tag:blogger.com,1999:blog-7785493.post8899454533420865992..comments2024-01-04T04:11:55.717+02:00Comments on Dienekes’ Anthropology Blog: The Indo-European invasion of the BalticDienekeshttp://www.blogger.com/profile/02082684850093948970noreply@blogger.comBlogger19125tag:blogger.com,1999:blog-7785493.post-58824789966670271662012-10-25T02:00:38.232+03:002012-10-25T02:00:38.232+03:00Actually it's striking how the rolloff date fo...Actually it's striking how the rolloff date for Poland, 2320 BC, fits well with the beginning of the bronze age in southern Poland, with the Unetice and Miersanovice cultures! <br /><br />So if the expansion of IE was linked to the expansion of the West_Asian component, and I deem this idea to be quite convincing, then it would follow that, surprisingly, the Corded Ware wasn't yet IE.<br /><br />There are two advantages in this insight: <br />1. There would no longer be any need to derive the Corded Ware IEs from the earlier West_Asianised Pit-Grave/Yamnaya culture.<br /><br />2. The Corded Ware/Battle axe/Single Grave culture followed immediately on the genetically rather Southern TRB (with Gok4) and was apparently associated with R1a, but probably also with I1 in the north. It may be one of the first agents to diffuse the Amerindian component more widely in Europe. And since this component is also found in Basques, the beginning of its wider diffusion must have antedated the arrival of IEs in central Europe. A pre-IE Corded Ware would suit this well.SimonWhttps://www.blogger.com/profile/05915003431050999048noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-53676264078738112922012-10-15T13:57:30.512+03:002012-10-15T13:57:30.512+03:00Eastern Slavs expanded into the eastern European p...<i><br />Eastern Slavs expanded into the eastern European plain in medieval times from Central Europe<br /></i><br /><br />Before the genetics, there were also several other theories, including the one of Slavs coming from the eastern steppes, or living in somewhere in Ukraine between steppes and forests. <br /><br />The idea that Slavs came from Central Europe (e.g. somewhere around Poland) was called in those days "autochtonous theory" and was pretty much considered not just unlikely, but even ridiculous theory.<br /><br />PS: got "internal error" while posting the comment. Hope i will not sent the same comment twice...szopenhttps://www.blogger.com/profile/04219188379320432806noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-72499617812423382112012-10-14T12:10:49.304+03:002012-10-14T12:10:49.304+03:00The Scytians lived in Ukraine for a considerable t...<i>The Scytians lived in Ukraine for a considerable time, also presence of Scytian connected groups (Sarmatians, possibly Alans) outlived them. Should not an Ukrainian (Lithuanian vs. Baluchi) Roloff pick-up a later signal because of this? BMAC is rather close to Baluchi. <br /></i><br /><br />Eastern Slavs expanded into the eastern European plain in medieval times from Central Europe. I see no reason why they would be particularly related to the Scythians who lived there a thousand years earlier.Dienekeshttps://www.blogger.com/profile/02082684850093948970noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-35720830832228955432012-10-14T09:14:27.365+03:002012-10-14T09:14:27.365+03:00In earlier debates you questioned that there were ...In earlier debates you questioned that there were IE on the steppe other than Indo-Iranians, so this is pretty much the same. But indeed the question was a bit too general and undefined. I reform it.<br /><br />The Scytians lived in Ukraine for a considerable time, also presence of Scytian connected groups (Sarmatians, possibly Alans) outlived them. Should not an Ukrainian (Lithuanian vs. Baluchi) Roloff pick-up a later signal because of this? BMAC is rather close to Baluchi. <br /><br />I can imagine multiple answers why it should not pick it up (Scytian genetic impact is insignificant in current Ukrainians, BMAC was too different from Baluchi, Scytians are not from the BMAC, ect.), but I think this result at least raise the question.Slumberyhttps://www.blogger.com/profile/05139930329199925111noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-50860908889754693002012-10-13T23:37:10.211+03:002012-10-13T23:37:10.211+03:00Don't you think that this result question the ...<i>Don't you think that this result question the theory that source of the steppe IE is the BMAC?</i><br /><br />I am not sure what you are referring to. I think the BMAC was the source of Indo-Iranian, and hence (ultimately) of the Scythians who migrated to Europe during the Iron Age. The language of previous denizens of the European steppe is unknown, and may have included some Indo-Europeanized groups.Dienekeshttps://www.blogger.com/profile/02082684850093948970noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-22033407223156533862012-10-13T20:10:46.504+03:002012-10-13T20:10:46.504+03:00Dienekes
Don't you think that this result que...Dienekes<br /><br />Don't you think that this result question the theory that source of the steppe IE is the BMAC? <br />Ukraine was on the main road of steppe migration after this admixture until the Medieval times. If the Eastern Steppe got a significant genetic impact from that South later, a Baluchi referenced roloff would pick that up.Slumberyhttps://www.blogger.com/profile/05139930329199925111noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-85658789808260586472012-10-13T09:33:15.402+03:002012-10-13T09:33:15.402+03:00And, @ Dienekes: The point I was trying to make do...And, @ Dienekes: The point I was trying to make doesn't hinge on the reality of the near absence of the Southern component in Balts, but rather on the relation West_Asian : Southern. If the West_Asian IEs took the route via Anatolia and/or the Balkans to the Baltic, then they must have largely avoided the locals on their way, otherwise the relation West_Asian : Southern in Lithuanians would be lower than it is. Of course, that's not impossible - the early farmers behaved exactly this way - but if the IEs instead sprung out of their West Asian homeland directly onto the Pontic plain, there would be no need to assume such avoiding behaviour. Admittedly, not a compelling argument.SimonWhttps://www.blogger.com/profile/05915003431050999048noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-85082031438132170612012-10-13T01:23:22.055+03:002012-10-13T01:23:22.055+03:00Well, true, physically there is quite a difference...Well, true, physically there is quite a difference between corded groups and the pit-grave people... And if the spread of IE languages was accompanied by the spread of the West_Asian component, then there must have been some gene flow. The question is, if this appears possible in the light of these differences. According to some dendrograms I've seen, Central European corded groups were rather similar to each other. The Polish groups just had a somewhat lower FI than the more western groups, though. Also Schwidetzky mentioned that the globular amphora people get broader faced towards the east and at least in this respect do resemble the pit-grave people. The spread of the globular amphora culture preceded and sort of prepared the spread of the corded ware. But I agree, the corded people can't be deduced from the pit-grave people and there is a lot of anthropological continuity with preceding groups.<br /><br />As for the origin of the corded ware culture, I've read that the development and implementation of the typical corded ware burial rite preceded the spread of the corded ware pottery by two centuries and was without a clear point of origin. The corded ware pottery on the other hand clearly spread from Poland to the other locales, and this spread was connected with globular amphorae.<br /><br />@ eurologist<br /><br />Bold assertions! No geographic overlap between the corded ware and Balto-Slavic?? Well the question is where we tentatively localise the proto-Balts and the proto-Slavs, and probably there are different methods to achieve this. At least the evidence from hydronyms and place names shows a lot of overlap between Balto-Slavs and corded ware. Germanic stands somewhat inbetween Balto-Slavic and Italo-Celtic... And a boundary with Uralic doesn't rule out another boundary with Balto-Slavic.SimonWhttps://www.blogger.com/profile/05915003431050999048noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-77843845264584396202012-10-12T12:49:47.327+03:002012-10-12T12:49:47.327+03:00At present, it is difficult to say much about the ...<i>At present, it is difficult to say much about the origins of the Corded Ware culture. But, its anthropological type is certainly related to that of the Neolithic inhabitants, and contrasts with both Mesolithic hunter-gatherers and Kurgan groups, so I would be much surprised if it was formed without input from the TRB population.</i><br /><br />I agree 100%. In most places (but not all) Corded Ware was not intrusive, but rather based on local continuation.<br /><br />I also see no relation between Corded Ware and Slavic or Baltic languages - there really is no geographic overlap, at the time. Conversely, proto-Germanic clearly shared a boundary with Uralic, and not with Slavic languages.eurologisthttps://www.blogger.com/profile/03440019181278830033noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-19095944402890570822012-10-12T11:26:02.633+03:002012-10-12T11:26:02.633+03:00while the K7b Southern component is virtually abse...<i>while the K7b Southern component is virtually absent</i><br /><br />Gok4-related farmer ancestry is estimated at 11-14% in Finns and North Russians by Skoglund et al. (2012). As I have emphasized before, absence of a component in an ADMIXTURE analysis indicates a minimum within a given set of populations, rather than complete absence.<br /><br />At present, it is difficult to say much about the origins of the Corded Ware culture. But, its anthropological type is certainly related to that of the Neolithic inhabitants, and contrasts with both Mesolithic hunter-gatherers and Kurgan groups, so I would be much surprised if it was formed without input from the TRB population.Dienekeshttps://www.blogger.com/profile/02082684850093948970noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-34418245432088792122012-10-11T20:43:43.845+03:002012-10-11T20:43:43.845+03:00Two remarks:
1. That the proto-Balts, and by infe...Two remarks:<br /><br />1. That the proto-Balts, and by inference presumably also the proto-Slavs and maybe the proto-Germanics, are descended from IEs that came from the east (Pontic steppe) rather than from the Balkans is also indicated by the fact that the K7b West_Asian component in Lithuanians is reasonably strong, while the K7b Southern component is virtually absent. This speaks against an IE derivation via the Balkans, at least for the Balto-Slavs, maybe also for the Germanics. <br /><br />2. The lower admixture date in Poles relative to Lithuanians may be due to later additional admixture with Scythians, Sarmatians and the like, which quite possibly affected Slavs more than Balts.SimonWhttps://www.blogger.com/profile/05915003431050999048noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-58291286011226480652012-10-11T03:43:52.696+03:002012-10-11T03:43:52.696+03:00OK I didn't get an answer, and I was also conc...OK I didn't get an answer, and I was also concerned about the markers with a low rate of recurrent mutagenesis (which may put them in correlation despite lack of IBD, artifactually increasing roloff signal). So I reread Moorjani et al., looking how they tested ROLLOFF, and what limitations they found. <br /><br />Firstly, adjacent SNPs pairs turned out to be unreliable: "We do not show inter-SNP intervals of <0.5cM since we have found that at this distance admixture LD begins to be confounded by background LD, and so inferences are not reliable (exponential curve fitting does not include inter-SNP intervals at this scale)."<br /><br />Testing ROLLOFF was conducted on simulated African_European mixed individuals with high (20%) European admixture. As Figure 4 shows, even with this deep admixture, and even under a simulated conditions of exact genetic distances and lack of recurrent mutagenesis, ages of admixture started falling out of confidence intervals after as few as 100 generations. They note that there is also a systematic upward bias in the ages of admixture when the admixture % is low, and/or admixture is old.<br /><br />CIs are calculated using permutations with individual chromosomes dropped, one at a time, so if the estimated age of admission is strongly influenced by something very local such as for example a single inversion event, it might be manifested by a wider CI, and then caught by inspecting the permutations manually...MOCKBAhttps://www.blogger.com/profile/05150628026789690963noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-28053069126312478922012-10-10T17:28:01.097+03:002012-10-10T17:28:01.097+03:00Dienekes, I like the conclusions of rolloff tests ...Dienekes, I like the conclusions of rolloff tests which you publish - all of them - but I feel somewhat uneasy about the boundaries of the confidence intervals. It seems that the decay rates are strongly influenced by the very few "LD survivors" at approximately 1 cM scale? How well have we checked that the genetic distances are measured correctly in those individuals who display the LD? Because individual variations in recombinations in LD rate are bound to happen. Local inversions suppress recombinations, large deletions bring loci closer together, and mutations in natural recombination hotspot may reduce effective genetic distances too. When we are relying on rare survivors of recombination, any such effects which slow down recombination may contribute to systematic error. <br /><br />Also, "plus-minus X generations" leaves an impression that the random error is equally likely to be positive or negative, but I suspect that the actual distribution of probabilities isn't symmetric.<br /><br />Also, translation to "plus-minus years" may need to take into account statistical uncertainty about the effective age of a generation? In other word, the confidence, when expressed in years, should be proprtionally wider than the CI expressed in generations, because of the added uncertainty about the conversion factor.<br /><br />Lastly, even a perfectly correct CI would have one in 20 probability of things happening outside of this interval; repeat the test several times, and you end up with a virtual guarantee that one of your estimates is off...MOCKBAhttps://www.blogger.com/profile/05150628026789690963noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-52606280014934458302012-10-10T15:58:42.725+03:002012-10-10T15:58:42.725+03:00Thanks for the additional explanation. I was havin...Thanks for the additional explanation. I was having trouble following some of the earlier Roll-off posts.<br />autCharles Nydorfhttps://www.blogger.com/profile/16291667302870991631noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-38531134749648296092012-10-10T14:22:37.058+03:002012-10-10T14:22:37.058+03:00http://sphotos-b.ak.fbcdn.net/hphotos-ak-snc6/2619...http://sphotos-b.ak.fbcdn.net/hphotos-ak-snc6/261926_4667528335980_494621359_n.jpgProject "Magnus Ducatus Lituaniae"https://www.blogger.com/profile/06764361071403376842noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-36898704276636533712012-10-10T14:21:46.869+03:002012-10-10T14:21:46.869+03:00Very interesting.
I have obtained the similar res...Very interesting.<br /><br />I have obtained the similar results in my project:<br />http://sphotos-b.ak.fbcdn.net/hphotos-ak-snc6/261926_4667528335980_494621359_n.jpgProject "Magnus Ducatus Lituaniae"https://www.blogger.com/profile/06764361071403376842noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-41364414604209630162012-10-10T13:06:43.054+03:002012-10-10T13:06:43.054+03:00At the other hand if this is not the same migratio...<i>At the other hand if this is not the same migration wave, then how you know that they have anything to do with each other at all?</i><br /><br />The dates have large confidence intervals. For example, the Polish_D rolloff analysis I just added has a date of 4,330 years.<br /><br />We know that they have something with each other because they involve the same reference populations (=South Asians) + populations likely to carry the greatest influence from substrata in southern (=Sardinians) or northern (=Lithuanians) Europe.<br /><br />But, certainly, they were probably not the "same" event; it would be more accurate to speak of a set of invasions that took place in Europe after its initial Neolithic settlement, much like the Americas were settled over centuries by different groups of Europeans who originated in different sub-parts of Europe and settled different sub-parts of the Americas.Dienekeshttps://www.blogger.com/profile/02082684850093948970noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-82800753383554809542012-10-10T09:20:32.554+03:002012-10-10T09:20:32.554+03:00This would mean that the Baltic region was reached...This would mean that the Baltic region was reached by IE much earlier than West/South Europe. The difference from the other admixture dates is 2 millenia, a way too much if you try to identify this as essentially the same migration wave. At the other hand if this is not the same migration wave, then how you know that they have anything to do with each other at all?<br /> <br />Also 5800 BP is not really Bronze Age. Do we have any bronze material from East Europe around this time? <br /><br />Another question: how you know that this particular admixture detected in the Lithuanians is IE? It is a plausible assumption that multiple groups migrated. Given how far the Baltic region from the better studied "main road", this can be easily a signal of some never before identified migration. (For example a Central Asian population that also contributed in modern South Asians.)<br /><br />I think you jumped into too much and too detailed conclusions. Slumberyhttps://www.blogger.com/profile/05139930329199925111noreply@blogger.comtag:blogger.com,1999:blog-7785493.post-8516747773894270802012-10-10T03:47:02.744+03:002012-10-10T03:47:02.744+03:00This looks neat, but I don't understand it.
D...This looks neat, but I don't understand it.<br /><br />Does the confidence interval (+/- number of generations) decrease by the square root of the sample size (both individuals sampled and number of SNPs)?Jhttps://www.blogger.com/profile/07501457160463457211noreply@blogger.com