Ancient Indo-European Folktales?

Trond Engen sent me links to “Ancient Roots of Indo-European Folktales,” by Sara Graça da Silva and Jamshid J. Tehrani (“In this study, we introduce new methods for tackling these problems by applying comparative phylogenetic methods and autologistic modelling to analyse the relationships between folktales, population histories and geographical distances in Indo-European-speaking societies”), and the (inevitably breathless) BBC News story about it, “Fairy tale origins thousands of years old, researchers say” (“Analysis showed Beauty And The Beast and Rumpelstiltskin to be about 4,000 years old”). My immediate reaction is skepticism, and Trond says:

Maybe I don’t understand their probabilities, but it seems like they use the statistical procedure to seek out the outliers and then say “I’m almost certain that these are cherries!” I mean, 1/3 of almost 300 adventures showing more vertical affinity than expected by chance. Really? Gradually fewer of those 100 adventures being reconstructible (with a low bar) for each node up, resulting in only one out of 100 being (weakly) assigned to the top node. What do they expect from a scattershot? And no attempt to control with e.g. Finnish or Hungarian.

But both of us are eager to hear what Hatters have to say about it.

Update. Mark Liberman has posted about this at the Log.

Further update.
Julien d’Huy has written me to say he was one of the first to use modern phylogenetic tools to study mythology and folktales; you can see samples of his work here, and compare the approach of Jean-Loïc Le Quellec here — he shows the existence of one folktale during the Palaeolithic period in Africa.

Comments

  1. I don’t think their basic assumption is valid, because folktales cross linguistic barriers easily.

    There can’t be any Indo-European folk tales, but only Eurasian ones.

    Some classical folktales have even greater distribution, including all of Eurasia and large parts of Africa.

    Cinderella (ATU510A in Aarne Thompson Uther (ATU) Index) is case in point.

    This tale well is known folklore of peoples living in Western and Eastern Europe, Middle East, Iran, Central Asia, Tibet, Siberia, Japan, China, South Asia, Indo-China, Western Africa, Eastern Africa.

    4000 years? Indo-European tale?

    No way. Cinderella is so widespread that it’s ought to be dated to Upper Paleolithic, at least.

  2. This tale is well known in folklore…

  3. @SFReader: So they do attempt to control for this. They’re using two variables — geographic location and common linguistic descent — and trying to figure out which of them explains more of the variations in the fairy tale plots.

    I’m still skeptical, but at least this basic objection is covered.

  4. I picked it from a different site earlier this morning too. Yes, the model with linguistic similarities worked better than the one with geographic distances, but we are talking cultural affinities and influences, religion, trade, rather than mere diffusion by distance?
    3 of the stories came particularly old-rooted, and two of them have also very old literary roots: Beauty and the Beast, and Rumpelstiltskin. It made me wonder if copycat children book writers, inspired by the Grimm brothers of Germany but introducing variations here and there, may have contributed to the present-day variability.
    The absolute champ (and the only one significantly predating the splits of the subfamilies) is (an unknown to me personally) story about a Smith and a Bad Magic Guy who make a deal, upon which the Smith obtains great powers over metals and materials, which he then uses to break the contract by chaining the Bad Guy to a tree or a rock with an unbreakable chain. One needs to mention (following Trond) that the statistical significance wasn’t properly corrected for multi-testing so the statistical significance is inflated. But the fairy-tale kind of a significance is definitely there, given a story thread of an ur-people which overwhelmed its neighbors after developing metallurgy….

  5. J. W. Brewer says

    One pretty obvious question is what level of generality do you use to decide if two tales from different cultures/languages are variants of the same ur-tale. (At a sufficiently high level of generality you can usually take a random ’70’s sitcom rerun and find a foreshadowing of the plot in some comedy by Plautus.) They seem to be using this ATU collection, which no doubt has a methodology (with no doubt both defenders and detractors in some relevant specialist literature). So on the one hand it’s good (and also convenient) that they’re using an existing dataset, with whatever pluses and minuses it has, rather than coming up with their own dataset that might be biased toward supporting their own hypothesis. But on the other hand, if the methodology used for the ATU dataset was done in a different context for a different purpose, how do we know if it’s the right sort of data for this exercise? It strikes me that in other folklore-corpus contexts, like the Child ballads in English, you clearly have both situations where text A and text B are plausibly historically related as actual texts even though there may be substantial differences in wording and then you have other situations where text C and text D appear, as texts, to be entirely independent compositions but nonetheless have some similarity of theme, setting, and plot. Using genetic metaphors may work pretty well for the first sort of relationship; for the second sort of relationship it’s, shall we say, a looser metaphor.

  6. Most of the tales Polish children hear come from the Grimms (and many of those have already filtered down into rural folklore). Of course some of the the recurrent motifs may be many millennia old (and found also in genuinely Slavic tales), but with so much horizontal transfer I’d be surprised if population histories and linguistic phylogenies were sufficiently correlated with folktale genealogies to be of much use. Even IE religious and heroic myths don’t tell you much about the history of the languages in which they were told. Yes, there was an important sky-god called *Diēus ph₂tēr, and probably a dawn-goddess called *H₂ausōs, and Bronze Age heroes valued *ń̥dʰgʷʰitom ḱléwos more than anything else, and in order to win it they apparently slew snake-like monsters. But this knowledge doesn’t help us to decide if the Italo-Celtic hypothesis is true, or whether Albanian is genetically closer to Greek, or to Indo-Iranian, or is in an outgroup relation to both.

  7. @Dmitry,

    One of my great-grandpas was a village blacksmith. His fame and local stories about his real or imaginary deeds circulated long after his death in the whole parish, especially because he was also a cheery sort of chap and a prankster. Some occupations just attract people’s attention in this way.

  8. “… and the (inevitably breathless) BBC News story about it, …”

    Oh, come on,. Hat. It is a *news* story, not a report to a conference (or blog) of linguistics experts, but to the great general public, based on an academic paper. Cut a bit of slack.

    The person who wrote it was probably on their umpteenth story of totally different subjects for the day with more to come, and in no position to question the merits of the paper, just report what was said on a general news cast, not a specialized slot. .

    I’ve been there countless times in my career – try, for instance,as a general reporter, being dropped in to cover a worldwide conference on AIDS – 900 experts and 2,000 papers – when it was only just beginning to be known to the wider public.

  9. Some occupations just attract people’s attention in this way

    Yes, if it feels like magic and raw power, then … But the blacksmiths also liked to cultivate the rumors of their special relations with the dark forces on purpose. Like I remember the story of Shaka Zulu’s assegai forged in the dead of the night using things which weren’t to be named.

    (The only time I knew I was in Hell was when we visited an aluminum factory as interns … imagine a cavernous pitch dark hall with the lights dimmed by hydrogen fluoride patina of corrosion of windowpanes and fat, thick graphite dust, where the black crust of the ovens glows red through the cracks, and flickers of blue flame dance over it, with the disquieting low buzz of electricity drowning all sounds … so a cherry-glowing giant scoop swings overhead, unheard, and fades into the darkness … and soon the devil himself appears, a diminutive foul-mouthed man completely covered in soot, wielding a jackhammer to break the crust and to prevent boiling-over of the liquid metal)

  10. I’d be surprised if population histories and linguistic phylogenies were sufficiently correlated with folktale genealogies to be of much use.

    I meant to say,

    I’d be surprised if folktale genealogies were sufficiently correlated with population histories and linguistic phylogenies to be of much use to those who study the latter.

  11. Trond Engen says

    There was a Norwegian experimental archaeologist a few years ago who added charred bones to the smelt to make carbon steel. He interpreted the tales of smiths’ black magic as trade guild metaphysics, interpreting the increased strenght of the metal as the spirit of the animal or the fallen enemy forged into the blade itself.

  12. David Eddyshaw says

    I have in front of me a Kusaasi (rural West African) folktale which is evidently just the same story as the Pardoner’s Tale from the Canterbury Tales.

    There is actually no great mystery about this; it’s a Buddhist Jataka story, incorporated into the Pancatantra, translated into Middle Persian and then Syriac and then Arabic (and thus to West Africa where it has metamorphosed into a folktale among a people in contact with Muslims but few of whom are themselves Muslim) and then Latin, and thus to Europe and Chaucer.

    No amount of Bayesian analysis would tell you this. It seems to me that this sort of enterprise is remarkably similar to the glorified mass-comparison stuff produced by non-linguists and published in journals like Nature by editors who seem unaware that there actually exists such a thing as comparative linguistics.

  13. It made me wonder if copycat children book writers, inspired by the Grimm brothers of Germany but introducing variations here and there, may have contributed to the present-day variability.

    Absolutely. “Goldilocks and the Three Bears” has a known point of origin, or at least a known point of dispersion: Robert Southey’s 1837 story (in which Goldilocks is not only an old woman, but the villain of the piece). All the folklore and literary versions since then descend from it.

    One pretty obvious question is what level of generality do you use to decide if two tales from different cultures/languages are variants of the same ur-tale.

    Tolkien complained in “On Fairy-Stories” about what he considered to be abuses of this very point:

    [T]here are many elements in fairy-stories (such as this detachable heart [which appears in the Egyptian “Tale of Two Brothers” and over and over in both folklore and literature since then], or swan-robes, magic rings, arbitrary prohibitions, wicked stepmothers, and even fairies themselves) that can be studied […]. Such studies are, however, scientific (at least in intent); they are the pursuit of folklorists or anthropologists: that is of people using the stories not as they were meant to be used, but as a quarry from which to dig evidence, or information, about matters in which they are interested. A perfectly legitimate procedure in itself—but ignorance or forgetfulness of the nature of a story (as a thing told in its entirety) has often led such inquirers into strange judgments.

    To investigators of this sort recurring similarities (such as this matter of the heart) seem specially important. So much so that students of folk-lore are apt to get off their own proper track, or to express themselves in a misleading “shorthand”: misleading in particular, if it gets out of their monographs into books about literature. They are inclined to say that any two stories that are built round the same folk-lore motive, or are made up of a generally similar combination of such motives, are “the same stories.” We read that Beowulf “is only a version of Dat Erdmänneken”; that “The Black Bull of Norroway is Beauty and the Beast,” or “is the same story as Eros and Psyche”; that the Norse Mastermaid (or the Gaelic Battle of the Birds and its many congeners and variants) is “the same story as the Greek tale of Jason and Medea.”

    Statements of that kind may express (in undue abbreviation) some element of truth; but they are not true in a fairy-story sense, they are not true in art or literature. It is precisely the colouring, the atmosphere, the unclassifiable individual details of a story, and above all the general purport that informs with life the undissected bones of the plot, that really count. Shakespeare’s King Lear is not the same as Layamon’s story in his Brut. Or to take the extreme case of Red Riding Hood: it is of merely secondary interest that the retold versions of this story, in which the little girl is saved by wood-cutters, is directly derived from Perrault’s story in which she was eaten by the wolf. The really important thing is that the later version has a happy ending (more or less, and if we do not mourn the grandmother overmuch), and that Perrault’s version had not. And that is a very profound difference […].

  14. Trond Engen says

    My reaction on this was immediately sceptical because of cultural transmission, more interested when I started reading saw that they tried to correct for geographic proximity, and increasingly incredulous as I saw what seemed to be a method for cherrypicking the best fits. But I also suspect I’m missing something. For one thing I have a hard time reading anything significant from Figure 3 (the scatterplot).

  15. David Eddyshaw says

    From Jeffrey Heath’s grammar of the Songhay of Timbuktu:

    “Among the other castes of greatest sociocultural interest, both feared and despised by mainstream Songhays, are the griots and the blacksmiths …
    … the blacksmiths …. are thought to have black-magical powers; most local blacksmiths are ethnic Tuaregs.”

  16. Trond Engen says

    Have a look at Figure 4 (the phylogenetic tree) and the occurence of different tales in different branches. The tales drop in and out of significance as we move up and down the tree, fairly unrelated to how secure they are in the constituent branches. No. 330, the only tale securely attributable to the top node, isn’t even on the list in Indo-Iranian. No. 328, only >50% secure at the top node, is >70% secure at most bottom and mid level nodes, and present in all of them. Obvious artifacts of statistics.

  17. Oh, come on,. Hat. It is a *news* story, not a report to a conference (or blog) of linguistics experts, but to the great general public, based on an academic paper. Cut a bit of slack.

    Of course, and I don’t expect news stories to resemble academic papers, I’m just having a bit of fun. No offense intended, Old Hack!

  18. Trond Engen says

    The hyperbole of the BBC article was probably copied straight from the press release, and the press release was probably (contrary to cherished misconception) written by the study’s authors.

  19. I don’t know enough about Indo-European linguistics to comment on the direct influence of the trees. I do, however, know a lot about statistics, and the results did not seem especially strong in that area. (I was particularly puzzled by what they were doing with the Markov chain Monte Carlo, but I suspect that is mostly a weakness of their explanation.)

  20. David Marjanović says

    One needs to mention (following Trond) that the statistical significance wasn’t properly corrected for multi-testing so the statistical significance is inflated.

    Ouchie. That’s a pretty basic mistake.

    Anyway, I join the chorus of those who think transmission has been underestimated. For starters, the Grimm brothers were full of shit; most of the oh-so-ancient German folk tales they collected had come from France within the previous 200 years – none less than Tolkien has beaten me to bringing up Red Riding Hood.

    The Nart sagas are all over the Caucasus, straight across the language families. Ossetic has them, and it only entered the area 1500 years ago.

    The epic of King Gesar is all over the Tibetan cultural sphere, including whole language families that are very distantly related to Tibetan. Bonus: some think that Gesar is Caesar.

    Alexander the Great has become legendary throughout his empire and beyond, sometimes showing up in what seem to have been unrelated tales.

    Then there’s St. Jehosephat, the bodhisattva…

    I wonder if the test for geography vs. phylogeny failed to take into account how much these two are correlated. Adding Finnish and Hungarian, as suggested above, would certainly have helped.

  21. The Nart sagas are all over the Caucasus, straight across the language families. Ossetic has them, and it only entered the area 1500 years ago.

    And some think they influenced the King Arthur legend!

  22. Re: Geser/Caesar

    One of the most common Mongolian mantras is “om Ochirvaani umpad” (Sanskrit oṃ vajrapāṇi hūṃ phaṭ)

    This is actually a prayer to Heracles…

    https://en.wikipedia.org/wiki/Vajrapani#Appearances_and_identifications

  23. On the topic of Indo-European: some game developers have hired linguists to create multiple dialects of “Proto-Proto-Indo-European” (i.e. Pre-Proto-Indo-European) for a game set 12,000 years ago. Says one dev: “When we first started recording, it felt very modern. It was like listening to a language coming from Eastern Europe or something.” [The Kurgan hypothesis is true!] “It was very verbose. So we worked with them to regress this language to something that would be even earlier than that.” Proto-Flintstone-European, then. But at that time depth, might they not better served to look at proposed Nostratic or Eurasiatic roots?

  24. The headline I saw was “Phylogenetic analyses suggests fairy tales are much older than thought” which initially made me think “Finally Julian Jaynes has been vindicated!” Less ambiguity in headlines, please.

    I am dubious about The Smith and the Devil going back to the Bronze Age, since the magical status of the smith in folklore depends on the use of iron.

    I don’t think it’s controversial that people enjoy a good story, and that stories can cross linguistic and cultural boundaries. Some Chinese folktales are related to European ones. However, as far as I know, no Native American stories are related to Eurasian ones. That would demonstrate a limit on how far back relationships can be traced.

    There’s a claim that Child Ballad No. 4 (The Outlandish Knight), which is related to an old Hungarian folktale, is illustrated on a gold Scythian scabbard dated to the 3rd century BC. (See discussion at http://mudcat.org/thread.cfm?threadid=22848, http://mudcat.org/thread.cfm?threadid=143147). If nothing else, it shows that people have been considering these matters for some time.

  25. Regarding the “Proto-Proto-Indo-European” dialects commissioned by a game developer, I had heard the news mentioned but did not realize that the game was set 12,000 years ago. That’s roughly twice as old as Proto-Indo-European itself. We might disagree on how old PIE is, but we can all still agree that this is many thousands of years before PIE for sure. The article says the game is set in the Carpathian Mountains, somewhere in Slovakia. At that time depth, it’s anybody’s guess where Pre-Proto-Indo-European would have been spoken, though the Carpathian area would have been speaking something else by the time of PIE. I would have just invented a language that looks like the various proto-languages reconstructed in Eurasia without claiming any relationship to PIE.

  26. The language of the Ulam tribe constructed by Anthony Burgess for Quest for Fire looks (at least lexically) like an Indo-European-based pidgin. Atra ‘fire’ is even distinctly Iranian (or maybe Proto-Albanian), tir ‘game’ is High German, and bratt ‘brother’ is Slavic.

  27. David Eddyshaw says

    “However, as far as I know, no Native American stories are related to Eurasian ones.”

    The rabbit as too-clever-for-his-own-good trickster, who is certainly a West African, also turns up in Muskogean stories; but although I think I’ve seen it asserted that he’s Amerindian, it seems on first principles pretty likely that Choctaws could have heard about him from enslaved Africans. (Not unlike the Sioux Norsemen we just heard about.)

    Of course, rabbits may just *be* like that.

  28. David Eddyshaw says

    https://en.wikipedia.org/wiki/Br%27er_Rabbit

    Cherokee, rather than Muskogean according to the ‘pedia; I’d forgotten about the Algonquian trickster-rabbit, too.

    Whatever else, it all demonstrates just how contagious good stories are.

  29. marie-lucie says

    In Alaska and the Northwest there are many stories about Raven, usually the trickster, as in some Siberian traditions. As you go South along the Pacific, very similar stories are attributed to another animal trickster, such as Mink or Coyote.

  30. Jim (another one) says

    “Some Chinese folktales are related to European ones. However, as far as I know, no Native American stories are related to Eurasian ones. ”

    I know of another non-exception exception to this. There used to be an anthropologist who was active in the Southwest collecting among other things folk stories. One time when he went out, he was astonished to hear a story that resembled the Bremen Town Musicians. Then it dawned on him what had happened. Collecting these stories is not just a matter of saying “Okay, the tape recorder is going; start telling your story.” You had to offer a story to get one back, and he had told the story of the Bremen Town Musicians. What threw him off at first was the substitutions. Instead of three animals there were four and the European animals had been replaced with ones that made more sense locally. I remember that one of them was a coyote.

  31. My understanding was that Br’er (/brə/ < brother) Rabbit was a Native American relexification of Anansi, who is definitely West African, or perhaps a merger of tricksters.

  32. marie-lucie says

    But a number of European-like tales in Canada are attributable to the shared lives of French and Indian paddlers and other lowly employees of the fur trade in past centuries.

  33. David Eddyshaw says

    @John Cowan:

    The rabbit is certainly West African. (He’s Asuong or Asumbil in Kusaal: “Mr Rabbit.”) He’s a rabbit in Hausa too. I’ve got the Tar Baby story in Hausa somewhere (it has more sex in it than in Uncle Remus’ version.)

    I seem to remember you yourself demonstrating that he was of the same mythological heritage as Anansi, despite what would seem to be fairly large barriers of species.

  34. David Eddyshaw says

    I suppose Bugs Bunny has a fair claim on the inheritance. He’s better at it than Brer Rabbit, though. Must be those extra centuries of practice.

  35. What threw him off at first was the substitutions. Instead of three animals there were four
    I’m puzzled by this – the Grimms’ version of the Bremer Stadtmusikanten has four animals (donkey, dog, cat, rooster). Maybe it was the other way round, four substituted by three? Or do you know a different European version with three animals?

  36. What could be said about the famous Aesop’s Fables then? Could they echo older, widespread tales shared by various people in SE Mediterranean, or were they inspired by his immediate environment? (He himself was supposed to have come from Asia Minor originally.) Were they the product of just one man’s imagination? Were they later received so well in other languages only because they were translated into them, or because their standard characters (smart animals, lazy people, etc.) were already part of the local folklore? One way or another, their perseverance is remarkable.

  37. marie-lucie says

    Maybe it was the other way round, four substituted by three?

    Many peoples have a specific, perhaps “magic” number which is important in story-telling. For Europe it is usually three, in the Northwest usually four, sometimes five or six. It is possible that the native version substituted its own number. Alternately, the story-teller might not have had a lot of suitable local animals to choose from.

    In many cases the traditional number applies to a group of N+1 rather than simply N: N brothers (let’s say) have identical adventures and each one in turn fails, but the last one (who may be the only one named) does things differently and wins. The group of animal musicians is different, as each one has his own distinctive “song”. They also fit the composition of the stereotypical four-voice choir (which may be why they need to be four rather than three). In any case they are a solid group, and stand or fall together.

  38. marie-lucie says

    Ariadne: Some of Aesop’s fables are also found in India.

    With ancient literature as well as folk tales, the role of the known “author”‘s was not to invent the stories but to tell them in a particularly memorable manner. Aesop, Ovid, Homer, or in France Perrault (tales) or La Fontaine (fables) (and similarly other authors in different countries) largely transmitted much older traditions, in their own distinctive manner and style.

  39. I thought that Br’er Rabbit originated in Lapine (Adams 1972).

  40. David Eddyshaw says

    @Brett:

    In that case we have to postulate an origin of the tale many millions of years ago. Unless, of course, there were human beings prior to Adams who spoke Lapine.

  41. I recall hearig that Michael Witzel’s The Origins of the World’s Mythologies had an even more grand scope and grand conclusions, organizing mythologies from all over the world into one historical account. I haven’t read the book. I’d be curious to know what others think of it.

  42. David Eddyshaw says

    Sounds like a rip-off of Edward Casaubon’s magnum opus, sadly unfinished at his death.

  43. David Eddyshaw says
  44. Ah, Neal Ascherson — always enjoyable! “Mr Casaubon, more than a century before his time, was a post-structuralist.”

  45. Trond Engen says

    I think it’s time to draw some lines here.

    1) The news stories take this as proof that some stories are very old. That’s not news. We know about Naticve American and Aboriginal Australian stories of natural disasters that are thousands of years old. The story of the great deluge may belong here too. But we don’t have to stop there. There’s hardly any doubt that stories have been continually told and retold since the first campfire, but I’m also in no doubt that they’ve been altered, split, merged and remerged beyond recognization for just as long minus one week,

    2) Some shrug and say horizontal transfer i more important. That may well be. But there’s no doubt that stories are told to children and grandchildren (“vertically”) as well as when meeting people from other places (“horizontally”). The authors of the paper admitts the importance of horizontal transfer up front and tries to correct for it. Given the mere existence of parents and grandparents anything but a clear vertical signal would be surprising for short timespans. For longer spans not so much..But the paper doesn’t go there.

    3) The claim of the paper is something much more specific: That the effect of vertical transfer associated with one cultural trait, language, can be identified, and that the linguistic affinity of the latest common origin of some fairytales, and hence their latest possible dates of creation, can be reconstructed with confidence through statistical methods. What I think we’ve concluded is that the paper falls well short of those claims.

  46. Have a look at Figure 4 (the phylogenetic tree) and the occurence of different tales in different branches.

    Where did they get that tree? Why is Indo-Iranian branching off first when it is clearly related to Balto-Slavic? It all looks bizarre.

  47. Trond Engen says

    Rick: Where did they get that tree? Why is Indo-Iranian branching off first when it is clearly related to Balto-Slavic? It all looks bizarre.

    Oh, that?

    Trees for our study were sourced from Bouckaert et al.’s [2,29] Bayesian phylogenetic analyses of Indo-European languages.

    … so it’s garbage. I didn’t mention it since I didn’t think it was important to the overall point. But it clearly is, since Bouckaert et al. essentially just measures the geographical distance anyway.

    As for the lingusitic tree, it would have been a nice thing.to try switching the branches around to see if the best fit matched some near-consensus linguistic tree more than geographical proximity. That would also allow adding Uralic, Turkic, Dravidian and the full compass of Caucasians as controls.

  48. David Eddyshaw says

    @SFReader:

    “One of the most common Mongolian mantras is “om Ochirvaani umpad” (Sanskrit oṃ vajrapāṇi hūṃ phaṭ)”

    Belated but heartfelt thanks. Mongols praying to Heracles (which seems oddly right, somehow) has now officially taken over as my poster child of cultural diffusion from Tibet’s national epic starring C Julius Caesar, beating St Bodhisattva into third place.

    Truly, all This is That …

  49. Aladdin and his lamp have also made their way into the Grimm collection.

  50. The Japanese wind god Fūjin is said to be derived from the Greek Boreas.

  51. David Eddyshaw says

    Those Greek-speaking Buddhists seem to be at the back of a lot of apparently unlikely long-range connections.
    The Pali Canon itself includes

    https://en.wikipedia.org/wiki/Milinda_Panha

    the Questions of Milinda, i.e. Menander, a Greek king in Bactria who conquered much of north India and promoted Buddhism.

  52. David Eddyshaw says

    Apropos of nothing much except that it’s cool cultural diffusion and involves Greek, the Nubian kingdom of Makuria in the first millennium

    https://en.wikipedia.org/wiki/Kingdom_of_Makuria

    used Greek as its primary written language (also Coptic and Old Nubian) and Greek may even have been the language of the court.

    Prince Georgios of Makuria is said to have travelled to Baghdad to intercede (successfully) for remission of tribute under the “Baqt” i.e “pact”, the treaty between Makuria and Muslim Egypt.

  53. Great heavens, if I’d ever known about that I’d forgotten it! This bit from the Wikipedia article is strikingly idiotic:

    The Nubians were a literate society, and a fair body of writing survives from the period. These documents were written in the Old Nubian language in an uncial variety of the Greek alphabet extended with some Coptic symbols and some symbols unique to Nubian. Written in a language that is closely related to the modern Nobiin tongue, these documents have long been deciphered. However, the vast majority of them are works dealing with religion or legal records that are of little use to historians. The largest known collection, found at Qasr Ibrim, does contain some valuable governmental records.

    In what universe are works dealing with religion or legal records “of little use to historians”?!

  54. Governmental records! How fascinating! 🙂

  55. Trond Engen says

    In what universe are works dealing with religion or legal records “of little use to historians”?!

    Heh. In a universe where historians are trying to understand the political history of a kingdom, I suppose. Actually, I thought something along those lines might have been edited out, but the sentence has ended just like that since the very first version of the article more than 10 years ago.

  56. David Eddyshaw says

    I don’t know about the stuff in Greek or Coptic, but the total amount in Old Nubian is agonisingly small. According to Gerald Browne’s Old Nubian Grammar it would amount to about a hundred pages of continuously printed text. This is actually a lot more than once was the case; the archaeological rescue work engendered by the Aswan Dam project quadrupled the amount of material.

    Of what there is, about half is translation from Greek, New Testament and other Christian writings, and the remainder “public contracts, private letters and similarly ephemeral material.”

    The translation stuff is of course invaluable for the language, but tells you nothing about Nubia except what they valued in reading material; the limitations of the corpus make interpretation of the “ephemera” hard, by the sound of it.

    There’s a lot of interest (I gather) in involving Africanists in the Old Nubian work, which has traditionally been carried out by people more from the “philological” Classicist side. There are various modern forms of Nubian which descend from the old language. I was mightily irritated by the blurb on the back, though, which says: “it is the only indigenous African language whose development we can follow for over a millennium.” Get a map of Africa, guys …

  57. David Eddyshaw says

    “tells you nothing about Nubia except what they valued in reading material”

    I suppose it also tells you there must have been a good many people who could read Nubian but not Greek (or else why make a translation?) which says at least something about the status of the language.

    I’ve worked in places where if someone says “He’s literate” they mean “He speaks English.” Evidently it was not like that in Makodia. (Well, I don’t suppose they spoke a lot of English, anyway.)

    With Nubian it may well have been a genre thing though; in Coptic as well there is lots of Christian (and heretical religious) material and some “ephemeral” stuff but little or no history or science or secular artistic literature or even systematic theology, because if you wanted to read (or write) that you’d do so in Greek, or – later on – in Arabic.

  58. David Eddyshaw says

    Makuria, not Makodia. I’ve carried out an Anschluß by Makuria of its upstream neighbour Alodia. Apologies to both parties involved.

  59. David Marjanović says

    And some think they influenced the King Arthur legend!

    Ah yes, King Arthur, protector of Wales… or England… or France.

  60. marie-lucie says

    David: King Arthur, protector of Wales… and of Brittany perhaps, rather than France. There are medieval French texts adapted from the King Arthur legend, which were known as la matière de Bretagne ‘the matter of Brittany’.

  61. I note that Nubian language was written in Cyrillic

    http://condor.depaul.edu/sbucking/extra/coptic.gif

    sort of…

  62. Re: ” However, as far as I know, no Native American stories are related to Eurasian ones.”

    This is not true at all.

    Berezkin’s catalogue of folklore motifs distribution clearly shows that majority of Native American folklore is of Eurasian origin. Small part, primarily centered in South America, is of, “Indo-Pacific” origin.

    Few concrete examples

    https://www.folklore.ee/folklore/vol36/berezkin.pdf

  63. note the maps.

  64. In the King Arthur legends, there simply is no English Channel: consequently, the Arthurian stories are called the Matter of Britain in English, rather than of Brittany. Lancelot can ride all the way to Arthur in Camelot/Winchester from his castle in Benwick, though it takes him quite a while to make the journey. Benwick is Benoic in French, and is said to be on the river Loire not too far from Bourges, possibly in the commune of Issou(n)dun. Note that Issoudon does have a castle and was near the old border between the Kingdom of France and the Angevin Empire (as well as, by chance, the border between occupied France and Vichy France).

  65. George Gibbard says

    Makuria may have been a more interesting place linguistically even that that: the king lived at Old Dongola where Old Dungulawi was spoken, but written Old Nubian is agreed to be the ancestor only of Nobiin, spoken further north than Dungulawi (Bechhaus-Gerst recommends Old Nubian be called Old Nobiin instead). It is likely that Old Nobiin writing originated in the kingdom of Nobatia (capital Pachoras, which was Faras in Egypt until it was submerged as a result of the Aswan Dam) which Makuria absorbed. The most northerly part of Nubia (Dar Kunuz until the Aswan Dam), which was apparently conquered by Silko of Nobatia from the Blemmyes (Beja), eventually wound up speaking Kenzi, a descendant of Old Dungulawi, though it was separated from the Dungulawi area by a 400-km stretch of Nobiin speakers.

  66. marie-lucie says

    JC, you must be right about Bretagne meaning “Britain” in this context, but since the Bretons came from Britain only a few centuries before, and they never lost all contact with the Celtic areas of Britain, the stories were common to both.

  67. David Eddyshaw says

    “Ah yes, King Arthur, protector of Wales… or England… or France.”

    Scotland, really. Arthur’s Seat is in Edinburgh.

  68. David Eddyshaw says

    “I note that Nubian language was written in Cyrillic” (linking to a Coptic alphabet)

    Although the script has some letters borrowed from Coptic (and Meroitic), Browne bangs on quite a bit about the Nubian script *not* being based on Coptic in the sense that “the ductus litteratum is Greek, not Coptic.” He says that this enabled the excavators of Qasr Ibrim to distinguish Nubian and Coptic fragments even without knowing the languages. He says published texts have used the Coptic letters for convenience. His grammar actually uses a purpose-designed Nubian font, which does look a bit unCoptic, though I can’t say I’d have noticed without prompting.

  69. David Eddyshaw: What’s St. Bodhisattva? Can’t find it on Google.

  70. David Marjanović says
  71. Trond Engen says

    SFReader: Berezkin’s catalogue of folklore motifs distribution clearly shows that majority of Native American folklore is of Eurasian origin.

    Thanks. The Berezkin articles are very interesting. Some of his motifs are clearly related, others less so. I read the one about pygmies without mouths very recently, I think, in a retelling of a medieval traveller’s account from the Volga. Now I’ll just have to find that book again…

  72. St. Josaphat.

    I see from the flyleaf it was almost twenty years ago that I bought (at the late lamented Soho Books in Manhattan) The Balavariani: A Tale from the Christian East, translated from Old Georgian by D.M. Lang; I well remember my astonishment to learn of the wide dispersal of this Buddhist story.

  73. @DavidEddyshaw: “The rabbit is certainly West African. (He’s Asuong or Asumbil in Kusaal: “Mr Rabbit.”) He’s a rabbit in Hausa too. I’ve got the Tar Baby story in Hausa somewhere (it has more sex in it than in Uncle Remus’ version.)”

    The Javanese trickster animal is Kancil the mouse deer, but several of the stories told about him are remarkably similar to Uncle Remus stories. In a field methods course when I was an undergraduate, our language consultant told us the story of Kancil and the tar baby (well, the doll covered with sticky stuff). A bit of search a few minutes ago turned up a version in a paper by R.V. Winsedtt, “Some Mouse-deer Tales“, Journal of the Straits Branch of the Royal Asiatic Society, 1906, who writes “Mr. George Maxwell and others have reminded me, that one of these tales of mine bears an extraordinary resemblance to that of Brer Rabbit and the Tar Baby.”

    I’ve always wondered how that story ended up in the American South and in Java. Aurelio Espinosa, “A New Classification of the Fundamental Elements of the Tar-Baby Story on the Basis of Two Hundred and Sixty-Seven Versions“, The Journal of American Folklore 1943, affirms his “belief in the India origin of the tale in the sense that India is as far back as we can trace it, and that it is not of African origin as some have believed”.

    But anyhow, neither linguistic phylogeny nor geographical proximity seems to have been especially important in this case…

  74. David Marjanović says
  75. David Eddyshaw says

    @Mark Liberman:

    Very interesting. Makes the tar-baby story look like a practical human universal (perhaps Brett is right and it’s proto-mammalian …)
    Interesting especially in that the Hausa version I know (tar baby as Hot Babe) is obviously not unique.

    Even so, the conjunction of Trickster-Rabbit and Tar Baby is a lot more specific, so the coincidence of the appearance of that form of the story in savanna-zone West Africa and the American SE is still striking.

  76. SFR, thank you for the Berezkin links.

    Prof. Berezkin was at the 2008 conference on the Dene-Yenisei hypothesis, among other non-linguist specialists in the relevant circumpolar cultures. He was especially interested in possible additional instances or even fragments of the “Cosmic Hunt” myth involving the stars of Ursa Major, which I was not familiar with.

    On the other hand, your other link, to the Dwarfs and Cranes story, was especially interesting to me as I had studied a story in which dwarfs fighting birds were an episode, with a footnote from Boas citing the Greek equivalent (that I had never heard of). My interest at the time was purely linguistic, but I should reread the story for its mythological content and Berezkin’s article for background.

  77. ə de vivre says

    Maybe those of you with more knowledge of statistical analysis and historical linguistics can help me out with this article. My understanding is that in historical linguistics mother-daughter relationships are established from regular sound correspondences and morphological similarities that then allow you to distinguish between genetic and borrowed or chance similarities between any two given languages. Is there analogous method in this study, or for any proposed genetic relationship is the alternate ‘borrowing’ hypothesis essentially unfalsifiable? Or did I completely miss what it’s claiming?

  78. Trond Engen says

    ə de vivre: Is there analogous method in this study, or for any proposed genetic relationship is the alternate ‘borrowing’ hypothesis essentially unfalsifiable? Or did I completely miss what it’s claiming?

    I haven’t looked into the actual statistics (indeed, I haven’t even checked if it’s available as an appendix to the paper), but my understanding is that there’s no comparative method in this study, How could it? To paraphrase Berezkin in one of his linked papers, tales are too fluid to be compared element by element in search of regular correspondences.

    What they do is use the standard catalogueization system of folklorists, the Aarne Thompson Uther (ATU) Index, picks a subset they judge suitable for the purpose, and check which catalogue numbers are known to be present in which countries/linguistic groups. They know that horizontal transfer (“borrowing”) is very important, so what they try to do is discern a signal from vertical transmision from the horizontal background noise.

    They establish a benchmark level of horizonal transmission through correlation of shared tales with geographical proximity. This level is presumably much higher than in historical linguistics. Then they look at the distribution of of tales to find those whose shared distribution among linguistic relatives is higher than would be expected from geographic proximity alone. After that (by some trick I’ll not pretend to knoe, but it may simply be the same routine again), they see if the shared distribution between presumed IE sub-branches is such that the tales can be attributed to the common linguistic ancestor. They come up with one whose broad distribution gives it more than 70% probability of being shared inheritance of all Core IE.

    There’s no falsification beyond statistics. And the statistics looks very weak. The sheer number of tales, already picked for their generality (“magic”), should be expected to produce false hits, at least hits with the bar of confidence as low as 70%. I wonder if what happened was that they went out looking for a clear signal in the data, and when nothing came up, they turned the control for geography upside down, using it instead to cherry-pick the best fits. Then they repeated until fade.

  79. ədv: phylogenetic approaches, like this one, don’t address the issue of chance resemblances. They take lists of cognates that were prepared in some other way, and attempt to use them to generate plausible clusters. If a certain set of words/folktales/etc. make the languages/traditions/etc. in question cluster nicely and consistently, you can argue that these all follow the same pattern of differentiation. If some other words/folktales don’t follow that pattern but are correlated geographically, you can argue for a borrowing scenario.
    For folktales there is nothing as precise as linguistic sound laws to judge whether two stories from different traditions are related by chance or historically.

  80. ə de vivre says

    Ah, thanks. Comparing correlations with linguistic proximity versus geographic proximity makes sense (or at least isn’t completely insane if you accept their premises). It seems like a strange mix of late 19th century structuralist folklore and modern large scale statistic analysis. I didn’t think people really did mytheme-style folklore analysis anymore except for lazy critics who use Joseph Campbell to avoid having to think about why they did or didn’t like something (not that I have unnecessarily strong opinions about that or anything).

  81. David Marjanović says

    Ah, we’re getting into the distinction of primary homology and secondary homology. 🙂 Primary homology must be established before any phylogenetic analysis; it means “systematically similar enough that it should be coded as the same character state; it could be secondarily homologous/cognate, but it could also be due to convergence/chance resemblance or to lateral gene transfer/borrowing”. You construct a whole dataset of these. Then you calculate the trees that explain the dataset most parsimonously; and then the trees tell you which primarily homologous features are secondarily homologous = are inherited from a common ancestor.

    In this case, the assessments of primary homology between tales weren’t made by the authors of the study, but taken from the ATU database. The tree wasn’t made by the authors either, but taken from Bouckaert et al. (2012 – not the best choice as of 2015, but fairly close as far as the topology rather than the length is concerned). They optimized their characters on the tree, or without the jargon they mapped the character states (each ATU number present/absent) on the tips of the tree and had the computer calculate their most parsimonious evolution along that tree, so that at each node they got a probability for the presence/absence of each ATU number. This is a test for how much phylogenetic signal the dataset contains: if there are few changes along the tree, there’s plenty of signal, while if there are many changes, it’s closer to random noise. The authors compared the results from this to an analysis without a tree that used geography as the predictor instead, and found that the tree fared better.

    Geographic distance, however, is a rather crude measure of the probability that a tale will spread across language barriers.

  82. Trond Engen says

    Thanks, David. I’ve been hoping that you and Dimitry would dissecate it for me. When I’ve become more and more outspokenly ignorant, it’s with the noble aim of coercing you into surrender.

  83. David Eddyshaw says

    Culture is likely to be more significant than simple proximity in the case of horizontal spread of folktales. Obviously culture and geography do correlate, but in the fairly frequent cases of mismatch culture is surely going to be the important thing.

    Islam is an obvious vector of stories; I’m also beginning to wonder from this thread if perhaps all the stories in the world are originally Buddhist.

    Although Arabic is in one way *the* Islamic language, genuine historic Islam (as opposed to the crippled parody which masquerades as Islam so often on the news) has always picked up cultural strands from many, many different linguistic groups and passed them on to many others; and as for Buddhism …

    (Apropos of nothing, reminds me of the catty remark that the Pali Text Society’s translations are in Buddhist Hybrid English.)

  84. Zhanara Dayrbekova says

    “Berezkin’s catalogue of folklore motifs distribution clearly shows that majority of Native American folklore is of Eurasian origin.”

    This is very true. I recently read a collection of Native American folktales or myths and many plots reminded me of several themes in the Epic of Manas.

  85. Trond Engen says

    For Dimitry read Dmitry. Sorry.

    Culture is obviously important, but how to measure it? By using religion, I suppose. Political entities are also important. As are trade connections. The folk tale inventory of different peoples and regions of the Balkans could be illuminating.

  86. ZhD: majority of Native American folklore is of Eurasian origin.”

    Some people will think that this means that Native Americans were taught the stories by Europeans, rather than they already knew the tales from their origin in Asia. Either way, some people will be offended.

  87. David Eddyshaw says

    The thing about Language is that the form of individual words is *arbitrary* except for particular domains like onomatopoeia and baby-words, so that if different languages show a great deal of resemblance in word forms, that calls far some sort of explanation via descent or borrowing. So far as statistical methods can give meaningful results in comparative work, this is critically dependent on the arbitrariness of speech. Even when you move from vocabulary to syntax within Language the method starts to break down, because there are only so many ways you can string a sentence together so that people can still parse it, and the various combinations of choices you might think it possible to make turn out not to be independently variable – hence all those Greenbergian partial universals (the Good Greenberg.)

    None of this arbitrariness is true for stories. There’s nothing very implausible about completely separate groups having come up with very similar stories by pure coincidence, and given the constraints that the story has to be coherent and memorable and relatable-to by actual people, and made out of things that actual people encounter (like animals) a further process of natural selection is going to result in a good bit of convergent evolution. If the groups in question live in similar material circumstances, it’s going to be even less significant if they tell similar stories.

    On top of all that, there aren’t really any good metrics for measuring similarity of stories; certainly nothing nearly as reproduceable and objective as you can think up for words.

    The whole notion that there is a workable analogy between language change and folktale transmission bears no real scrutiny.

  88. -Some people will think that this means that Native Americans were taught the stories by Europeans, rather than they already knew the tales from their origin in Asia. Either way, some people will be offended.

    The former is likely to account only for a small minority of folktales.

    What I gathered from Berezkin’s articles (and books. But they exist only in Russian, I am afraid) is that most of the folk motifs around the world are very, very old and date from Upper Paleolithic Eurasia (20-30 thousand years ago).

    People had to do something in these caves around fire during long, cold nights of the Ice Age, right?

    And some stories are even older and were brought by modern humans from Africa (50 thousand years ago).

  89. Who do you think was the Wolf in Little Red Riding Hood?

    Neanderthal or Denisovan?

  90. David Eddyshaw says

    Well, how many potential stories are there?
    According to some of the more boring structuralists out there, there seem to be about three plots.

    While I have no sympathy for that sort of thing at all, I do think a microscopic kernel of truth lurks in it; and we can’t say anything meaningful about the likelihood or unlikelihood of similarities among the tales of different groups unless we have some sort of idea of the denominator. Even if we had a reliable way of assessing similarity.

    Neanderthal. Obviously. No Denisovans in Grimm-land.

  91. -Islam is an obvious vector of stories; I’m also beginning to wonder from this thread if perhaps all the stories in the world are originally Buddhist.

    Berezkin has a very short (one page!) article on the Big History (titled “What happened in history?”).

    It’s in Russian, but try google translate.

    He ends his big history of humankind with the establishment of permanent links between East Asian and Western Eurasian centers and formation of the “world-system” which he dates to last centuries of the 1st millenium B.C.

    That’s the last important thing which happened in history so far.

  92. I like one instance of an independent but similar folk-story theme: people marrying seals. There are versions at least from Scotland (the Selkies), the Pacific Northwest and Tierra del Fuego. In detail they differ quite a bit, but if you were to ignore these differences, you could waste a lot of time on trying to explain why all of them merit the same checkmark on the theme list.

  93. David Eddyshaw says

    Indeed. Lots of stories about people marrying animals (foxes, in Japan.)
    But I don’t think it’s surprising, or needs any sort of explanation by vertical or horizontal transmission.

    Story A:

    Neolithic Storyteller: Hey! Listen to this! A man killed a seal!
    Neolithic Audience: Triffic. What’s on television? How long do we have to wait?

    Story B:

    Neolithic Storyteller: Hey! Listen to this! A man married a seal!
    Neolithic Audience: Disgusting, yet strangely compelling. Tell us more …

  94. No, the Google translate would not do.

    The article on Big History is short, so I translated it myself (using Google Translate, of course).

    Here it is in its entirety.


    Yu.E. Berezkin

    “What happened in history ”

    “What happened in history?” was the title of Gordon Childe’s famous book, in which he formulated the concept of the Neolithic and Urban revolutions. Since its release in 1942 and until the end of the 1970s, the idea of such revolutions structured our vision of pre-literate history. However, with the progress of archaeological research it became clear that the heuristic value of this concept is limited, especially outside of Asia Minor. Since the late 20th century, it became possible to reconstruct the specific history of human societies instead of universal stages of development. When the geneticists have found that modern humans came out of Africa, c. 60 thousand years ago and then spread over the ecumene on certain routes, we got a new base model, which helps to structure the knowledge of the past.

    Partly a precondition and partly a consequence of the adoption of this model is the idea of higher, than it was customary to think, inertia of the culture in which its traditional forms, if they do not reduce the competitiveness, can be reproduced indefinitely. Elements of culture are exposed to natural selection influenced by a variety of factors, including random and unique, so full parallelism in the development of societies which lack exchange of information is impossible. Since the history has no laws, but only probabilistic regularities, we should focus on specific events and processes which resulted in development going this particular route. And the farther away from us the time of possible bifurcation points, the greater would be the effect of the realized choice. It can seem very important that the intelligent beings did not evolve from dinosaurs, but if the deuterostomes did not appear in the Precambrian, we probably would have had to evolve from the octopus.

    Reconstruction of the genetic history of mankind is getting more detailed each day, as well as getting updated and filled with new content thanks to the discoveries of archaeologists. Materials of comparative study of folklore and mythology associated with the processing of large volumes of data (tens of thousands of texts), make a picture of the past even more multidimensional.

    About 60 thousand years ago, modern humans formed in Africa began to migrate to Arabia and moved further to the east along the Indian Ocean coastline. No later than 45 – 40 thousand years ago, they settled the subcontinent of Sunda (now partially submerged southeastern edge of Asia) and the Sahul continent (Australia with Tasmania and New Guinea). In the Persian Gulf region, some groups who separated from these migrants have journeyed to the north and by 50/45 – 35 thousand years ago, they occupied roughly the same territory in which the Neanderthals lived earlier. About 30 thousand years ago, modern humans replaced in East Asia the Homo erectus who who lived there before. Soon afterwards, the humans reached northeast Asia (Yanskaya site), but it is difficult to say whether they came from the south-west across Siberia or from the south from China, continuing to spread northward from South-East Asia. In the era of glacial maximum 22 – 19 thousand years ago, the population of Northern Eurasia was reduced and the groups which passed through this “bottleneck” likely have changed their culture. 18-19 thousand years ago, they again began to settle in the territory north of about 55 ° N lattitude in north-eastern Europe and to the east of the Urals (as the dating of Diuktai culture in Yakutia in northeastern Siberia and the monuments of the Late Kama and Pechora basin show). Analysis of materials (P.Yu. Pavlov) has recently shown that a new population of northeastern Europe came from the southern regions of Siberia.

    About 15 thousand years ago, from a foothold in Beringia (in what is now the shallow waters of the Chukchi and Bering Seas) the colonization of the New World began, in which both the groups of continental-Eurasian and East Asian origin have participated. The first inhabitants of the Americas both anthropologically, and, most importantly, culturally, more resembled the Ainu and even Melanesians and Australian Aborigines than the modern American Indians. However, in the Holocene, these archaic groups were gradually replaced by Amerindian-like populations. Materials of physical anthropology with great likelihood point out the Sayano-Altai region as the ancestral home of the latter. In middle and late Holocene, the displacement of protomorphic (close to Melanesian-like populations) populations in the south-eastern edge of Asia by colonizing populations from southern China of Mongoloid type (familiar with the productive economy) has occured. In Melanesia and New Guinea, where the separate center of agriculture has developed, the non-Mongoloid populations have survived, although in Melanesia they had lost earlier languages.

    In the Near East, 10-12 thousand years ago, the formation of productive economy and relatively complex societies has occured. The tendency to develop in this direction emerged several millennia earlier. Having a much greater density than the surrounding areas, the Near Eastern farmers in 8 th century BC. began to expand in all directions, resulting first in neolithization of Asia Minor, Cyprus and Zagros, and then of the Southern and Central Europe, the Iranian plateau and the bordering areas of Central Asia, Caucasus, Eastern and Northern Africa. Then the cultural achievements spread to the west and north of Europe, the Eurasian steppes and in the west of South Asia. In northeast Africa, 10 thousand years ago, a separate (with no or minimal influence from Asia Minor) center of ceramics production and animal husbandry emerged. In Sub-Saharan Africa, the productive economy and the accompanying cultural achievements began expansion only in the 2nd millennium BC. under the influence of cultural impulses from the north.

    Of the above mentioned macro-processes, the most important – after migration out of Africa – is the separation of populations of the Old World into the continental Eurasian and Indo-Pacific groups, which for tens of thousands of years (45-15 thousand years ago or so) were separated from each other by sparsely populated or even uninhabited territories of Tibet and surrounding arid areas to the west and to the north of Tibet. The evolutionists up to the mid-20th century unsuccessfully tried to comprehend within the stadialist concept the accumulated differences in the culture of these two regions. The presence of these two distinct cultural complexes, which then partially (but not completely) mixed in the New World, as well as in Southeast Asia and Siberia, is reflected in the differences between the complexes of folklore and mythological motifs. This Indo-Pacific complex has African roots while the Eurasian continental complex is virtually devoid of such links. The mythology of the Sub-Saharan Africa is strikingly poor and basically boils down to a set of motives, which was known to the beginning of modern human colonization of Eurasia and has moved to the Indo-Pacific world. It is about a dozen motifs associated with explanations of mortal nature of people, and about as many related to ideas about the sun and the moon, the starry sky (the Milky Way, the Pleiades and Orion’s Belt, and the African astral mythology in general is poor), the appearance of the people from underground or from the sky.This also appears to include some Trickster motifs, some of which are familiar only to the people of Africa, especially to the Bushmen, and to the Australian Aborigines. All or almost all of adventure motifs of folklore common in Africa have probably penetrated here from Asia in the Holocene.

    The mythology of Native Americans, by contrast, is extremely rich because in its formation both sets – continental Eurasian and Indo-Pacific- were involved. By the beginning of the settlement of the New World, the culture of East Asia has been very complex, as evidenced by the spread of ceramics here already in the late Pleistocene. This technological discovery is important not only in itself, but as an indicator of the ability of the culture to innovate. It is possible that the bottle gourd first cultuvated in Asia, was known if not to the first inhabitants of the New World, then at least to the people who lived in America at the beginning of the Holocene. Iconography of cultures of the Northwest Coast of North America and Nuclear America (from Mesoamerica to the Central Andes) shows common elements with the East Asian iconography. This is due not to the Trans-Pacific sea contacts (only with the Polynesians, whose influence on the culture of Southern California, and possibly Chile was purely local), but to preservance of the ancient heritage of migrants who came to America through Beringia. This does not allow us to consider the development of the cultures of the New World as quite independent, since they began to develop on already quite formed Eurasian basis. Accordingly, the historical experiment of the parallel development of the civilizations of the Old World and the New World is not entirely pure.

    In Sub-Saharan Africa and Australia, spontaneous fundamental changes in the culture did not occur. The most common explanation is probably as follows. Such changes are rare and occur under unique set of circumstances. Huge spaces, a variety of natural conditions and their frequent changes under the influence of climatic variations, distant migrations and, as a consequence, the development of new natural niches – all this contributed to the cultural dynamics of Eurasian populations and after the Late Glacial Maximum led to formation of two powerful centers of cultural genesis: Near Eastern and East Asian . This East Asian center at a late stage also experienced external borrowings (chariots and horses, bronze, then iron), while the Near Eastern center, having expanded to include Europe, North Africa and part of western Asia, remained self-sufficient. Both centers were in a state of constant expansion, both demographic (its last stages – settling by Polynesians of Far Oceania and finally, the expansion of Europeans in the modern times) and cultural. Expansion of Bantu peoples in southern Africa who met here Europeans who arrived from Holland and then England is nothing, but a distant consequence of the process which began at the upper Euphrates and the Levant in the terminal Paleolithic. Since the end of 1st millennium BC., regular contacts (as opposed to previous sporadic and indirect contacts) between the East Asian and Western Eurasian centers were established, leading to the formation of so-called “world-system”.

    And in it, we have lived since then.

  95. ə de vivre says

    I think the problem is that in historical linguistics genetic and borrowing relations aren’t just abstractions over the data, they correspond to two different types of language acquisition (or maybe two poles on a spectrum of language acquisition). You acquire your native language(s) during a critical period of infancy when you learn lots of complicated stuff that for the most part isn’t available to reflection. After that any language acquisition is much more conscious, simpler, and more available to reflection. For stories, everything is borrowed. There is no privileged genetic transmission, so just because story X is highly correlated with language family Y, that in itself doesn’t tell us anything about the relationship between the most recent common ancestor of the language and the origin of the story.

  96. Either way, some people will be offended.

    Some people are offended by the claim that Native Americans came from Eurasia in the first place. Or for that matter the claim that Navajos and Apaches came from the North.

  97. Trond Engen says

    David Eddyshaw: None of this arbitrariness is true for stories.

    This is true, but not completely true. Stories are made up of elements that are put together arbitrarily, These elements or subplots may be borrowed wholesale into completely different stories. Some of these elements are in themselves complex, some are simple, but strung together they can make something that hardly could have occured by chance. Repeat this over several stories or whole mythologies, and you have evidence of cultural ties.

    It follows that the more general the plots, the weaker the results. The phylogenetic study uses ATU numbers, a classification designed to be general, with no attempt to build an internal phylogeny (or, since this is philology, provenance). This is like doing mass comparison without weeding out borrowings. I’m not sure which way it skews the results, but it certainly makes them weaker.

    Berezkins approach is quite different. By comparing stories on the same theme in detail, he’s able to find some instances of quite complex plotlines or metaphysical concepts that are shared between different peoples far apart. I don’t think it’s evidence of pseudo-genetic inheritance along with language, but it does point to (recent or ancient) cultural contacts. I’m e.g. convinced that the circumpolar cultural sphere is one of contact and shared metaphysics rather han common descent. The tundra and the arctic coast have been underappreciated as a spreadzone. Samoyedic and Yukaghir each cover a stretch of coastline equalling roughly half the distance between Dene and Yeniseian, and Inuit even more. All are suspected to be recent arrivals in most of their range.

  98. The whole notion that there is a workable analogy between language change and folktale transmission bears no real scrutiny.

    I agree; while it may be true that “strung together, [plots] can make something that hardly could have occurred by chance,” there is no way to quantify it, and anyone looking to build a theory of inherited stories is obviously (based on human nature) going to overinterpret. There is simply no way for a believer to convince a skeptic that such inherited links are true. Contrast the Indo-European thread (still going strong at well over 800 comments!), where elaborate attempts to support insufficiently based theories are being efficiently shot down by people who actually know what they’re talking about. Facts are facts.

  99. Trond Engen says

    I don’t think we disagree much. I’m usually annoyed by attempts to attribute certain minimal and very general plotlines (“guy stealing a cow”) to PIE mythology based on scattered evidence, but they gain strength with complexity (“ginger-haired warrior with a thunderbolt weopon killing dragon and stealing cows”). But as you say, there’s never any way to qualify the impression. In this way “comparative folkloristics” is more like archaeology, but without the dating: But as with archaeology, one might discern patterns and build hypotheses that might turn out to be powerful in coordination with other disciplines. The recent advances in ancient population genetics is transforming European archaeology as we speak, as pots turn out to have language and genes after all. I wouldn’t be surprised if general patterns of distribution of complex plotlines over a large number of myths turned out to correlate with genetic evidence. If so, it might predict even linguistic relatedness. But we’re not there yet.

  100. Sure, and I’d be much more inclined to take an interest in such theories if they weren’t so prone to be embraced by cranks with no idea of what the word “evidence” means. I agree we don’t differ much if at all; I’m just more inclined to be impatient and dismissive.

  101. Stories (or better motifs, which are an element of a story) are memes and reproduce by retelling. From one generation to another or from one tribe to another, there is no functional difference. So there is no point in contrasting inheritance and borrowing, both are natural pathways for story transmission.

    Once born, a story is basically immortal quite unlike languages.

    The reason is simple. Reproduction of story is millions of times easier than transmission of language. It takes several years to teach a child a language, but five minutes to retell Cinderella story!

    To kill a story which people find interesting, you’d have to exterminate all tribes where it was ever told without a single survivor.

    And fortunately such things are rather rare.

    A story told in dozens of unrelated languages across large distances is the most natural thing in the world and doesn’t require explanation.

    Explanation is needed when the same story is told in very distant places which don’t have contacts with each other, but nowhere in between. It could be either a parallel invention or very ancient migration.

    Now, if we have not one story, but several dozen or several hundred stories which are found only in Alaska and Tierra del Fuego, but nowhere in between, the ancient migration argument is the only valid.

  102. marie-lucie says

    Retelling stories does not always mean saying the same thing over and over again. As Lévi-Strauss showed, it is very common for a story to be altered in a way that differentiates one people’s tradition from another. For instance, if in one story the hero travels to the sky, the neighbours might tell a similar story but with the hero diving to the bottom of the sea. When enough details are changed, the story might be unrecognizable.

    Marrying animals: Stories of men or women marrying animals are not just for entertainment, much less presented as interesting news (“Hey, listen to this, a man married a seal!” as David E suggested), but placed in a mythical time when humans and animals were part of the same world – they might all have been created as rather undifferentiated beings which were given shape and other attributes as well as appropriate ways of living by a Transformer.

  103. Trond Engen says

    marie-lucie: When enough details are changed, the story might be unrecognizable.

    True, or be similar only on the surface. But all the more power when a complex storyline is shared.

  104. Lévi-Strauss having been invoked, I recall that he thought the circumpolar cultural sphere might be behind resemblances between certain Algonkian myths and the Grail cycle. Whether this is tenable in 2016, I don’t know.

  105. Retelling stories does not always mean saying the same thing over and over again.

    Indeed. I learned “The Rhyme of the Nancy Bell”, whose meter and rhyme tend to stabilize it, at my father’s knee. Only much later did I find out that it was by W. S. Gilbert, who I already knew and loved, and discover that the text I had memorized was somewhat different from Gilbert’s 1866 version (even to the title; Gilbert called it “The Yarn of the Nancy Bell”). I still preserve many of those variations when I recite it today. Nor are any two recitations quite alike: I accidentally dropped two whole stanzas once when I recited it publicly, no doubt contaminating the memories of who knows how many children who were listening at the time.

  106. marie-lucie says

    JC, I wasn’t thinking only of slight changes of words or forgetting a few paragraphs, but retelling stories in order to avoid disgraceful episodes and add more praiseworthy ones, make some characters more or less acceptable, and the like. Political correctness is not restricted to our own time. For instance, in one myth of origin from British Columbia that I know, two brothers are dropped on earth and proceed to explore and delimit the territory which will become that of their future descendants, meeting neighbouring tribes peacefully and acquiring wives from them. So far so good, but one episode involves a much more distant tribe, and some details show that this episode must have been added at a later date, after the distant tribe had expanded its territory towards the one that created the myth in question, and that expansion was not rexactly peaceful.

  107. Witzel’s recent talk about the hydronyms, like Volga or Saraswati, also “found” thousands miles away from their “canonic locations”, and about a large non-IE layer in the Avestan (and non-IE Vedic personal names of kings and poets) (identified through sound-combinations disallowed in these languages’ native words)… is there anything novel and surprising in there, or it just surprises the non-linguist listeners?

  108. David Marjanović says

    Is that talk online somewhere in some form? What you’re saying isn’t new as a short summary; Witzel published this 118-page paper on the subject in 2001. But of course he’s probably found a lot more in the 17 years since then…

  109. he thought the circumpolar cultural sphere might be behind resemblances between certain Algonkian myths and the Grail cycle.

    I would read that retelling. Canoes replacing horses, and Arthur’s warriors paddling forth from his wickiup at Camelot in search of noble adventure?

  110. @David, link gives 404 from here.

  111. Trond Engen says

    We tend to overlook the relative ease of travel and the short distances along the arctic coast. The circumference of the Arctic Circle is 16 000 km, of the 70° line 13 700 km. The distance from the North Cape to the Bering Strait is some 6 600 km — or 2,5 times the seaway along the Norwegian coast.

  112. I only saw cell phone images of the slides of Witzel’s presentation on @amwkim’s twitter, but of course it wasn’t fleshed out in too great detail

  113. David Marjanović says

    The Electronic Journal of Vedic Studies hasn’t published since 2015, so perhaps somebody took all the pdfs down but left the whole misleading site up anyway. Fortunately the pdf is still accessible on Witzel’s own website.

    The distance from the North Cape to the Bering Strait is some 6 600 km —

    Going straight across the North Pole used to be impossible. The Northeast Passage used to be frozen year-round along much of the eastern Siberian part of the coast, too.

  114. The Northwest Passage by sea was not achieved until 1906. Amundsen spent more than two years making the trip, although a lot of the time was spent stationary, learning polar survival skills from the local Inuit. What he learned in the Arctic made it possible for him to make the even rougher journey to the South Pole a few years later.

  115. David Marjanović says

    images of the slides of Witzel’s presentation on @amwkim’s twitter

    Here; recommended. Click on twitpics to embiggen.

  116. Trond Engen says

    Yes. I didn’t mean across the pole. I meant travel on the frozen tundra or along the coast, and ease of it for those used to the environment, and relative to more densely vegetated regions. I think it may have been underappreciated as a corridor for exchange.

  117. David Marjanović says

    Forgot to mention that I love Witzel’s typo of “Sintastha” for Sintashta.

    travel on the frozen tundra

    But why would anyone do that before knowing there’s a place worth going to on the other side?

    or along the coast

    Frozen for a large part of the way.

  118. Trond Engen says

    People do live there. To maintain their existence in a marginal environment they have to travel more than most — exploiting seasonal resources and exchanging their surplus for necessities. It’s not that far. I don’t mean to say that there were people who traveled along the whole Arctic coast all the time, but before the taming of the horse opened the steppe and navigation opened the seas it would have taken fewer steps for an idea or an object or a gene to move between east and west in Eurasia along this route than maybe any other.

  119. Considering that a single language family (Eskimo-Aleut) in found in the North American Arctic, and that another (Uralic) once dominated the Arctic in Eurasia, and since we have established that folktales can spread much more quickly and easily than (proto-)languages, I have no difficulty in believing that there must have been repeated diffusion of folktales across the entire Arctic zone.

    Since some language families represented in or near the Arctic have spread from the Far North to the South (cf. Athabaskan in the Americas), and others from the South to the North (the spread of Algonquian throughout the Canadian Subarctic, from a homeland which was surely located much further South, or the spread of Turkic [Yakut/Dolgan] from Southern to Northern Siberia), we cannot be certain whether folktales/folktale themes common to the Arctic and to various more Southern regions first arose South and spread North and thence (potentially) to the entire Arctic region or vice-versa (i.e. arose somewhere in the Arctic and spread South and/or around the Arctic). Such spreads could have taken place repeatedly, with various survivals of earlier spreads found in areas growing ever-more peripheral and thus unaffected by later spreads: for ought we know perhaps some folktale/folktale theme now found in the North American Arctic only might owe its origin to a spread from Central Asia…much in the same way, for instance, that Greenlandic, the most viable language of its family and likely to be the only surviving Eskimo-Aleutian language in a few generations, is geographically the furthest away from the Eskimo-Aleutian homeland.

    Such acculturation can take place quite rapidly: Folktales of French origin are quite widespread among North American Indians, for instance, and it is telling that in most cases there is among the keepers of this oral tradition zero awareness of the foreign (European) origin of said folktales.

  120. ə de vivre says

    It’s a little off topic, but if there’s a right place to ask these questions, this is it: when did the Aleuts on either side of the Bearing Straight stop having contact with each other? Or did they? The Russian group must be a back-migration, right? Otherwise we’d expect them to be much more linguistically diverse if they’re stay-behinds. Are there estimates on the chronology of the first Eskimo crossing and/or the Aleut re-crossing (if that’s what happened). None of my usual more-or-less superficial internet research has found any explanation about how the Aleuts came to straddle Asia and North America.

  121. Aleuts in Russia were settled in Komandor islands by orders of the Russian-American Company in 1825.

    They were sort of unfree labor back then, so they went where the managers told them to go (in this case, for seal hunting in rich hunting grounds of previously uninhabited Komandor islands).

  122. ə de vivre says

    Er, yeah, I meant Yupik, not Aleut. My mistake. Mutatis mutandis, I have the same quesitons.

  123. a single language family (…) (Uralic) once dominated the Arctic in Eurasia

    Hm, when is “once”, and how far is “Arctic”? The Barents Sea coast was still Paleoeuropean speaking as recently as about 600 years ago, ditto the Finnmark coast about 1500 years ago, and probably no part of the tundra zone east of Taimyr Peninsula has ever been Uralic-speaking. (Yukaghir, rather more likely, but even they have probably come in from the southwest.)

    I did see a lecture by Johanna Nichols just two days ago on how the entire Northern Hemisphere acts as one mega-linguistic area with major east-to-west clines, but even though a route across the Arctic is often the shortest connection, we don’t really know of big spread events all across it. It seems to me that the Taimyr peninsula instead acts as a minor residual zone, with two unconnected spread zones then running to its east and west, both connected to a N-S spread zone bundle running along the Ob, Yenisei and smaller rivers in-between.

  124. David Eddyshaw says

    Otherwise we’d expect them [Siberian Yupik] to be much more linguistically diverse if they’re stay-behinds

    Sirenik, though usually regarded as a Yupik language, was very divergent, and is sometimes classified as a third branch of Eskimo on a level with Yupik and Inuit.

    Chaplinski/St Lawrence Island Yupik is all one language, and the speakers were certainly in regular contact up until Soviet times; mind you, I suppose St Lawrence Island doesn’t really qualify as being on the American side of the Bering Strait. There’s a lot of Chukchee influence, including in the Island language, so the migration there at least must have been from the Asian mainland to the island rather than vice versa (unsurprisingly.)

    (This is cribbed from Steven Jacobson’s grammar of Siberian Yupik.)

  125. Trond Engen says

    j: even though a route across the Arctic is often the shortest connection, we don’t really know of big spread events all across it.

    I’ve spent the better part of the evening trying to retrieve something I read recently about East Siberian material culture and genetic material spreading all the way to Northern Scandinavia. It’s Lamnidis et al: Ancient Fennoscandian genomes reveal origin and spread of Siberian ancestry in Europe (full text available as pdf).

    Abstract:

    European history has been shaped by migrations of people, and their subsequent admixture. Recently, evidence from ancient DNA has brought new insights into migration events that could be linked to the advent of agriculture, and possibly to the spread of Indo-European languages. However, little is known so far about the ancient population history of north-eastern Europe, in particular about populations speaking Uralic languages, such as Finns and Saami. Here we analyse ancient genomic data from 11 individuals from Finland and Northwest Russia. We show that the specific genetic makeup of northern Europe traces back to migrations from Siberia that began at least 3,500 years ago. This ancestry was subsequently admixed into many modern populations in the region, in particular populations speaking Uralic languages today. In addition, we show that ancestors of modern Saami inhabited a larger territory during the Iron Age than today, which adds to historical and linguistic evidence for the population history of Finland.

    From the discussion:

    The large Siberian component in the Bolshoy individuals from the Kola Peninsula provides the earliest direct genetic evidence for an eastern migration into this region. Such contact is well documented in archaeology, with the introduction of asbestos-mixed Lovozero ceramics during the second millenium BC​, and the spread of even-based arrowheads in Lapland from 1,900 BCE​​. Additionally, the nearest counterparts of Vardøy ceramics, appearing in the area around 1,600-1,300 BCE, can be found on the Taymyr peninsula, much further to the east​. Finally, ​the Imiyakhtakhskaya culture from Yakutia spread to the Kola Peninsula during the same period​.

    The Imiyakhtakhskaya (Ymyjakhtakh) culture was a neolithic culture originating in central Yakutia. It developed a characteristic wool-admixed type of ceramics that spread across northern Siberia from Chukotka to Northern Fennoscandia in the 2nd millennium BCE. Too bad it’s too late for Dene-Yeniseian.

  126. David Marjanović says

    What is it, then? The once mighty Yukaghir empire?

  127. Trond Engen says

    Maybe. It’s surely a good geographical match. Also, the bronze age reached the Arctic when Sejma-Turbino prospecters (or people trading with Sejma-Turbino) came down the Yenisei and set up shop in Western Taymyr. Further east an Ymyyakhtakh* settlement also took up Bronze production, and these two seem to have coexisted for a long time. This seems like a good point of entry for old Uralic loans in Yukaghir.

    According to Wikipedia on Ymyyakhtakh, it’s also been associated with the Chukchi and Koryak ethnic groups. If so, we may hypothesize that the core region took up Yukaghir from Sejma-Turbino-connected traders, but the shift never reached the eastern periphery.

    *) English spelling.

  128. Trond Engen says

    Me: Too bad it’s too late for Dene-Yeniseian.

    It’s still intriguing that the paper on Native American origins that Rick linked to in another thread mentions an admixture event from a population related to Koryaks.

    We also show that after 11.5 ka, some of the northern Native American populations received gene flow from a Siberian population most closely related to Koryaks, but not Palaeo-Eskimos, Inuits or Kets, and that Native American gene flow into Inuits was through northern and not southern Native American groups.

    Surely this gene flow must be a result of the event that brought Dene-Yeniseian to America. Except that the source population was not closely related to Ket.

  129. Trond Engen says

    *) English spelling.

    Sometimes my Russian transcriptions mix i, j, and y indiscriminately, and then I try to be consistent. I’m clearly confused.

  130. Trond Engen says

    Me: Surely this gene flow must be a result of the event that brought Dene-Yeniseian to America.

    A very infomative chart of the geneflows. The 12% admixture of Siberian ancestry to Athabascans is shown as quite recent. The same goes for the formation of Inuit as 67% Paleo-Eskimo and 33% Northern Native American.

    The Paleo-Eskimo origin of Inuits is very different from everything I’ve read until now. I don’t know what to make of that.

  131. So Athabaskan- and Eskimoan-speakers were a small elite that passed their language to a lot of locals. Nothing too surprising about that: we wouldn’t expect a huge migration across the Strait.

  132. Trond Engen says

    The Athabaskan speakers look like a small foreign elite. Eskimoans less so. But it’s not easy to understand what’s going on in Eskimoic.

    Here’s a version of the chart with more populations added.

  133. Yukaghirs came to the Arctic from the Urals or somewhere thereabouts in the 2nd millenium BC.

    They weren’t quite Bronze Age warriors, but came in close contact with them (perhaps Yukaghirs were chased by Bronze Age peoples into the Arctic).

    People displaced by the Yukaghirs (or maybe another group which migrated with them) crossed the Bering strait and are likely to have been ancestors of the Na-Dene people.

    That’s the summary of Russian research on the topic I’ve read somewhere on the net.

  134. in the 2nd millennium BCE. Too bad it’s too late for Dene-Yeniseian.

    If we are talking about crossing Bering strait, it’s just proto-Na-Dene, not Dene-Yeniseian.

    Unless you believe that the Kets crossed the Bering strait in opposite direction.

  135. Kets going the opposite direction are called bras (according to Dirac).

  136. Trond Engen says

    Me: The Paleo-Eskimo origin of Inuits is very different from everything I’ve read until now. I don’t know what to make of that.

    Me, again: But it’s not easy to understand what’s going on in Eskimoic.

    A 2014 paper from the Copenhagen lab: Raghavan et al.: The genetic prehistory of the New World Arctic.

    Abstract

    The New World Arctic, the last region of the Americas to be populated by humans, has a relatively well-researched archaeology, but an understanding of its genetic history is lacking. We present genome-wide sequence data from ancient and present-day humans from Greenland, Arctic Canada, Alaska, Aleutian Islands, and Siberia. We show that Paleo-Eskimos (~3000 BCE to 1300 CE) represent a migration pulse into the Americas independent of both Native American and Inuit expansions. Furthermore, the genetic continuity characterizing the Paleo-Eskimo period was interrupted by the arrival of a new population, representing the ancestors of present-day Inuit, with evidence of past gene flow between these lineages. Despite periodic abandonment of major Arctic regions, a single Paleo-Eskimo metapopulation likely survived in near-isolation for more than 4000 years, only to vanish around 700 years ago.

    This summary of archaeological consensus is handy for context:

    Humans first peopled the North American Arctic (northern Alaska, Canada, and Greenland) from the Bering Strait region beginning around 6000 years before the present, leaving behind a complex archaeological record. Over successive millennia, the pioneering Arctic cultures developed into distinct lifestyles and cultural stages grouped within two broad cultural traditions known as Paleo-Eskimo and Neo-Eskimo. Early Paleo-Eskimo people representing the Denbigh, Pre-Dorset, Independence I, and Saqqaq cultures (~3000 to 800 BCE) lived in tent camps and hunted caribou, musk ox, and seals with exquisitely flaked stone tools similar to those used by northeast Siberian Neolithic cultures. In northern Alaska, the Denbigh cultural groups were succeeded by the Paleo-Eskimo Choris and Norton cultures starting around 900 BCE, with the Norton material culture further developing into the Ipiutak culture around 200 CE. Simultaneously, during the cold period beginning around 800 BCE, innovations in housing and hunting technologies accompanied the formation of the Late Paleo-Eskimo or Dorset culture in eastern Arctic (eastern Canadian Arctic and Greenland), with population growth and more intensive use of marine mammals, including walrus. The Dorset culture is divided into three phases: (i) Early Dorset, ~800 BCE to 0 BCE/CE; (ii) Middle Dorset, ~0 BCE/CE to 600 to 800 CE; and (iii) Late Dorset ~600 to 800 CE to 1300 CE. The Paleo-Eskimo tradition in the eastern Arctic ended sometime between 1150 to 1350 CE, shortly after the sudden appearance of the Neo-Eskimo Thule whale-hunters from the Bering Strait region.

    The Siberian Old Bering Sea culture is the earliest expression of the Neo-Eskimo tradition ~2200 years before the present, developing into the Punuk culture around the sixth century CE. Almost concurrently, the Old Bering Sea culture developed into the Birnirk culture in the northern parts of the Bering Strait region. Interactions between people of the Birnirk and Punuk cultures gave rise to the western Thule culture on both sides of the Bering Strait, with contribution from the Paleo-Eskimo Ipiutak culture in Alaska By the early second millennium CE, western Thule cultural groups began their movement into the eastern North American Arctic. With the Thule culture came more effective means of transportation like dog sleds and large skin boats, complex tool kits like sinew-backed bows, and harpoon float gear for hunting large whales. Thule culture spread quickly throughout the eastern Arctic, rapidly replacing Dorset in most, if not all, regions. […]

    On the genetics I can’t find any short passages that easily lend themselves to be used as summaries, but the essence is that all Paleo-Eskimos were one close and continuous population for its whole 4000 year existence, until it was completely replaced by Neo-Eskimos in the early 2nd millennium CE. Admixture with the Neo-Eskimo lineage would have occured before both entered the new world. There was little evidence of admixture with Athabascans or other Native Americans. But all these lineages are more related to eachother than to any other except Eastern Siberians.

    And from the discussion, an expanded version of the abstract:

    We have shown that Paleo-Eskimos likely represent a single migration pulse into North America from Siberia; separate from the migration events giving rise to Native Americans and Inuit. However, while being genetically distinct from other New World populations, Paleo-Eskimos are still more closely related to these populations than to non–New World populations, which is in agreement with a single ancestral population giving rise to many subpopulations and possibly many migration pulses into the Americas, as suggested by the Beringian standstill model and a three-stage colonization model. Moreover, although our data are in agreement with Reich et al., we find no support for Saqqaq or the rest of the Paleo-Eskimo tradition being a part of one of the two waves of Native American ancestors entering the more southern regions of the Americas. Therefore, an additional Paleo-Eskimo migration wave should be added to the three-wave hypothesis in explaining the peopling of the Americas.

    So there seems to have been a major reinterpretation of the origin of Neo-Eskimos and the contact with Native Americans. Now we have a very interesting pair of crossing geneflows, with Athabascan being formed when a branch of Northern NA takes up genes from a lineage ancestral to Paleo-Eskimo, and Neo-Eskimo being formed when Paleo-Eskimo takes up genes from Northern NA. Is this simpler to explain as a single disruptive event? Northern Native Americans having back-migrated to Alaska are being taken over by migrants from across the Strait. The result is an immediate and irreparable split of the community. In one part the migrants make themselves the ruling elite. This group ends up inland fishing and hunting caribous. The other part ends up at the coast (e.g. represented by the strangely wood-dependent Ipiutak culture), eventually joining the Paleo-Eskimos to form the Neo-Eskimos.

  137. Trond Engen says

    Brett: Kets going the opposite direction are called bras (according to Dirac).

    Hah. This made me go reading. Dirac’s notation was adapted from Grassmann, so it’s linguistically sound.

  138. marie-lucie says

    TE: The Athabaskan speakers look like a small foreign elite

    A small foreign elite, with a morphologically very complex language, able to pass it on over thousands of miles East-West and North-South?

    it’s not easy to understand what’s going on in Eskimoic

    I don’t know much about the family, but it may be relevant that Eskimos were able to travel in both directions in the winters until the Bering Strait was closed to them during the Cold War.

  139. ə de vivre says

    I don’t know much about the family, but it may be relevant that Eskimos were able to travel in both directions in the winters until the Bering Strait was closed to them during the Cold War.

    This was my question. Was there unbroken cross–Bering Strait contact before European sailors came along? Is there evidence of that cross contact? When/why did that contact stop being a migration that brought new languages/DNA in one direction or the other.

  140. Trond Engen says

    marie-lucie: A small foreign elite, with a morphologically very complex language, able to pass it on over thousands of miles East-West and North-South?

    That occurred to me too, but

    1. Proto- (or Pre-Proto-)Na-Dene, while morphologically complex, was at least much more regular at the time it acquired new adult speakers.

    2. A great part of the spread may have been by filling up practically empty space in the northern forests.

    3. Non-speakers being taken up in the community may not become fluent, but their children will grow up as native speakers.

    I have a follow-up comment on Eskimoan in moderation.

  141. Imiyakhtakhskaya culture (…) spread across northern Siberia from Chukotka to Northern Fennoscandia in the 2nd millennium BCE

    OK, color me corrected. At least assuming that this really was a culture in the, well, cultural sense and not merely a popular type of pots that got traded all across the Arctic.

    There may be possible common substrate vocabulary in Sami + Nenets (in a few cases also in other Samoyedic or in Ob-Ugric), but this generally doesn’t have parallels in Yukaghir that I know of.

    “Siberian admixture in Uralian speakers” is still a different issue though. There’s every reason to think that this spread west along the usual more southern forest-steppe route and north only secondarily, for starters because it also turns up in the more southern Leväluhta discoveries in central Finland (possibly Sami speakers, possibly Paleo-European, probably not Finnic), and even modern Finns. So it has nothing to do with the geographically short but culturally inconvenient route along the Arctic.

    Here’s a version of the chart with more populations added.

    Mmmm, those are very small n-values. And then you have risks such as that a single East Asian great-grandparent, perhaps tagging along with colonial Russians, would be sufficient to introduce the entirety of 12% East Asian ancestry in the modern Athabaskans.

    But it’s not easy to understand what’s going on in Eskimoic.

    Surely we should make this Eskaleut in general to have a good resolution on the situation?

  142. marie-lucie says

    ə de vivre: The fall of the Soviet Union caused some Eskimos on either side of the Strait to be reunited with friends and relatives on the other side. At the other end of North America, there are Eskimos in both Canada and Greenland (the latter being geographically North American).

  143. marie-lucie says

    Trond: What could have been the superior cultural element(s) brought by the Athabaskans?

  144. ə de vivre says

    marie-lucie: So there was no cross–Bering Strait contact after the initial Neo-Eskimo migration until the arrival of Europeans in the area?

  145. Difficult or not, the Navajo language was successfully imposed on many non-Navajos, as is shown by the Navajo clan names. The Mexican Clan (Naakai dine’é) still remember that their ancestress (clans are matrilineal) was a Mexican slave woman. The same may well have been true of the remote ancestor of Navajo.

  146. marie-lucie says

    ə de vivre: .So there was no cross–Bering Strait contact after the initial Neo-Eskimo migration until the arrival of Europeans in the area?

    That is not at all what I said. In the winter the ice on the Strait enabled people on each side to cross over to the other side as well as meet in the middle (contact could also be achieved by boat in the summer). The Soviet regime put a stop to those human contacts, which have now resumed.

    What probably happened is that people in various areas of the High Arctic had settled several large territories where they could hunt and fish and find other resources, and they did not necessarily travel much outside of these territories (hence the existence of several Eskimo dialects or even languages). But one of those territories was the Strait and its shores, along both of which Yupik Eskimo is/was spoken.

  147. Trond Engen says

    j: OK, color me corrected. At least assuming that this really was a culture in the, well, cultural sense and not merely a popular type of pots that got traded all across the Arctic.

    I thought the same thing. It’s clearly a culture east of Taimyr/Yenisei, roughly modern-day Yakutia. It’s less clear further east and west. But either way it does show “horizontal” contact in the circumpolar region.

    “Siberian admixture in Uralian speakers” is still a different issue though. There’s every reason to think that this spread west along the usual more southern forest-steppe route and north only secondarily, for starters because it also turns up in the more southern Leväluhta discoveries in central Finland (possibly Sami speakers, possibly Paleo-European, probably not Finnic), and even modern Finns. So it has nothing to do with the geographically short but culturally inconvenient route along the Arctic.

    You obviously know more about this than me. But if I remember correctly the authors suggest a combination of both — a rapid western spread along the Tundra, but also southward spread along the usual north-south routes contributing to the Uralic Bronze Age cultures. But the presence in Northern Fennoscandia is too strong for that to be the only route.

    Mmmm, those are very small n-values. And then you have risks such as that a single East Asian great-grandparent, perhaps tagging along with colonial Russians, would be sufficient to introduce the entirety of 12% East Asian ancestry in the modern Athabaskans.

    Agreed. I was going to write something about the inferred tree possibly being sensitive to slight changes in the input.

    Surely we should make this Eskaleut in general to have a good resolution on the situation?

    Yes. I don’t know what’s going on there either…

  148. ə de vivre says

    marie-lucie: I was confused because I wasn’t sure what part of my questions you were responding to. I wasn’t asking about modern cross-Bering movement, and I’m aware of the distribution of Eskimo languages. What I’m curious about is the chronology of cross-Bering movement after the initial Neo-Eskimo migration and before the arrival of Europeans. Am I understanding you correctly that you’re saying, yes there is evidence of (more of less) continuous cross-Bering movement during that time?

  149. Trond Engen says

    marie-lucie: Trond: What could have been the superior cultural element(s) brought by the Athabaskans?

    I don’t know. If it’s material culture, something that made it possible to increase production inland like willow bark fishing nets? Or it could be social culture. A clan system that drew people into the cultural sphere and kept them there?

    (My comment on Eskimoan is out of moderation now.)

  150. Trond Engen says

    ə de vivre: What I’m curious about is the chronology of cross-Bering movement after the initial Neo-Eskimo migration and before the arrival of Europeans. Am I understanding you correctly that you’re saying, yes there is evidence of (more of less) continuous cross-Bering movement during that time?

    There clearly has been cross-Beringian contact for as long as Eskimoan groups have lived on both sides, but it won’t show up in the study without recent genomes from the contact zone. That may seem like a weakness but may well have been intended. Using Greenlanders they all but eliminate later Eurasian and Athabascan admixture and get a clearer view at the stage immediately before the Eskimoan expansion.

  151. j.: There may be possible common substrate vocabulary in Sami + Nenets (in a few cases also in other Samoyedic or in Ob-Ugric), but this generally doesn’t have parallels in Yukaghir that I know of.

    Would you expand on that?

  152. ə de vivre says

    There clearly has been cross-Beringian contact for as long as Eskimoan groups have lived on both sides

    Why clearly? I’m being maybe annoyingly persistent on this point, but I’m curious: if there was continuous cross-Bering movement why didn’t more East Siberian material culture make its way into the Americas. Did the Siberian Yupik jealously guard the secret of another continent (not implausible)?

  153. Using Greenlanders they all but eliminate later Eurasian and Athabascan admixture and get a clearer view at the stage immediately before the Eskimoan expansion.

    Why so? Is it shown beyond all doubt that Greenlanders did not pick up Athabaskan ancestors before occupying Greenland?

  154. Trond Engen says

    Why clearly?

    It’s been my impression that the intermediate position of Naukan between Siberian and Central Alaskan Yupik points to a less than clear break. But I may be overstating my case.

  155. ə de vivre says

    Is the consensus that Siberian (in a strictly geographic sense) Yupik is a back migration? Otherwise it seems plausible that the proto–Central Alaskan end of the dialect continuum was the one that happened to migrate across the strait, leaving its closer (Naukan) and farther (Siberian) relatives behind, never to be heard from again?

    Or more succinctly: why is a fuzzy break the best explanation of Naukan’s intermediate position?

  156. Trond Engen says

    Why so? Is it shown beyond all doubt that Greenlanders did not pick up Athabaskan ancestors before occupying Greenland?

    I expected that one. No, I don’t think so, but they may still have wanted to avoid the effect of later admixture in the Beringian contact zone.

  157. Trond Engen says

    why is a fuzzy break the best explanation of Naukan’s intermediate position?

    I’ll have to pass on that one. Some prefer an Eskimo-Aleut homeland in Siberia, some in Alaska. Some see Siberian Yupik as a stay-behind. Others, I believe, see Sireniki as a stay-behind and the other Siberian Yupik languages as later back-migrations. I don’t even know if the interpretation Naukan would be clearer, or any different, in any of those cases.

  158. David Marjanović says

    Surely we should make this Eskaleut in general to have a good resolution on the situation?

    Yes. I don’t know what’s going on there either…

    Nobody does.

  159. Trond Engen says

    A very useful summary. I knew Aleutian was messy, but not that it’s part of the trading & raiding sprachbund of the Pacific Northwest. In that case I doubt that genetics will be much help.

  160. Would you expand on that?

    Sure. So, the fact that the Sami have shifted their language from something else to the current Finno-Ugric fare has been long obvious; back in the 20s Y. H. Toivonen devised a theory that, since their endonym sounds similar to “Samoyed”, the substrate should have been Samoyedic, who were at the time assumed to have been “roaming” all across the European & West Siberian Arctic for thousands of years; he datamined a bunch of lexical parallels between Sami varieties and mostly Nenets in support. By now we know though that the Sami were linguistically distinct already on much southern latitudes, and that the Nenets expansion west is somewhat recent (closer to 500 than 1000 years ago), so this theory is firmly out of the question.

    The data behind this has so far only received one detailed later review: “Protolappisch und Samojedisch”, Eugen Helimski 1994. He concludes the data divides as follows:
    – 25 etymologies are discardable as mere superficial similarities;
    – 13 are common Proto-Uralic inheritance;
    – 10 are up interpretation as either of the above two;
    – 11 could be analyzed as late loanwords of some sort.

    It’s this last group I’m talking about. It includes mainly words limited to Eastern Sami which seem likely to be recent loans from Nenets (e.g. Kildin jurka ‘corral for reindeer’), or words limited to Nenets which seem likely to be recent loans from Eastern Sami (e.g. jāmću ‘lasso’). But at least one looks like a candidate for substrate origin: a group of terms for polar duck or fowl species, which UEW reconstructs as *aŋɜ, with alleged reflexes also from Mansi and Khanty. These clearly cannot be all straightforwardly cognate, but unlike Helimski I would not dismiss entirely some more roundabout connection. A few items form Toivonen are also comparisons such as Sami-Enets or Sami-Selkup, where routing would be more difficult.

    It’s hard to go much further than this offhand. There is very little research so far on possible early loan contacts between Ob-Ugric and Samoyedic, or on the Western Siberian substrate in Ob-Ugric. (The latter is fairly obvious IMO if we allow for the possibility of substrates at all; but in the Hungarian research tradition no such tool exists and the material is treated simply as “common lexical innovations in Proto-Ob-Ugric”. I am hoping to contribute something to this some years in the future.)

    (I have no idea what to do with the fact that there are also Altaic parallels for *aŋɜ.)

  161. David Marjanović says

    Guide for the Perplexed: Uralistic ɜ is not IPA ɜ, it just means “any vowel”.

  162. Trond Engen says

    There should be DNA from the Nenets substratum coming in soon. Or maybe not. Shallow tundra graves aren’t the best environment for preserving ancient DNA. That’s why the studies of ancient Americans are thin on Paleo-Eskimos.

  163. j.: From what you are saying, especially considering what you call Altaic parallels, *aŋɜ might be an Wanderwort, perhaps with something sound symbolic in it; those are tough to trace down. Could PIE *h₂enh₂-ti- ‘duck’ also be somehow connected?

  164. David Marjanović says

    I’m sure I recently read that this has been proposed, but I can’t find where; it’s not in the obvious-seeming place.

  165. David Marjanović says

    Found it – it’s on the last page of a report from a conference where Hyllested was present, but it’s not by him. I quoted the paragraph here, but the spam filter doesn’t let it through because it’s in Russian.

    In short, nasal + laryngeal clusters in PIE are said to correspond to */ŋ/ in Uralic (2 examples given) and Altaic (1 Proto-Tungusic and 1 Proto-Turkic example, the latter being *aŋɨt corresponding to PIE “*h₂enh₂t-“). Makes sense (as inheritance or as a loan substitution) if *h₂ was [χ] or thereabouts and preceding nasals assimilated by becoming [ɴ].

  166. A sign of the times?

    “the spam filter doesn’t let it through because it’s in Russian.”

  167. David Marjanović says

    No, I don’t think this is part of the Warmed-Up War. The spam filter has learned from experience that comments consisting mostly of non-Latin characters (it’s not limited to Cyrillic) tend to be spam, and so it overgeneralizes.

  168. Stu Clayton says

    Then it’s high time to disown that experience and generality. It would suffice to verify that all non-Latin characters in a comment belong to a Russian codepage (there seem to be several, some of great antiquity).

  169. I was provoked to look into this the other week — the Akismet spam recognition service can in fact be told to expect alternative languages such as Russian in the comments, but the plugin available for WordPress cannot be configured to give it that information. Someone could enter a feature request with the plugin author, of course.

  170. Stu Clayton says

    I just posted this in the WordPress Akismet plugin forum:

    # A linguistic website on wordpress that I visit uses the Akismet plugin. For over half a year now, unfortunately, comments posted in Russian are being rejected as spam.

    Another site visitor did some research and told us this: “the Akismet spam recognition service can in fact be told to expect alternative languages such as Russian in the comments, but the plugin available for WordPress cannot be configured to give it that information.”

    Would it be possible to extend the plugin to expose the configuration API of Akismet ? #

  171. Thanks! Let us know if you get any useful info.

  172. Stu Clayton says

    Here’s how simple it all is: specify a list of approved codepages in the “blog_lang” parameter of the comment_check call:

    blog_lang
    Indicates the language(s) in use on the blog or site, in ISO 639-1 format, comma-separated. A site with articles in English and French might use “en, fr_ca”.

    API documentation

    The standard list contains only the value of getLocale(), as can be seen in class.akismet.php. It seems to me that all Songdog would need to do is modify this class appropriately on your server. You don’t need a fancy API to configure this.

  173. Great, I’ll alert him accordingly.

  174. Trond Engen says

    Remember adding Dyirbal, Cree, Romanized Egyptian Arabic, Kildin Saami, and Hieroglyphic Luwian.

  175. Could PIE *h₂enh₂-ti- ‘duck’ also be somehow connected?

    There was a recent lively session at Academia.edu on Indo-Uralic lexical parallels. This was indeed brought up, and one opinion was that names for geese of the shape KANK or HANH might very well be onomatopoetic. I’m not sure if this generalises to all waterfowl though, and certainly not for the likes of tarmigans*.

    *omission of gratuitous pt- intentional.

  176. You don’t need a fancy API to configure this — no, but such changes will be overwritten when the plugin is updated, and being able to update plugins in a timely manner is essential to security these days.

    The better class of plugins will define a hook so you can register a local function to modify such defaults, but no such luck here. There might be a hook defined at a lower level that can intercept all HTTP callouts, but then you’re modifying raw URLs and not having fun either.

    Having been in charge of a WordPress site once I will say that this is the sort of thing you do if someone pays you good money, and not just to help out a friend.

    That said, if you are actually in a situation where maintaining a parallel version of the Akismet plugin makes sense, it’s not much extra work to add an extra field in the admin GUI (basically cut and paste) and I’d definitely do that instead of having constant strings in my patch.

  177. There was a recent lively session at Academia.edu on Indo-Uralic lexical parallels.

    I wish I’d at least listened in on that!

    names for geese of the shape KANK or HANH might very well be onomatopoetic.

    If that were so, you’d expect words like that to be more widespread. In California, the area I’m most familiar with, there are hardly any ‘goose’ words of that form. The possible velarization of *n under the influence of the following *h₂, which David mentioned, would show at least some period of regular development.

  178. Stu Clayton says

    @Lars (the original one)

    Yes, patching is not a good way to go, but “paying good money” is another undesirable extreme, absent more information. I asked about the matter at the forum to elicit initial responses and thus get a feel for “their” attitudes.

    I’m not even sure that this blog_lang parameter can be used to avoid the Russian-suppressing behavior of Akismet. The API doc seems to suggest it, but I’m speculating.

    A list of locale tags is not something that cries out for frequent modification via an admin UI. Even if you want such a UI, it doesn’t have to be an extension of the plugin UI. Any UI technology (including a simple text editor) that produces a file with a list of tags will do, provided there is a way to get the plugin to read it in – in other words, a hook for reading a config file. The other settings must be stored already in some form locally on the server.

    These questions are essentially design probes, not gun-jumping “how can I hack this” thoughts. Whether they are “good money” questions remains to be seen. It can’t hurt to try out whether the blog_lang parameter has the effect I am guessing at.

    A large part of how I earn money is by deconstructing developer claims that they can’t correct their code because it would require too much work, they don’t have the budget etc. I do this by writing a few lines of alternative code.

  179. Lars (the original one) says

    @Stu, the problem is not to fix the code, because that is easy, it’s that we as outsiders cannot impose on Hat or Songdog to implement a less maintainable solution than they have now. If it were my blog it would have been fixed a while ago, but very different tradeoffs apply when I’m both the tech nerd in charge and the person aggravated by suboptimal functionality.

    The world is full of things that could be improved, just volunteer at your nearest open source project 🙂

  180. Stu Clayton says

    No question of imposing ! Let’s wait for someone to respond at the forum.

    “Open source” means “enjoy, it’s free!”. It’s a gateway drug. How long will free code be maintained for free ? I suspect there are few large companies that did not become addicted to it over the past 10 years. Now the day of security-hole reckoning has come upon them, and they don’t know what to do. Their dealers have moved on to more fun things.

  181. Lars (the original one) says

    That’s why there’s room for new people to have fun fixing those security holes. But yes, if you aren’t prepared to help maintain an open source project you shouldn’t build your business on it.

  182. Stu Clayton says

    Too late for shouldn’t, in many cases. The problem is now one of dependency bulk and sheer business size, not of will-power and big pocket money.

    Consider a bank system with 200-plus Java application projects, each of which over the years made itself dependent on dozens of open-source jars. In the system overall there are tens of versions of hundreds of jars, some of them dependent on specific versions of others (I mention only Struts, Spring and Log4J).

    Then consider similar approaches in the newer Javascript applications.

    There is no one “open source project”. There were dozens and dozens of them, many of which just stopped existing. Most people involved in these projects tended to be unconcerned with other projects, and usually failed to learn from the others. It was not so much about learning, but about fun.

    Years ago there may have been conscious decisions at management level to in some sense “build their business” on open-source. But consciousness is compatible with starry-eyed ignorance (the phrase “the human condition” springs to mind here). Also, whether or not there were such decisions, the individual application projects were staffed with young programmers who rushed into jar addiction anyway, because internet. I remember once, from a code review, an application dependent on an 800 KB jar of which exactly one class and one method was used. The method String-formatted a long.

    languagehat is dependent primarily on WordPress. Maintaining a single dependency, or trading it in for another, has costs and benefits that are easier to estimate.

  183. Lars (the original one) says

    Stu, we are in violent agreement.

  184. Stu Clayton says

    Shun fun, embrace
    The chasteness of a nun !

  185. But yes, if you aren’t prepared to help maintain an open source project you shouldn’t build your business on it.

    That seems an extreme position. I once worked for a company that depended on FreeBSD to run its servers, without the slightest intention (or indeed possibility) of contributing anything to even the applications, never mind the kernel, though they depended on the latter even more than the former. I know they weren’t alone in this respect.

    (I don’t mention Linux in this connection because many businesses do help maintain it indirectly — by paying money to Red Hat, Oracle, SUSE, or Canonical, to say nothing of Intel and IBM. There is no analogue to this in the *BSD world.)

  186. Lars (the original one) says

    Yes, well, the *BSD’s and the GNU project are sort of too big to fail, so you can piggyback on them for free. Though you still don’t have any guarantee that the feature you need will be there in the next release.

    But server OS’es can be replaced, though maybe not conveniently — by building your business on an open source project I was thinking of things like selling a webmail service that runs on Hula.

  187. Yeah, that would suck. Although Hula was always an over-the-transom-in-reverse open-source project, and those often grind to a halt.

    Stu: Which nun?

  188. Stu Clayton says

    “A nun” in this usage means “given any nun”. It’s no accident that “A” is used for the universal quantifier. It’s upside-down merely to protect the innocent.

  189. Stu Clayton says

    Here is my WP forum question. One of the developers has already responded !

  190. Not very helpfully, but thanks for bringing this to their attention!

  191. Stu Clayton says

    I think he merely didn’t understand the issue. That’s one reason why I formulated it again in my response, in more detail than originally. I’m hoping he’ll speak to another developer, especially if he feels out of his depth. A sign of that is his remark: “you did mention a linguistic website, which might get more complex, if the same site sports countless languages.” Countless languages ! Sports !

  192. According to Wikipedia on Ymyyakhtakh, it’s also been associated with the Chukchi and Koryak ethnic groups. If so, we may hypothesize that the core region took up Yukaghir from Sejma-Turbino-connected traders, but the shift never reached the eastern periphery.

    A new DNA paper probes an early Bronze Age (9.8 kya) individual from Duvanny Yar in the Northern fringes of the Kolyma Basin, and indeed shows that it’s highly similar to the contemporary Koryak, Itelmen, and Chukchi, as well as to late Bronze Age individual from the shores of Magadan. This stream of ancestry seems to have been quite widespread after the Last Glacial Maximum, reaching West and Southwest towards Altai and Ural mountains (and especially contributing to the present-day Ket and Na-Dene peoples), but it has been pushed to the very NE fringes of Asia by the waves of “Neo-Siberians” of largely East Asian DNA makeup in the rest two millennia.
    https://www.nature.com/articles/s41586-019-1279-z

  193. Trond Engen says

    A draft version from last year on Biorxiv. I think we discussed it back then.

    The Ancient North Siberians are most closely related to Western Hunter Gatherers, indicating a westerly route of migration. THey must have been all over the northern continent before the Last Glacial Maximum forced them out of the northern parts of their range.

    The Ancient Paleo-Siberians came north from East Asia, admixed with the ANS in a complex way, and became the main contributors to Native Americans and some peripheral Modern Siberian populations as well as extinct populations like Paleo-Eskimos and Ancient Beringians. Both the admixture events and the subsequent splits are dated to 20 kya, so it must have been a dynamic time. The Eskimo and Na-Dene have more AP ancestry than other Native American groups. evidence of later contact across the Bering Strait, but the the time depths and the source populations on both sides are different. The Kets are among the Siberian populations with much AP ancestry, but it’s specifically linked to Paleo-Eskimos, so I don’t see any clear genetic link with Na-Dene. However, the relation to Paleo-Eskimos might indicate that the AP ancestors of the Kets were located further north and east. The Paleo-Eskimos in turn had more East Asian ancestry than any of the other groups, Messy, but the patterns will probably become clearer with more data.

    Neo-Siberians are a later incoming population from East Asia. They are the main contributors to most Modern Siberians, though to a different degree.

    I don’t know how the Ymyyakhtakh fits in all this. They are certainly later.

  194. David Marjanović says

    Both the admixture events and the subsequent splits are dated to 20 kya, so it must have been a dynamic time.

    A glacial maximum will do that to people, I guess.

  195. Trond Engen says

    Here’s Flegontov et al (2016): Na-Dene populations descend from the Paleo-Eskimo migration into America, a paper on the genetic relation between Paleo-Eskimos and Na-Dene that I must have overlooked last time this was up. I can't evaluate the genetics, but the study uses rare alleles to demonstrate a closer relation between Paleo-Eskimos and Na-Dene peoples than what can be gleaned from the whole-genome approach. I'll quote the discussion in extenso;

    Discussion

    Our results are consistent with a gene flow from the Saqqaq Paleo-Eskimos (19 – 25% admixture ratio) exclusively into the First American ancestors of Na-Dene, and a much later and less extensive bidirectional gene flow was detected between the Na-Dene and Neo-Eskimo branches. A somewhat lower level of Saqqaq-related ancestry of 16% was reported using the admixture graph method in Chipewyans, a Northern Athabaskan ethnic group. We emphasize that only a fraction of modern Na-Dene individuals displays this level of Saqqaq ancestry, with most Na-Dene being admixed with other native groups and/or Europeans.
    Methods of rare allele and autosomal haplotype analysis are especially sensitive for reconstructing recent population history within a few thousand years, and in some cases were demonstrated to outperform traditional methods based on unlinked common genetic variants, such as ADMIXTURE, PCA, TreeMix, f3-, f4-, and D-statistics. In this light, we consider discrepancies between our results and those of previous studies in Suppl. Text 2.

    We dated the Paleo-Eskimo admixture at about 3,600 YBP using GLOBETROTTER analysis based on autosomal haplotypes (Table 2). Much older dates of the Siberian-Athabaskan gene flow obtained in our analysis based on rare allele sharing, about 6,500-7,000 YBP (Fig. 5, Table 1), probably correspond not to the admixture of Paleo-Eskimos and First Americans (that must postdate the Paleo-Eskimo immigration), but to a time point when Paleo-Eskimo ancestors branched off from the Siberian-Arctic stem. The split date suggested here for this unsampled “ghost population” fits the archaeological record of Siberia remarkably well, as discussed below. And split dates for other nodes inferred here broadly agree with the dates produced with independent methods.

    The new wave of population from northeastern Asia that arrived in Alaska at least 4,800 years ago displays clear archaeological precedents leading back to Central Siberia. The rise of the Syalakh culture that flourished across much of Northeastern Siberia between 6,500 and 5,200 YBP involved migrants from the Transbaikal area who possibly mixed with local remnants of the earlier Sumnagin culture (10,500-6,500 BP), bringing the bow and arrow and new types of pottery to Northeastern Siberia. As the Bel’kachi culture (5,200-4,100 YBP) developed from Syalakh along the Lena and Aldan rivers, at least one group of these people might have crossed the Bering Strait into Alaska around 4,800 YBP, giving rise to Paleo-Eskimos. Thus, the Syalakh culture peoples, spreading across Siberia after 6,500 YBP, might represent the “ghost population” that split off around 6,500-7,000 YBP and later gave rise to migrants into America.

    The geographic connection between Paleo-Eskimos and the related Siberian groups probably became severed as subsequent waves of hunter-gatherers entering Eastern Siberia from the west during the Late Neolithic (Ymyakhtakh culture, 3,700-2,800 YBP) brought new cultures and new language groups. This phase of North Asian prehistory most likely involved the spread of Yukaghir, Chukchi-Kamchatkan and Eskimo-Aleut languages, whose presence in the extreme northeast of Asia intervened geographically between Paleo-Eskimos, Na-Dene, and their Old World cousins. Notably, the dates of the Siberian-Arctic split obtained under our model (~4,000-4,200 YBP, Table 2) also agree with this scenario that links the spread of the Ymyakhtakh culture (after 3,700 YBP) with the Arctic meta-population, i.e. ancestors of modern Chukchi-Kamchatkan and Eskimo-Aleut ethnic groups.

    The success of Paleo-Eskimos and Na-Dene in occupying territories previously populated by First Americans, in some cases (Southern Athabaskans) moving very far from the original homeland in Alaska and northwestern Canada, might be partially attributed to archery, a technological advance lacking among the local populations. Paleo-Eskimos quickly spread from Alaska to Greenland and Labrador and have been credited with introducing the bow and arrow to populations in Eastern Canada by 4,000 YPB, though the Dorset people, the last wave of Paleo-Eskimos, seem to have given up this technology for handheld lances.

    Another important observation concerns the distribution of Siberian (Paleo-Eskimo) ancestry among modern North Americans. The methods used in this study detected Central and West Siberian ancestry in a fraction of Na-Dene individuals belonging to all major branches of the language family existing today: Tlingit, Northern Athabaskan (Chipewyans, Dakelh, etc.) and Southern Athabaskan. Importantly, the Central and West Siberian ancestry is almost exclusive to Na-Dene, and missing in other North or South American native ethnic groups, including Haida, a group previously considered a divergent member of the Na-Dene language family. Thus, the current consensus view of the Na-Dene language family and the distribution of recent Siberian ancestry match remarkably well. Although the small population sizes do not allow statistically valid comparisons, individuals with noticeable Saqqaq ancestry are likely more frequent among Northern Athabaskans as compared to Tlingit and Southern Athabaskans, the latter being mixed with southern Native Americans (Suppl. Fig. 1B).

    We speculate that a migrating population, starting from Siberia around 6,500 YBP (the Syalakh culture), entering the New World around 4,800 YBP, and later mixing with First Americans, might have carried the Dene-Yeniseian languages into North America. This hypothetical language macrofamily unites multiple Na-Dene languages and Ket, the only surviving remnant of the Yeniseian family, once widespread in South and Central Siberia. For a further description of the Dene-Yeniseian hypothesis and a review of lexicostatistical dating estimates see Suppl. Text 3. Although the Dene-Yeniseian macrofamily is not universally accepted among historical linguists (cf. Hamp), and correlation of linguistic and genetic history is far from universal, the existence of the exclusive Siberian-Dene gene flow makes a genealogical relationship of the language families, either as the closest sister-groups or within a wider clade, an attractive area of future research. Inferred age of the gene flow, 6,500-7,000 YBP, possibly corresponding to the split of Na-Dene and Yeniseian precursors in Siberia, is comparable to the age of the classic Indo-European language family, suggesting that investigation of the Dene-Yeniseian connection lies within the reach of current methods in historical linguistics.

    I gather that the admixture came from a limited part of the Paleo-Eskimo population, i.e. somewhat divergent from Saqqaq, and into an American Indian population located in the far northwest of North America. Since it’s only a partial ancestry detected in only a fraction of the population, it looks very much like an incoming elite (and it would be fun to see if there are correlations with the clan system). The main difference from Raghavan et al (2014) has to do with the timing and the relation to Neo-Eskimos.

  196. John Cowan says

    So we can now construct a deeply flawed theory (to be sent posthaste to Nature) that the Palaeo-Eskimo spoke Proto-Dene-Yeniseian.

  197. John Cowan says

    And in it, we have lived since then.

    We provincials tend to think that the Columbian Exchange was a pretty important historical event even at the big scale. Maize in Africa, cucurbits in Australia, wheat on the Great Plains, and those stupid starlings from a deeply stupid attempt to introduce all the birds mentioned in Shakespeare.

  198. Starlings are objectively the worst birds.

  199. January First-of-May says

    and those stupid starlings from a deeply stupid attempt to introduce all the birds mentioned in Shakespeare

    This is one of those real stories that sounds like it belongs in a comedic historical fiction novel.

  200. Gary Vellenzer says

    I understand they tried to introduce some of the same birds into Australia, but the birds insisted on flying south for the winter.

  201. Stu Clayton says

    Birdbrained.

  202. Starlings are objectively the worst birds.

    In Uzbekistan (and, possibly, Tajikistan and Turkmenistan), common starlings have met their nemesis in the form of common mynas, which have completely displaced them. Mynas are slightly larger and much feistier. I remember seeing them for the first time in the early 1970s, and in the mid-1970s they were even sold in pet shops…

    These days, common starlings can be seen only in the spring and autumn, during their migrations.
    In the early 2000s, there was an attempt to cull mynas, which, predictably, came to nothing.

    More on them (in Russian):
    http://sreda.uz/rubriki/bio/pojdem-v-ataku-na-krysu-v-peryax/
    http://sreda.uz/rubriki/bio/majna-pobedila-oxotnikov/

  203. Mynas were accidentally released in Israel a decade or two ago. They have become ubiquitous, and devastating to local songbirds. Even sparrows are getting scarce at places.

  204. Even sparrows are getting scarce at places.

    Same in Uzbekistan.

  205. January First-of-May says

    Even sparrows are getting scarce at places.

    Come to think of it, it’s been a while since I’ve seen a sparrow in Moscow… though that’s probably more about pigeons (and partly about land mismanagement).

  206. David Marjanović says

    Sparrows have become much rarer in much of Europe. Not in front of my windows, though, where they have impenetrable bushes to nest in and a hole in the parking lot to take dust baths in.

  207. Trond Engen says

    John Cowan: So we can now construct a deeply flawed theory (to be sent posthaste to Nature) that the Palaeo-Eskimo spoke Proto-Dene-Yeniseian.

    Letter to Nature 570 (2019):

    Flegontov et al: Palaeo-Eskimo genetic ancestry and the peopling of Chukotka and North America

    Abstract:

    Much of the American Arctic was first settled 5,000 years ago, by groups of people known as Palaeo-Eskimos. They were subsequently joined and largely displaced around 1,000 years ago by ancestors of the present-day Inuit and Yup’ik. The genetic relationship between Palaeo-Eskimos and Native American, Inuit, Yup’ik and Aleut populations remains uncertain. Here we present genomic data for 48 ancient individuals from Chukotka, East Siberia, the Aleutian Islands, Alaska, and the Canadian Arctic. We co-analyse these data with data from present-day Alaskan Iñupiat and West Siberian populations and published genomes. Using methods based on rare-allele and haplotype sharing, as well as established techniques, we show that Palaeo-Eskimo-related ancestry is ubiquitous among people who speak Na-Dene and Eskimo–Aleut languages. We develop a comprehensive model for the Holocene peopling events of Chukotka and North America, and show that Na-Dene-speaking peoples, people of the Aleutian Islands, and Yup’ik and Inuit across the Arctic region all share ancestry from a single Palaeo-Eskimo-related Siberian source.

    The supplements are open access. Here is the supplementary discussion in extensio:

    The time and place of the Eskimo-Aleut founder admixture event remains uncertain. Under our demographic model, the admixture event that is shared by all members of this lineage was dated by two independent methods, ALDER and Rarecoal, at 2,700-4,900 ya and 4,400-4,900 ya, respectively (Fig. 2b, Supplementary Information section 12), and involved a substantial (~55-62%) genetic contribution from a Northern First Peoples population distantly related to Athabaskans (Fig. 2). There is no clear archaeological evidence for a Native American back-migration to Chukotka, increasing the weight of evidence that this admixture event occurred in Alaska. Indeed, the Alaskan Peninsula and Kodiak Archipelago have long been suggested as a source of influences shaping the Neo-Eskimo material culture (Fig. 3b). Some of the earliest maritime adaptations in Beringia and America are encountered in this region associated with the Ocean Bay tradition (~6,800 – 4,500 calBP). Around 4,000 calBP, the Ocean Bay tradition was succeeded by the Early Kachemak tradition, which is seen as a dramatic departure from the preceding phase, with some archaeological evidence for contacts with the Paleo-Eskimo Arctic Small Tool tradition. Given the new genetic results, it seems possible that this cultural discontinuity is associated with the emergence of the ancestral Eskimo-Aleut population. Early Paleo-Eskimo people used marine resources on a seasonal basis only, depended for the most part on hunting caribou and muskox, and lacked sophisticated hunting gear that allowed the later Inuit to become specialized in whaling. It is conceivable that a transfer of cultural traits and gene flow between Paleo-Eskimos and First Peoples happened simultaneously.

    An important further clue is given by our finding that the ancestors of Inuit/Yup’ik experienced bidirectional gene flow with Chukotko-Kamchatkan ancestors, while Aleuts did not. This is consistent with a scenario of PPE/First Peoples admixture in Alaska, and a subsequent migration of Aleut ancestors into the Aleutian Islands (Fig. 3b), which might have occurred around 4,000 calBP according to known discontinuities in the Aleutian archaeological record (the onset of the Margaret Bay phase, which saw an influx of ASTt and Kodiak elements8). Conversely, ancestors of Inuit and Yup’ik migrated back to Chukotka, where around 2,200 calBP they established the earliest culture securely assigned archaeologically and genetically to Neo-Eskimos, i.e. the Old Bering Sea culture, admixed with local populations, most likely in interior Chukotka, and re-expanded from there to Alaska and later throughout the American Arctic. The Thule expansion was likely driven by innovations in hunting and the food surplus created by whaling. The oldest Old Bering Sea individual in this study was dated to ~1,500-1,900 calBP, which also overlaps our estimated time of the bidirectional admixture between Inuit/Yup’ik ancestors and Chukotko-Kamchatkan-speaking groups (~1,700-2,300 ya).

    A succession of western Alaskan cultures, namely the Old Whaling, Choris, Norton, and Ipiutak (with the earliest dates around 3,100, 2,700, 2,500, and 1,700 calBP, respectively), combined cultural influences from earlier local Paleo-Eskimo sources as well as sources in Chukotka and southwestern Alaska. Parallels between these cultures and subsequent Neo-Eskimos are notable, and they might represent partial links between the founding population at 4,800 ya and the Old Bering Sea culture at 2,200 calBP (Fig. 3b). The location and source populations for early Eskimo-Aleuts will likely be resolved if future analyses can include samples from these western Alaskan traditions, as well from the Ocean Bay and Kachemak traditions in southwestern Alaska.

    The descendants of the proto-Paleo-Eskimo lineage speak widely different languages, belonging to the Chukotko-Kamchatkan, Eskimo-Aleut, and Na-Dene families. Based on lexicostatistical studies of languages surviving in the 20th century, the time depth of the former two families is likely shallow, and the Na-Dene family is probably much older, on the order of 5,000 years (Supplementary Information section 13). Thus, while the linguistic affiliation of Paleo-Eskimos is impossible to determine from genetic data, the finding that the most diverse linguistic group whose speakers carry large proportions of PPE ancestry is Na-Dene and that Na-Dene linguistic variation may reach back to the Paleo-Eskimo period suggests that proto-Na-Dene may have been spoken by a Paleo-Eskimo population. A Siberian linguistic connection was proposed for the Na-Dene family under the Dene-Yeniseian hypothesis. This hypothetical language macrofamily unites Na-Dene languages and Ket, the only surviving remnant of the Yeniseian family, once widespread in South and Central Siberia. Although the Dene-Yeniseian family is not universally accepted among historical linguists, and correlations between linguistic and genetic histories are far from perfect, evidence of a genetic connection between Siberian and Na-Dene populations mediated by Paleo-Eskimos suggests that future research should further explore Dene-Yeniseian as a genealogical family or as part of a wider clade.

    To celebrate multiple links, the open version of the letter at Academia.edu

  208. Trond Engen says

    In Flegontov et al (2019) the breakup and complex admixtures forming Na-Dene, Paleo-Eskimo and Eskimo-Aleut take place in short order just after 5000 BP.

    The archaeologist Andrew Tremayne has an interesting series of papers on the chronology and relation of the cultures of western Alaska. In Tremayne and Brown (2017): Mid to Late Holocene Population Trends, Culture Change and Marine Resource Intensification in Western Alaska he and William Brown analyse population development within the AST tradition and quantifiy a collapse around 3500 BP and reemergence as the Norton tradition centuries later.

    Keeping the branching diagrams from Flegontov et al (2019) but applying them on the events in Tremayne and Brown, we could suggest the following shallower timeline:

    Ca. 5000 BP: Peoples belonging to the Arctic Small Tool tradition cross the Bering Strait and find a subsistence niche built on caribou hunting, inland fishing and seasonal stints at the coast. A good niche, because they multiply until they make up around half the population of Alaska.

    Ca. 3500 BP: A collapse in the caribou population, maybe caused by a major volcanic event, the Aniakchak eruption of 3700 BP, forces the Alaskan caribou hunters into new ways of life. Some go east looking for other caribous, join the local forest people and become the Na-Dene, others specialize in marine hunting. Some of these absorb local groups of marine hunters and become the Eskimo-Aleuts, others become the Norton tradition of Paleo-Eskimos.

    Ca. 2500 BP: The Norton people have consolidated in the nortwestern region and start spreading north and east, eventually replacing earlier ASTt’ers and becoming Dorset and Saqqaq. The Eskimo-Aleuts start expanding through the Bering Strait region. One group crosses into Chukotka where they mix with Chukchi-Kamchatkans and eventually become the Eskimos. The Na-Dene expand and absorb peoples across the northern forests.

  209. Thanks for these updates!

  210. Trond Engen says

    I should stress that I haven’t quoted all those papers from different sources to support one coherent, overarching story. They’re just bits and pieces that somehow are part of an emerging bigger picture. The Tremayne and Brown paper shows that a near simultaneous three-way split-and-admixture well after the 5000 BP crossing is possible archaeologically, but an earlier date is probably better from purely genetic considerations. In both scenarios the ASTt people established themselves in Western Alaska with a mixed economy based on caribous, fish and marine mammals. In the earlier scenario one group soon joined the forest people and became the ancestors of the Na-Dene. Those who stayed behind on the coast developed in different directions. The northerners quickly spread out to fill the empty niche on the tundra and are known as Paleo-Eskimos, the southerners merged with local marine hunters and eventually became the Eskimo-Aleut. As is said in Flegontov et al (2019), the picture will probably become clearer with more genomes from Western Alaska.

    Either way, the linguistics is disconcerting. In the genetic models the “invaders” were a single population until the three-way split-and-admixture event. This is reasonable since they would have been few and dependent on eachother. Still they must have carried at least PP-Eskimo-Aleut and PP-Na-Dene with them if the Old World origin of both is to be upheld. It would be tempting to attribute Eskimo-Aleut to a later wave, after the back-migration to Chukotka, but that can’t account for the Aleuts. Is my impression correct that the Old World features of Eskimo-Aleut are mostly found in Eskimoic? What if Aleut is the conservative branch of Eskimo-Aleut and the Old World features in Eskimoic came about by contact with (Para-)Yukaghir in the Old Bering Sea phase?

  211. David Marjanović says

    Interesting idea.

  212. John Cowan says

    Nature 570

    Ghordjazz. Life imitates art.

  213. There is no any PIE or Indo-European language in reality. The term Indo-Eropean was created based on British & Eropean political targets in India.

  214. Sigh.

  215. Stu Clayton says

    Mr. Mellow Yellow can muster only a sigh now. I noticed that you had two yellow things for dinner recently: chicken curry and lemon meringue pie.

  216. True, and they were both yellowly delicious.

  217. yellow: from Middle English yelwe, yelou, from Old English ġeolu, from Proto-Germanic *gelwaz, from Proto-Indo-European *ǵʰelh₃wos.

  218. John Cowan says

    I think that all of Ardich’s words are PIE in origin, with the possible exception of Europe, which may be < some relative of Akkadian erebu ‘go down, set’ (of the sun) = ‘the West’, an analogue of Orient (thus Etymonline). Or it may not. Then again Erope is certainly not PIE either.

    Historical linguistics vs. the irredentist agenda.

  219. David Marjanović says

    There is no any PIE or Indo-European language in reality. The term Indo-Eropean was created based on British & Eropean political targets in India.

    Ha! That’s what they want you to believe!

    (And they want me to believe I just passed through Bielefeld an hour and a half ago.)

  220. Lars (the original one) says

    There is a pleasing symmetry to using a word originally from the East to denote the West (Europe) and vice versa (Orient), especially when they refer to the same natural phenomenon.

    It took me the longest time to figure out why we kept crossing streets in (Heróica) Puebla called Avenida NN Ote and Pte. Well, we crossed them to get to the other side, but what the last parts meant.

  221. Dmitry Pruss says

    Wasn’t it Marr who taught that all language families are an invention of the imperialists, while the toiling masses always spoke philogeny-less, non-family Japhetic languages?

  222. Stu Clayton says

    Lars TOO: Avenida NN Ote and Pte

    I had to check the RAE about that. Oriente/Poniente for East/West – “poniente” because that’s where the sun goes down, I suppose. Did you find out more ? Is this rebellion against the (apparently) English words “east/west” ?

  223. Lars (the original one) says

    At a guess, a worry that este and oeste would sound too much alike. But the Mexicans I was there with weren’t native to the city and couldn’t even supply the expansion for Ote/Pte, I found it somewhere else.

    The street numbers are fun, east-west Avenidas have odd numbers increasing to south of the Zocalo, even numbers to the north, while north-south streets are also Avenidas, odd towards the west and even towards the east — those are divided into Sur and Nte parts. Except that if some stretch is too narrow to be an Avenida, it’s a Calle with the same number and direction. And except that the whole system is at a 30 degree angle to the cardinal directions, so Norte is almost north-east, but still more north than east. There is some redundancy in this; if you are where two even-numbered streets cross you know it’s the Oriente and Norte parts of those Avenidas, and vice versa. And you will never run out of street numbers or house numbers as the city grows.

    Avenida 1 in all four directions have ‘real’ names — A. de la Reforma to the west, Cinco de Mayo to the north, Don Juan de Palafox y Mendoza to the east, and 16 de Septiembre to the south.

  224. A review of the shared fairy-tiles about the constellations, including the celestial hunters chasing Ursa Major with the help of Alcor the bird, shared between Siberians and Native Americans
    https://www.livescience.com/pleiades-constellation-origin-story.html

  225. Trond Engen says

    The Dene migrations are discussed in Moccasins and Dene Migration.

  226. @David M. “Guide for the Perplexed: Uralistic ɜ is not IPA ɜ, it just means “any vowel”.”

    To mean ‘any vowel’ it is better to use the symbol V, which contrasts with C ‘any consonant’.

  227. Clearly, and I think basically no-one actually actively uses “ɜ” anymore (it got stamped out in the typewriter era together with a few other similar quirks); but the tradition is out there; if we cite items from the UEW as “aŋV” when they’re actually written as aŋɜ, there is bound to be some gaggle of people who would fail to find the word when checking themselves.

  228. John Cowan says

    crossed [the streets] to get to the other side

    A purely Covid riddle:

    Why didn’t the chicken cross the road?

    It was delivered.

Trackbacks

  1. […] as does the BBC. (How did this happen?) January 21: a robust conversation begins over on LanguageHat. January 22: William Pooley documents some of the difficulties glossed over by da Silva and Tehrani […]

Speak Your Mind

*