shift occurred as a result, it was outmatched by parallel processes in extra-European West Eurasians, as seen in your [URL="https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedC1XUjJMNnFBRnRvLXhpUFdRSWp0bGc&authkey=CJ7C6ugB&hl=en&authkey=CJ7C6ugB#gid=0"]spreadsheet[/url].

Dienekes said...

There are two alternatives regarding the general trend. One is that Mesolithic Europeans were Amerindian-shifted relative to Neolithic ones. Two is that the Y-haplogroup R folk, who seem to have invaded Europe during the last 5,000 years or so were Amerindian-shifted relative to the "Palaeo-West Eurasians" (I/J/G folk).

The latter makes sense to me, due to the ancient link between R and Q, although there seems to be some evidence for the former as well in the assessment of the Iberian Mesolithic Brana specimens.

hamarfox said...

Thanks for the reply. I definitely see merits to both theories. One thing that strikes me as odd about the idea of East Eurasian admixture being perculiar to Europe is that West Asians are equally Asian-shifted to Atlantic-Baltics/most individual European populations in Fst distance, and also in your African/Asian shift of West Eurasian populations spreadsheet.

At any rate, Finns, like Russians, are outliers in the trend, and I'd suppose a separate source/period of geneflow into those groups, and the Iron Age theory is interesting.

mikej2 said...

I think that the time 2500 years ago is quite a good estimation for the second Uralic movement from east. The first one consisted of the people who brought Finnic languages to the shores of Baltic Sea (Estonians and Finns). It happened 3500-4000 years ago according the linguistic timing. The second arrival is estimated to 2500 BC. It consisted of the people who brought Saami language. The original Saami population (being likely those Mesolithic Europeans with aforementioned older Asiand admix, claassified now Am. Indian?) mixed with this latter arrival, as well as Finns after crossing Gulf of Finland to the north. In fact Nganassans have very little to do with this considering datings. They however mixed with other Arctic groups being a proxy for North Siberian genes in light of the for me known history.

Onur said...

Hamar,

Patterson et al.'s and Dienekes' ADMIXTOOLS tests confirm that:

1) Finns, Russians and various Uralic and Uralic-admixed Caucasoid minorities of European Russia possess more Mongoloid admixture than all other Caucasoid populations of Europe and West Asia,

2) and, in turn, northern Caucasoids (whether from west or east) in general possess more Mongoloid admixture than southern Caucasoids.

As an example, see Dienekes' F4 ratio test:

http://3.bp.blogspot.com/-sK38RWxUSdw/UFbt-9RV3jI/AAAAAAAAGZ4/InEHr1SYOjs/s1600/F4ratio_Sardinian_Karitiana_seven.png

Note that according to this test Finns, Russians and the Mordva are followed by all the other northern European populations in general, who are in turn followed in general by the southern European and West Asian populations in Mongoloid admixture from the highest to the lowest.

You are correct in noting Finns' and Russians' comparatively elevated levels of Mongoloid physical features. But you are ignoring the fact that in general southern Caucasoids possess more distinctively Caucasoid and less Mongoloid facial features than northern Caucasoids (whether from west or east). This fact is quite often ignored by Nordicists and the Nordicist physical anthropological establishment. To see Caucasoids par excellence you have to look to south, not to north. So there is nothing in Patterson et al.'s and Dienekes' ADMIXTOOLS tests that contradicts physical anthropological results.

jackson_montgomery_devoni said...

So the ''Siberian'' component picked up by ADMIXTURE that is seen in Finns is of a rather young Uralic origin? Does that seem about right?

Fanty said...

@Mikej2:

That very "first wave" must have been something that included even France and Britain and is usualy not rated Uralic or Sibirian but even Amerindian (American Native)

It cant be connected to "Finns and Estonians".

Dienekes muses the original R1a and R1b people to be responsable for these signals, because R1a/R1b are related to "Q", the main haplogroup of American natives. The parental haplogroup of them was "P".

If I recall it right, this tool calculated the age admixture between "Northern European" and "American native" to have happend 17K years ago.

Matt said...

One thing that strikes me as odd about the idea of East Eurasian admixture being perculiar to Europe is that West Asians are equally Asian-shifted to Atlantic-Baltics/most individual European populations in Fst distance

Hamarfox,

I had some trouble with this at first. Dienekes or anyone else can correct me if I'm wrong (and perhaps I'm explaining something you already know):

Europeans can be the Caucasoids who are most drifted away from Native Americans, (even if less pushed away without African admixture than some other Caucasoids), as well as those who have been most admixed with a Native American like population relatively recently.

That Europeans are pushed further due to drift and pushed closer due to admixture, rather than just purely pushed away due to drift, would not obviously until you look at a parental like population which is just as drifted away from East Eurasians and less admixed (which Sardinians are and the various Middle Eastern and West Asian populations are not).

The two do not contradict. It's a combination of mixtures and drift (i.e. separate evolution, bottlenecking) that creates the total FST distance.
FST is a measure of total population differentiation - it doesn't measure when the differentiation happened or how, or how many events occured that promoted similarity or differentiation.

For comparison (I think) a group of Papuans each with a single European great grandparent would have higher FST to Europeans than East Asians do (?), even though the "Papuan" group would be admixed and the East Asians would not.

Slumbery said...

Fanty: You talk about something different than Mikej2. He did not talk about any "very first wave". The earlier Uralic wave he talks about is not picked up by this analysis, because this type analysis shows only the last significant admixture. These results leave place for other Uralic waves.

mikej2 said...

@Fanty

No, you are wrong. The first wave I was talking about was the Baltic Finnic wave about 3500-4000 years ago. The Saami people and language did not belong to this group, and Saami language is a distinct Uralic language.

What you are speaking about is IMO a different thing, being related to Central and West European and later to Scandinavia.

In rough datings without Central/West Europeans

- before 2000BP Mesolithic Europeans still lived in Finland and in some regions of Sweden and Norway. They had already disappeared from more southern places. They spoke unknown languages that are identified by linguists.

- 3500-4000BP Baltic Finnic people arrived the Baltic Sea region, maybe not seasides. They lived to the south of Gulf of Finland.

- 2500BP Uralic people speaking Saami language arrived to the Ladoga region and a bit later to Finland continuing to the north by seasides of Baltic Sea (Ostrobothnia). Neolithic Europeans shifted to this new language. Now all Saamis speak dialects or diverged languages of the old Saami.

- 1700-1800BP Baltic Finns who still lived on the southern shores of Gulf of Finland crossed the sea to the north and mixed with those Saami people who had already lived there 700 years. Maybe Finns were pushed by some other groups to the north. As a result of this Finns show today this genetic mixture, but because Finns pushed remains of Saamis northward, the more northern genetic pattern increases today in the north.

What happened next? The Slavic expansion about 1500-800BP. I dont go in this, it is another story in Baltic and Finnish areas.






Onur said...

On West Asians and their eastern shift:

West Asians (including the non-ethnic Russian peoples of the Caucasus) in general, the Iranic speaking ones and some groups in the northern Caucasus in particular, possess minor ASI (=Ancestral South Indian) admixture. Unlike the situation in Europe, where the eastern shift is entirely or almost entirely of Mongoloid origin, the eastern shift in West Asia is only partially of Mongoloid origin, the rest of it being of ASI origin.

Matt said...

But you are ignoring the fact that in general southern Caucasoids possess more distinctively Caucasoid and less Mongoloid facial features than northern Caucasoids (whether from west or east).

One thing to bear in mind is that the population mixed with has relatively more Native Americans and Siberian affinities than East Asian affinities (the signal of admixture is stronger with them than East Asians and South East Asians), but we don't know quite what they may have looked like.

In Tsuniko Hanihara's data (and I think this is generally supported by physical anthropology), Native Americans have fairly transversely projecting faces without much prognathism (I think particularly when accounting for facial size) and Siberians have very pinched, sharp nasalia (even compared to West Eurasians).

The less Caucasoid population that mixed to create Europeans may have had some or all of these features, leaving the Europeans not really noticeably more "Eurasian" like than the Caucasoids who did not mix in this fashion, at least in a way that we would naively expect (e.g. broad, flat nasals, shovel shaped incisors, high rounded orbits with low supraorbital projection, transversely flat faces, etc.), particularly since we're talking 10% influx and once the effects of natural selection and European unique drift are taken into account.

hamarfox said...

[i]Note that according to this test Finns, Russians and the Mordva are followed by all the other northern European populations in general, who are in turn followed in general by the southern European and West Asian populations in Mongoloid admixture from the highest to the lowest.[/i]

The problem is in calling it Mongoloid. On the chart you link to, populations in which Mongoloid phenotypes manifest are placed side by side with populations lacking in such types. Remember, 'East Eurasian' encompasses phenotypes from Mongoloid to Dravidian. Yes, the affinity in this case is NE Asian, yet the age is undetermined.

As to the chart, Hungarians (in whom Mongoloid phenotypes are relatively common) being lower in this 'Mongoloid' admixture than the Cornish (in whom Mongoloid phenotypes are absent) is patently absurd. Also in the chart, it appears SSA admixture is an anchor to the true Karitiana levels of certain populations

[i]You are correct in noting Finns' and Russians' comparatively elevated levels of Mongoloid physical features. But you are ignoring the fact that in general southern Caucasoids possess more distinctively Caucasoid and less Mongoloid facial features than northern Caucasoids (whether from west or east). [/i]

Physical anthropology (and Cavalli-Sforza and ADMIXTURE) all track a distinctive cline of Mongoloid influence peaking in NE Europe, troughing in SW Europe, but already petering out by the time it reaches Germany. Physical anthropology notes Uralic forms (Lappoid, East Baltid etc.) in Eastern Europe, not in NW Europe. And from experience, there are no 'Mongoloid' features to speak of in the British Isles. Certainly not at comparable levels with, say, Poles or Swedes, who occupy a similar position on the chart.

[i]This fact is quite often ignored by Nordicists and the Nordicist physical anthropological establishment. To see Caucasoids par excellence you have to look to south, not to north. So there is nothing in Patterson et al.'s and Dienekes' ADMIXTOOLS tests that contradicts physical anthropological results.[/i]

The difference between NW and Southern Eurasians is the impact of Mesolithic Cro-Magnoids. CMs are proto-West Eurasians, and, I agree, not Caucasoids par excellence. It's possible that their features are, in part, related to the ancient East Eurasian population responsible for the genetic phenomenon in question. However, they're not overty Mongoloid shifted. The anthropology of the British Isles is essentially the product of the two forms discussed: CM and caucasoids par excellence. Claiming a Mongoloid component equal (or surpassing) that of Eastern Europe [i]is[/i] contradictory to physical anthropology.

On a brief side note: In Dienekes' earlier table, the ordering of Ameriasian admix was slightly different:

http://1.bp.blogspot.com/-mGxYfQ2nxPs/UDRHAYItFjI/AAAAAAAAFpA/3jzvCNG1m10/s1600/amerasian.png

Matt, thanks for the explanation. I'd considered that possibility. This is where my ignorance of how the various software work comes to bite me, since I don't know which programmes (if any) are measuring direct admixture or simply drift. I'd previously considered that the 'Southern' component (and the SW Asian in some Dodecad runs) experienced drift versus all other populations compared to Atlantic Baltics and Gedrosia and Caucasus populations, and that this is responsible for part of the observed effect. I also considered that populations that appeared off-cline in the Karitiana-affinity phenomenon did so because of SSA admix, and that all West Eurasians would occupy a osition on the cline if they lacked SSA admixture (note that the least SSA admixed Southern Europeans, such as Basques, North Italians, and Tuscans are part of the cline).

Onur said...

One thing to bear in mind is that the population mixed with has relatively more Native Americans and Siberian affinities than East Asian affinities (the signal of admixture is stronger with them than East Asians and South East Asians)

As you may know, East Asians and Southeast Asians (including the island Southeast Asians and Oceanian Mongoloids) are not the only Mongoloids, as Siberians and Native Americans too are certainly part of the Mongoloid race. So when I say "Mongoloid", I am always including Siberians and Native Americans in Mongoloids.

but we don't know quite what they may have looked like.

Physical anthropology may give some clues about that. I already mentioned the in general elevated Mongoloid physical affinity of current northern Caucasoids (whether from west or east) compared to that of current southern Caucasoids. A thorough craniometric study of various Caucasoid groups and the non-Caucasoid or racially hybrid groups they interacted with from various times and places would be informative to see how the various Caucasoid groups took their current physical form in time.

mikej2 said...

Correction to my first message concerning Saamis

"The second arrival is estimated to 2500 BC"

should be

"The second arrival is estimated to 2500 BP"

Fanty said...

I refered to what was talked about in the first 2 posts, rather than in the article.

My claim was that Dienekes muses about these things.

And you claim I am wrong and that he doesnt.

Read his post (second post) ;-)

I am right, not wrong about this. :-P

Dienekes said...

Re: Saami theory could be tested depending on admixture age estimates in Vologda Russians and Mordvins, since these are unlikely as far as I can tell tell to have Saami ancestry.

hamarfox said...

Physical anthropology may give some clues about that. I already mentioned the in general elevated Mongoloid physical affinity of current northern Caucasoids (whether from west or east) compared to that of current southern Caucasoids.

You're not taking into the massive clines of Mongoloid presence within Northern Europe. I'm sure you know that there are essentially no Mongoliform features in the following: Britain and Ireland, the low countries, France. You also know that they're considerably weak in Germany. This is against populations such as NE and Eastern Europeans whose genetic pools frequently produce people who look like this Pole:

http://i674.photobucket.com/albums/vv103/camelsloop/7521_1218250307292_1559932989_568381_7697792_n.jpg

Can you honestly find me anyone from the above populations who resembles her?

Indeed, it's the absence of Mongoloid characteristics in NW Europeans, despite their comparable levels of Karitiana-admixture, that suggests precisely that the admixture isn't Mongoloid.

Mongoloid traits in Europe broadly follow haplogroup N distribution.

mikej2 said...

Dienekes

"Re: Saami theory could be tested depending on admixture age estimates in Vologda Russians and Mordvins, since these are unlikely as far as I can tell tell to have Saami ancestry."

Just notice what I wrote; the people who brought Saami language were Uralic people who came to the Ladoga region 2500BP. The Saami history in Finland is quite well known and their movement to the north happened during the last 2000 years. The Saami people living today mostly in Norway are a mix of these "newcomers" and old Mesolithic people. So it is likely that same Uralic admix is present among Vologdas and Mordvins, but the Mesolithic part of Saami ancestry is not present there. Saami toponymes has been found also in Northern Russia, which supports the original linguistic theory.

So we are back in the starting point; Finns are what they are. They have admix of Saamis including both Mesolithic and Uralic-Saamic ancestry, plus the older Uralic admix from the southern seaside of Gulf of Finland. And of course later Scandinavian admix.

truth said...

You don't necessarily need to show mongoloid features to have mongoloid admixture, see the case of the Ainu people in Japan, same with amerindian-like admixture, some North Amerindian tribes don't show these traits.

hamarfox said...

You don't necessarily need to show mongoloid features to have mongoloid admixture, see the case of the Ainu people in Japan, same with amerindian-like admixture, some North Amerindian tribes don't show these traits.

Most Ainu show Mongoloid features. I'm not sure whether you meant, 'You don't necessarily need to have mongoloid features to have East Eurasian admixture'. I agree with that. But having 15% 'mongoloid' admixture in a population, yet not having mongoloid features manifest in that population with any regularity, especially when other populations with comparable 'mongoloid' admixture do manifest mongoloid features, is rather nonsensical.

Rather, it's more parsimonious to suggest an original, deep Asian-shift, over which selection and such had worked its magic in its own (non-Mongoloid) direction, which was originally ubiquitous among N. Europeans, more or less, and then a more recent infiltration of Y haplogroup N carriers that brought the Mongoloid phenotype unevenly to parts of Northern and Eastern Europe.

This latter wave is what ADMIXTURE detects, and also what physical anthropologists and even casual observers also detect.

Onur said...

The difference between NW and Southern Eurasians is the impact of Mesolithic Cro-Magnoids. CMs are proto-West Eurasians, and, I agree, not Caucasoids par excellence.

The physical type of that Caucasoid element that is not Caucasoids par excellence is closer to that of Mongoloids than the physical type of the Caucasoid par excellence element is in a continuum of physical types from the most Caucasoid to the most Mongoloid. Whether they are Proto-Caucasoids or later steppe types with Mongoloid admixture or a mix of both is open to debate, as Dienekes points out. Dienekes favors the later steppe tpes with Mongoloid admixture scenario more due to the fact that Y-haplogroup R, which is today (but apparently not before the Copper Age) the main Y-haplogroup of a large portion of Europe, especially northern Europe, is a sister clade of Q, the main Y-haplogroup of Native Americans, and a first degree cousin clade of O and N, the main Y-haplogroups of East and Southeast Asia and a large portion of Siberia respectively.

Claiming a Mongoloid component equal (or surpassing) that of Eastern Europe [i]is[/i] contradictory to physical anthropology.

I have never claimed that, even exluding Russians and Uralic and Turkic speakers of NE Europe. In Dienekes' F4 ratio test, Scandinavians and non-Russian northern Slavs (Ukrainians, Poles, etc.) show up more East Eurasian shifted than NW Europeans (Baltic peoples are not included in that test, but they too would surely show up more East Eurasian shifted than NW Europeans, as is clear from Dienekes' other test you linked to).

On a brief side note: In Dienekes' earlier table, the ordering of Ameriasian admix was slightly different:

http://1.bp.blogspot.com/-mGxYfQ2nxPs/UDRHAYItFjI/AAAAAAAAFpA/3jzvCNG1m10/s1600/amerasian.png


The only differences in the orderings are in some populations whose East Eurasian shift levels are pretty close. So their orderings are actually pretty similar.

Indeed, it's the absence of Mongoloid characteristics in NW Europeans, despite their comparable levels of Karitiana-admixture, that suggests precisely that the admixture isn't Mongoloid.

We don't know what type of Mongoloids are responsible for the general East Eurasian shift trend in Europe. According to genetics, they seem to be closest to Native Americans among the existing Mongoloid groups. Native Americans are widely thought to be physically less evolved (thus closer to Proto-Mongoloids) forms of the Mongoloid race than current Eurasian Mongoloids, who are the type of Mongoloid that first comes to mind today when talking about Mongoloids in general. These may give some clues about the type of Mongoloids who are responsible for the general East Eurasian shift trend in Europe.

This latter wave is what ADMIXTURE detects, and also what physical anthropologists and even casual observers also detect.

The latter wave is detected by ADMIXTURE because of its relative recency. Nothing to do with physical characteristics.

Onur said...

BTW, part of the Mongoloid admixture in current Europeans may be from as early as the Paleolithic. That may explain the significant East Eurasian shift of the hunter-gatherer ancient European samples sequenced so far.

hamarfox said...

I appreciate the response, Onur.

The physical type of that Caucasoid element that is not Caucasoids par excellence is closer to that of Mongoloids than the physical type of the Caucasoid par excellence element is in a continuum of physical types from the most Caucasoid to the most Mongoloid. Whether they are Proto-Caucasoids or later steppe types with Mongoloid admixture or a mix of both is open to debate, as Dienekes points out.

But the CM type isn't shifted toward Amerindians relative to Caucasoids par excellence. Rather it's the opposite case: prominent noses, longer, thinner faces, higher-cheekbones relative to Cro-Magnoids are shared in common between Amerindians and classic Caucasoids.

Nor are these features inherently Mongoloid or Asiatic at all. They've been variously classified as North African and even sub-Saharan. The point is that calling them 'Mongoloid' is speculative, and not even particularly relevant, given that the connection is to Amerindians (who don't have particularly broad faces, snubbed noses etc. as CMs do).

I have never claimed that, even exluding Russians and Uralic and Turkic speakers of NE Europe. In Dienekes' F4 ratio test, Scandinavians and non-Russian northern Slavs (Ukrainians, Poles, etc.) show up more East Eurasian shifted than NW Europeans..

Dienekes' second test found, however, that the English are more Karitiana than, for example, Hungarians. But regardless of the precise ordering, the salient point is that the contrast between Mongoloid-absent and Mongoloid-present populations is too weak (even virtually nonexistent) to be in harmony with physical anth. We should use this fact to help guide our interpretation of the precise nature of the admixture, and in how many waves it occurred (and over what duration).

Regardless of whether you think ordinary CM forms are Mongoloid influenced, I'm sure you agree that Mongoloid-influenced types manifest differently (and less ambiguously) in Y haplogroup N influenced populations. This snaps back to my original point: we're looking at clear clines within N. Europe, and to equate NW and NE Europe is to eclipse some serious historical events, IMO -- events that heavily impact on physical anthropology.

hamarfox said...

The only differences in the orderings are in some populations whose East Eurasian shift levels are pretty close. So their orderings are actually pretty similar.

See above. Even Dienekes states that the E. Eurasian in W. Europe isn't predominantly Mongoloid. Again, Hungarians being less Mongoloid than Britons is patently absurd and 100% against physical anthropology.

We don't know what type of Mongoloids are responsible for the general East Eurasian shift trend in Europe. According to genetics, they seem to be closest to Native Americans among the existing Mongoloid groups. Native Americans are widely thought to be physically less evolved (thus closer to Proto-Mongoloids) forms of the Mongoloid race than current Eurasian Mongoloids, who are the type of Mongoloid that first comes to mind today when talking about Mongoloids in general. These may give some clues about the type of Mongoloids who are responsible for the general East Eurasian shift trend in Europe.

SSA ancestry seems to have a substantial effect on the figures. I'd like to wait until numbers that take this into account are available.

The latter wave is detected by ADMIXTURE because of its relative recency. Nothing to do with physical characteristics.

The recency has everything to do with physical characteristics. Y haplogroup N, Uralic physical types (look up Lappoid, East Baltic, Ladogan), and ADMIXTURE Siberian correspond almost perfectly. Can you explain what, if not recency, is the reason Bjork (or the girl I posted) could never be born to Dutch or Cornish parents? They're both from populations almost equivalent to the Dutch and Cornish in Karitiana affinity.

Honestly, anthropological revisionism is sadly common in lay circles whenever some unexpected finding occurs. But let's keep a level head: physical Asian-shift in Europe still follows the same patterns everyone observed before Patterson's paper was ever released.

Onur said...

But the CM type isn't shifted toward Amerindians relative to Caucasoids par excellence. Rather it's the opposite case: prominent noses, longer, thinner faces, higher-cheekbones relative to Cro-Magnoids are shared in common between Amerindians and classic Caucasoids.

Nor are these features inherently Mongoloid or Asiatic at all. They've been variously classified as North African and even sub-Saharan. The point is that calling them 'Mongoloid' is speculative, and not even particularly relevant, given that the connection is to Amerindians (who don't have particularly broad faces, snubbed noses etc. as CMs do).


You are focusing on some relative differences and mainly on intra-Caucasoid and intra-Mongoloid differences rather than the differences between Caucasoids and Mongoloids and ignoring the big picture. Cro-Magnons group with all other Caucasoids in craniometric analyses, while Native Americans group with other Mongoloids in detailed craniometric analyses.

Dienekes' second test found, however, that the English are more Karitiana than, for example, Hungarians. But regardless of the precise ordering, the salient point is that the contrast between Mongoloid-absent and Mongoloid-present populations is too weak (even virtually nonexistent) to be in harmony with physical anth. We should use this fact to help guide our interpretation of the precise nature of the admixture, and in how many waves it occurred (and over what duration).

Regardless of whether you think ordinary CM forms are Mongoloid influenced, I'm sure you agree that Mongoloid-influenced types manifest differently (and less ambiguously) in Y haplogroup N influenced populations. This snaps back to my original point: we're looking at clear clines within N. Europe, and to equate NW and NE Europe is to eclipse some serious historical events, IMO -- events that heavily impact on physical anthropology.


As I pointed out, not all Mongoloid groups are physically particularly close to each other. According to craniometric analyses, there is more physical variation between Mongoloid types than between Caucasoid types. This may explain the different effects of different Mongoloid types on the physical characteristics of Caucasoid groups in the cases of similar levels of total Mongoloid admixture.

See above. Even Dienekes states that the E. Eurasian in W. Europe isn't predominantly Mongoloid.

No, Dienekes is ambiguous on that point. He seems to largely avoid that subject, at least for now.

Again, Hungarians being less Mongoloid than Britons is patently absurd and 100% against physical anthropology.

I don't see any noticeable difference between the average Briton and the average Hungarian on the level of their Mongoloid traits. It seems to me that only certain sub-groups of Hungarians possess more Mongoloid traits on average than Britons. But they seem to be in the minority in the total Hungarian population.

Onur said...

SSA ancestry seems to have a substantial effect on the figures. I'd like to wait until numbers that take this into account are available.

Dienekes has already done that:

http://dienekes.blogspot.com/2012/08/scrubbing-sardinians.html

"Another way to demonstrate that east Eurasian-like admixture in CEU is inflating the perceived level of African-like admixture in Sardinians is to comprehensively "scrub" Sardinians of all traces of African ancestry by replacing segments of their DNA when there is even a hint of such ancestry with missing values."

"With regular Sardinians:

f4(San,Papuan;Sardinian,Karitiana) = 0.0084678 (Z=21.2137)
f4(San,Papuan;Sardinian,CEU) = 0.00118099 (Z=10.6838)

So, CEU appears = 0.00118099/0.0084678 = 13.9% East Eurasian"

"With scrubbed Sardinians:

San,Papuan;Sardinian_scrubbed,Karitiana 0.00774427 0.00056725 13.6523
San,Papuan;Sardinian_scrubbed,CEU 0.000678108 0.000167341 4.05225

So, CEU appears = 0.000678108/0.00774427 = 8.8% East Eurasian"

Note that this is an extreme scrubbing, as Dienekes states:

"Using the byseg mode of DIYDodecad, I created ancestry maps of the 28 HGDP Sardinians, and I only kept windows where the African admixture was exactly 0%. This is a very aggressive scrubbing, designed to remove virtually all African admixture from the population. For example, if a window has 99.9% West Eurasian admixture and 0.01% African, I will nonetheless remove it, even though chances are extremely high that the 0.01% represents only noise. I did not want to leave any doubt that any trace of identifiable African ancestry remained in my "scrubbed Sardinians"."

So, CEU must have more East Eurasian shift than 8.8% in reality.

The recency has everything to do with physical characteristics. Y haplogroup N, Uralic physical types (look up Lappoid, East Baltic, Ladogan), and ADMIXTURE Siberian correspond almost perfectly. Can you explain what, if not recency, is the reason Bjork (or the girl I posted) could never be born to Dutch or Cornish parents? They're both from populations almost equivalent to the Dutch and Cornish in Karitiana affinity.

Honestly, anthropological revisionism is sadly common in lay circles whenever some unexpected finding occurs. But let's keep a level head: physical Asian-shift in Europe still follows the same patterns everyone observed before Patterson's paper was ever released.


The key is not the existence of Bjork-like people but their proportion. Individual examples do not tell us anything on that issue. I am sure I would find Bjork-like Britons or Irish too if I search enough (I already know some British and Irish people with less salient Mongoloid traits), but that would say nothing about their proportion.

hamarfox said...

You are focusing on some relative differences and mainly on intra-Caucasoid and intra-Mongoloid differences rather than the differences between Caucasoids and Mongoloids and ignoring the big picture. Cro-Magnons group with all other Caucasoids in craniometric analyses, while Native Americans group with other Mongoloids in detailed craniometric analyses.

Well, back to the original point you made: that CM deviates in the direction of Mongoloids compared to classic Caucasoids, and that craniometric comparison should reveal the degree of deviation (and hence degree of admixture). The type also deviates in the direction of Negroids and Dravidians versus classic Caucasoids, and moreso than it does Mongoloids. Calling the type Mongoloid, as you do, is as (or, actually, more) presumptive as calling it Negroid or Dravidian.

As I pointed out, not all Mongoloid groups are physically particularly close to each other. According to craniometric analyses, there is more physical variation between Mongoloid types than between Caucasoid types. This may explain the different effects of different Mongoloid types on the physical characteristics of Caucasoid groups in the cases of similar levels of total Mongoloid admixture.

Maybe. But, again, it's pure presumption. Instead of making the same points again, I'll introduce a fresh one. Check out this spreadsheet (counterpart to the graph you showed me):

https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedERDdFNVVjY1NVMzdE9wTlJ6QlZKSHc#gid=2

How Mongoloid are Indians? Barely any percentage. Just looking at the ordering of Burusho and Pathan vs Chuvash, for example, suggests to me that this is no clear-cut process. Indians are highly East Eurasian, but not very Mongoloid. But they attain a high percentage of Karitiana, even exceeding their E. Eurasian component.

The point is, we need to start thinking outside of the box to explain these relations, and labelling it all 'Mongoloid' is as within-the-box as can be.

I don't see any noticeable difference between the average Briton and the average Hungarian on the level of their Mongoloid traits. It seems to me that only certain sub-groups of Hungarians possess more Mongoloid traits on average than Britons. But they seem to be in the minority in the total Hungarian population.

This is of course consistent with a higher average in Hungarians. Let's say features manifest at X, and Hungarians on average are X minus 1. Britons are X minus 2. Obviously, deviation from Hungary's average is more likely to cross the boundary.

Dienekes has already done that

I was talking about a statistic for all populations with SSA taken into account. The figures seem very sensitive to SSA ancestry. Observe the CEU's position versus the Britons it usually clusters with. CEU is below all NW European populations features. Why? Presumably a 0.2% presence of SSA. And, obviously, the massive negative scores for ~10% SSA-admixed populations (that at K3 also have E. Eurasian admixture too) are worth consideration. Also compare the SSA-admixed Makrani with non-SSA-admixed sister populations.

hamarfox said...

Finally, consider the African shift of populations lacking any SSA-ancestry (including Hungarians):

https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedC1XUjJMNnFBRnRvLXhpUFdRSWp0bGc&authkey=CJ7C6ugB&hl=en&authkey=CJ7C6ugB#gid=0

This is possibly related to E31b. But regardless, it should be considered.

The key is not the existence of Bjork-like people but their proportion. Individual examples do not tell us anything on that issue. I am sure I would find Bjork-like Britons or Irish too if I search enough (I already know some British and Irish people with less salient Mongoloid traits), but that would say nothing about their proportion.

Well, as I said before, the higher the baseline, the more salient the extremes. Any population is going to have oddities, some of which appear to shoot in a certain direction, whether or not the population has admixture. Milder Bjork-like features are present in a fairly high percentage of Icelanders. None of these features are a staple of British physical characteristics, which is why you don't find Bjork-oids in Britain.

It's also interesting to mention that the most Mongoloid looking 'Briton' I've ever seen (Michael McIntyre) is 50% Hungarian.

You seem like an intelligent person, and I'm sure that if I could communicate my experiences to you, you'd agree with me. But as it goes, you'll just have to take my word for it that every time I see a Mongoloid influenced European (which is quite striking in a crowd of Britons), they always turn out to be E. European immigrants, and not even necessarily ones from very far East either.

hamarfox said...

Sorry, I've no idea where I got E31b from. I meant E1b1b.

Onur said...

Well, back to the original point you made: that CM deviates in the direction of Mongoloids compared to classic Caucasoids, and that craniometric comparison should reveal the degree of deviation (and hence degree of admixture). The type also deviates in the direction of Negroids and Dravidians versus classic Caucasoids, and moreso than it does Mongoloids. Calling the type Mongoloid, as you do, is as (or, actually, more) presumptive as calling it Negroid or Dravidian.

According to the craniometric analyses I have seen, Cro-Magnons deviate from modern Europeans more in the Mongoloid direction than in the Negroid direction. It is not even clear whether they craniometrically deviate in the Negroid direction. These are consistent with the genetic results of the European hunter-gatherer ancient samples sequenced so far, as they genetically deviate much more in the Mongoloid direction than in the Negroid direction. See for instance the PCA analysis result of the Mesolithic Iberians from La Braña:

http://4.bp.blogspot.com/-GPh5_xses4o/T-yfvfiTB4I/AAAAAAAAE7A/BOTvHVxxkPg/s1600/brana_global.png

Notice that geographically speaking, these individuals are as far from the Mongoloid region as can be while a late Upper Paleolithic connection with Africa is quite possible. Despite that, they deviate much more in the Mongoloid direction than in the Negroid direction. I wonder how geographically more eastern Mesolithic Europeans would show up on that plot.

Maybe. But, again, it's pure presumption. Instead of making the same points again, I'll introduce a fresh one. Check out this spreadsheet (counterpart to the graph you showed me):

https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedERDdFNVVjY1NVMzdE9wTlJ6QlZKSHc#gid=2

How Mongoloid are Indians? Barely any percentage. Just looking at the ordering of Burusho and Pathan vs Chuvash, for example, suggests to me that this is no clear-cut process. Indians are highly East Eurasian, but not very Mongoloid. But they attain a high percentage of Karitiana, even exceeding their E. Eurasian component.

The point is, we need to start thinking outside of the box to explain these relations, and labelling it all 'Mongoloid' is as within-the-box as can be.


European populations have no or almost no ASI admixture. So we are left with Mongoloids or an extinct race X to explain their East Eurasian shift. Physical anthropological results and the principle of parsimony seem to favor Mongoloids (including their possible extinct varieties) as the explanation.

I was talking about a statistic for all populations with SSA taken into account. The figures seem very sensitive to SSA ancestry. Observe the CEU's position versus the Britons it usually clusters with. CEU is below all NW European populations features. Why? Presumably a 0.2% presence of SSA. And, obviously, the massive negative scores for ~10% SSA-admixed populations (that at K3 also have E. Eurasian admixture too) are worth consideration. Also compare the SSA-admixed Makrani with non-SSA-admixed sister populations.

CEU is a fairly good representative of NW Europeans (your area of interest), so my example was good enough.

Onur said...

Finally, consider the African shift of populations lacking any SSA-ancestry (including Hungarians):

https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedC1XUjJMNnFBRnRvLXhpUFdRSWp0bGc&authkey=CJ7C6ugB&hl=en&authkey=CJ7C6ugB#gid=0

This is possibly related to E31b. But regardless, it should be considered.


As in ADMIXTURE analyses, differences between admixture times are distorting that PCA analysis.

Well, as I said before, the higher the baseline, the more salient the extremes. Any population is going to have oddities, some of which appear to shoot in a certain direction, whether or not the population has admixture. Milder Bjork-like features are present in a fairly high percentage of Icelanders. None of these features are a staple of British physical characteristics, which is why you don't find Bjork-oids in Britain.

It's also interesting to mention that the most Mongoloid looking 'Briton' I've ever seen (Michael McIntyre) is 50% Hungarian.

You seem like an intelligent person, and I'm sure that if I could communicate my experiences to you, you'd agree with me. But as it goes, you'll just have to take my word for it that every time I see a Mongoloid influenced European (which is quite striking in a crowd of Britons), they always turn out to be E. European immigrants, and not even necessarily ones from very far East either.


I think genetic analyses should be reinforced with craniometric analyses. By comparing their results we can arrive at more robust conclusions about the racial history of Caucasoids and Homo sapiens in general.

hamarfox said...

According to the craniometric analyses I have seen, Cro-Magnons deviate from modern Europeans more in the Mongoloid direction than in the Negroid direction. It is not even clear whether they craniometrically deviate in the Negroid direction.

As you say, 'Mongoloids' contain a great degree of internal diversity. But CMs clearly lack the full complement of Mongoloid features (however broadly defined), and at least on a superficial level, they're more consistent with SSA, since, unlike with the Mongoloid supposition, fewer key Negroid features are salient in their absence in CMs.

Maybe these features were yet supplied by an extinct E. Eurasian people, but these were a people who lacked the distinct bone structure and eyeforms of classic Mongoloids (assuming that their features are fully represented in CMs), and so shouldn't be grouped with them at this point in our understanding.

These are consistent with the genetic results of the European hunter-gatherer ancient samples sequenced so far, as they genetically deviate much more in the Mongoloid direction than in the Negroid direction. See for instance the PCA analysis result of the Mesolithic Iberians from La Braña

This is the question that's most interesting to me. These Asian-shifted folk mixed later with haplo R Amerindian-connected folk, and the result was descendants significantly shifted away from East Eurasians relative to them. This is precisely the reason I'd refrain from any discrete conclusion on the matter. It's an open mystery.

European populations have no or almost no ASI admixture. So we are left with Mongoloids or an extinct race X to explain their East Eurasian shift. Physical anthropological results and the principle of parsimony seem to favor Mongoloids (including their possible extinct varieties) as the explanation.

Western Europeans have connections with the Pathan people. Only about half of the South Asian component is ASI, so if you halve the amount of South Asian in Pathan at K7 and add the small amounts of Siberian (they have no East Asian), then I generously round it to 20%. But they're 35% Karitiana. This suggests that at least some of the 'Karitiana' in Western Europeans is of the same (unknown) nature as is found in the Pathan.

As in ADMIXTURE analyses, differences between admixture times are distorting that PCA analysis.

Ah, I know little about how the PCA analysis was performed. Are you saying these genetic distances aren't absolute, or simply that they mask the nature (and time) of the admixture that might cause the relative positions? (A genuine question.)

Onur said...

This is the question that's most interesting to me. These Asian-shifted folk mixed later with haplo R Amerindian-connected folk, and the result was descendants significantly shifted away from East Eurasians relative to them. This is precisely the reason I'd refrain from any discrete conclusion on the matter. It's an open mystery.

You are ignoring the large-scale migration of the Sardinian-like Neolithic folk to all over Europe between the Mesolithic era and times of the migration of the Y-haplogroup R bearing folk. They apparently had a huge impact on the overall genetic landscape of Europe. Also, you are ignoring that the Y-haplogroup R bearing folk significantly admixed with the West Eurasian Neolithic folk, as they largely took their wives from the locals during their migrations, thus significantly diluted their own original autosomal genetics (more in the south than in the north).

Western Europeans have connections with the Pathan people. Only about half of the South Asian component is ASI, so if you halve the amount of South Asian in Pathan at K7 and add the small amounts of Siberian (they have no East Asian), then I generously round it to 20%. But they're 35% Karitiana. This suggests that at least some of the 'Karitiana' in Western Europeans is of the same (unknown) nature as is found in the Pathan.

No, ASI is the major element in the makeup of the "South Asian" component, its ANI or Caucasoid part as a whole being the minor element. Anyway, if you are going to calculate the overall Mongoloid or overall ASI ancestry of a population, you should stop using results of ADMIXTURE, which has problems in detecting ancient racial admixtures and their effects (see below), as proxy, but should instead use direct measurements such as the ones performed by softwares like ADMIXTOOLS. Pathans possess about 20% ASI admixture and obviously possess more Mongoloid admixture than detected by ADMIXTURE. As I pointed out previously, the East Eurasian shift in South Asians and South Asian-admixed populations in Central and West Asia certainly only partially Mongoloid-related, the rest being ASI related. ASI, because of being genetically closer to Mongoloids than to Caucasoids, have a positive effect on the East Eurasian shift of populations. That is why Indians have such a huge East Eurasian shift. But we know from direct measurements that Europeans possess no or almost no ASI admixture but carry significant levels of ancient Mongoloid and/or an extinct race X admixture, that significantly decreases from the north to the south and less significantly decreases from the east to the west.

Ah, I know little about how the PCA analysis was performed. Are you saying these genetic distances aren't absolute, or simply that they mask the nature (and time) of the admixture that might cause the relative positions? (A genuine question.)

Non-formal admixture tests such as ADMIXTURE, STRUCTURE, PCA and MDS are prone to the distorting effects of linkage disequilibrium decay and drift and are not good at detecting ancient racial admixtures and their effects. The Pan-European East Eurasian shift is so ancient that its effects are only partially detected, if at all, by non-formal admixture tests. That is why you should use direct admixture measurements such as the ones performed by softwares like ADMIXTOOLS and TreeMix, which are much better at detecting ancient racial admixtures and their effects.