The Great Macrofamily thread: Indo-Uralic, Altaic, Eurasiatic, Nostratic etc.

[Tropylium]

And how does Sahara drying out force people in the Near East to die out? As we can see from the current situation, one can be overwhelmingly a desert while the other mostly isn't. Do we have some kind of biogeographical data showing massive desertification periods also in the Near East?

Additionally, the coasts and river valleys have remained inhabitable all along. I've never been clear on what part of the Sahara Pump theory of human dispersal prevents people from just following the Nile and discovering the Mediterranean that way (and from there perhaps going down Red Sea; or up into eastern Anatolia and from there down Euphrates). The current system where the Nile flows at least from Ethiopia to the Mediterranean is apparently 800,000 years old (Atbarah River connecting the earliest) — this is pretty recent geologically, but not anthropologically.

[mèþru]

You succeeded in making me doubt what I thought.

[WeepingElf]

These discussions on where and when exactly Homo sapiens came into being have very little to do with the question of reconstructable macrofamilies. There must have been many migrations between Africa and Eurasia - in both directions. We are dealing with at least 100,000 years here; this is the linguistic equivalent of interstellar spaceflight - there will be no way of reconstructing "Proto-World" anytime in the foreseeable future.

Even a convincing demonstration of Indo-Uralic would be a milestone which would be worth the Nobel Prize in linguistics if there was such a thing! As Tropylium has written here, Indo-Uralic could be a research project that supports several linguists' entire careers, and it would be hubristic for amateurs like me to expect to crack the riddle one day. I don't expect to be the next Michael Ventris!

A popular hypothesis is that Proto-Nostratic was the language of the first Neolithic farmers of the Near East and the family spread with agriculture, but this has problems. The ancient Near East was a crazy quilt of small language families - Sumerian, Elamite, Hurrian-Urartian, Hattic, etc. - usually not claimed for Nostratic. The Neolithic Near East was at least as diverse linguistically. On the other hand, the inclusion of Greenberg's Eurasiatic into Nostratic brings some languages of peoples who never had agriculture into the family. As I already said, I doubt that Afrasian and Dravidian have anything to do with the Mitian cluster (or with each other), and even the MItian connection of Kartvelian seems doubtful to me.

[Tropylium]

As Tropylium has written here, Indo-Uralic could be a research project that supports several linguists' entire careers

Yes, that's kind of the point I'm getting at here. We have no guarantee at all that something like Indo-Uralic is "just beyond" the horizon of what current historical linguistics has figured out. But what if something else is? African linguists are handicapped by much scantier data on most languages, and regardless they continue working on their own macrofamilies, and keep finding morphological and typological markers. It's a real possibility that even something like Nilo-Saharan (at core, after removing some number of areal stragglers that do not belong in after all) is a younger, and therefore more workable, grouping than the "notorious" proposals like Indo-Uralic/Mitian/Nostratic.

And if this were the case, maybe we really should be working on these hypotheses first, to learn the lessons on how exactly long-range relationships ought to be investigated. So far there are no well-reconstructed language families of 10k+ age, but a few of the African families seem like they could end up as that one day.

I guess there's also the philosophical question of why we are interested in macrofamily research at all? Just for the sake of finding more relatives for some specific language of focus like Indo-European; or for the general purpose of pushing our knowledge of linguistic history further back?

The ancient Near East was a crazy quilt of small language families - Sumerian, Elamite, Hurrian-Urartian, Hattic, etc. - usually not claimed for Nostratic.

At least Elamite has been seriously compared with both Afrasian and Dravidian. Quite feasibly one of these can be a layer of contact influence of course, maybe the former in particular, since I think it is likely that the spread of Semitic wiped out a few other Afrasian groups. Compare IE and the Balkans, where Greek/Latin/Slavic wiped out several smaller languages like Dacian, Thracian and Phrygian. Or just the known history of Mesopotamia itself, which has clear mini-spread-zone traits: Akkadian wiped out Sumerian, then Aramaic wiped out Akkadian, and now Aramaic is just about being wiped out by Arabic. There were probably a few other waves of this sort between the rise of agriculture and the invention of writing. And the lesson is not just that there is turnover, but also that new languages mostly come in from the southwest — and are therefore likely to have been Afrasian. So an "Eteo-Sumerian" that was a sister group of Semitic and left some influence also in (pre-)Elamite would be a fairly natural hypothesis.

On the other hand, the inclusion of Greenberg's Eurasiatic into Nostratic brings some languages of peoples who never had agriculture into the family.

Eurasiatic relates to wider Nostratic much in the same way Nostratic relates to other macrofamily hypotheses: it's made up of more thoroughly researched language families, and yet, it does not appear to be too much stronger at all. If the wider version of Nostratic is onto anything, I don't think it is going to have a clear Eurasiatic subgroup versus all other members splitting off earlier. A lot of the binary pairings within Nostratic have never been firmly explored, and perhaps there is actually rather e.g. a "Southwestern Nostratic" family that comprises Kartvelian + Afrasian.

[Salmoneus]

These discussions on where and when exactly Homo sapiens came into being have very little to do with the question of reconstructable macrofamilies.

They are, however, very useful in bringing out of the shadows the broader political-ideological motivations that underpin some people's linguistic theories.

A popular hypothesis is that Proto-Nostratic was the language of the first Neolithic farmers of the Near East and the family spread with agriculture, but this has problems.

Notably, there wasn't a big genetic expansion across the area with the development of agriculture - it seems to have been spread culturally, rather than demically, at least at first (unlike in Europe). The two big expansions into the middle-east - Afro-asiatic speakers out of Africa, and the mysterious expansions out of somewhere in the caucasus/armenia area (which spread as far as the Minoans), occured later, though presumably (given their scale) must have been related to some sort of 'technological' (/sociological, etc) development.

The ancient Near East was a crazy quilt of small language families - Sumerian, Elamite, Hurrian-Urartian, Hattic, etc.

You're very keen on crazy quilts, aren't you? It's not an expression I've ever heard before. Is a crazy quilt actually a thing, or are the quilters in your area just notoriously erratic?

But seriously, we maybe should point out that Sumerian, Elamite and Hurian-Urartian do not come from the ancient Near East, although Hurrian did end up there for a while. That doesn't undermine your broader point, however.

[Pabappa]

Crazy quilt means you mix different fabrics together, I think. Pronounced with stress on the first word.

[Salmoneus]

Crazy quilt means you mix different fabrics together, I think.

But isn't that inherent to being a quilt?

Pronounced with stress on the first word.

Huh! OK.

[mèþru]

The quilting shall not be civilized!!!!!!!!

[Tropylium]

"I must stress, first of all, that the study of macrofamilies is not a quilt…"

[WeepingElf]

As Tropylium has written here, Indo-Uralic could be a research project that supports several linguists' entire careers

Yes, that's kind of the point I'm getting at here. We have no guarantee at all that something like Indo-Uralic is "just beyond" the horizon of what current historical linguistics has figured out. But what if something else is? African linguists are handicapped by much scantier data on most languages, and regardless they continue working on their own macrofamilies, and keep finding morphological and typological markers. It's a real possibility that even something like Nilo-Saharan (at core, after removing some number of areal stragglers that do not belong in after all) is a younger, and therefore more workable, grouping than the "notorious" proposals like Indo-Uralic/Mitian/Nostratic.

Surely, we can't expect to achieve much ourselves. We are a bit like the Middletown Astronomical Society discussing theories of dark matter at their monthly meetings, with no hope of making meaningful contributions to the solution of that riddle. OK, I hyperboled a bit - comparative linguistics, unlike astrophysics, is no "big science" requiring equipment worth millions, and we can still hope to throw up useful ideas.

Almost certainly, in Africa, there are discoveries to be made the like of which have been made in Eurasia decades ago, as many African languages are poorly explored. For instance, I have been told that reconstructions of Proto-Chadic and Proto-Cushitic leave much work to be done, even though these two families are probably not deeper than Indo-European (though they lack the ancient literary languages that were so helpful with IE - but this also holds for Uralic, which is nevertheless quite advanced).

And if this were the case, maybe we really should be working on these hypotheses first, to learn the lessons on how exactly long-range relationships ought to be investigated. So far there are no well-reconstructed language families of 10k+ age, but a few of the African families seem like they could end up as that one day.

I agree with you that there are some very promising research areas in Africa. As I said earlier, Niger-Congo and Afrasian may be as deep as Indo-Uralic or even deeper, yet they have won greater acceptance among linguists. Africanists tend to be more liberal in accepting language families for which some evidence has been found but reconstruction not yet achieved, than Eurasianists.

I guess there's also the philosophical question of why we are interested in macrofamily research at all? Just for the sake of finding more relatives for some specific language of focus like Indo-European; or for the general purpose of pushing our knowledge of linguistic history further back?

Well, to me it is just curiosity. When a pattern is found somewhere, we tend to ask "why?". Of course, little practical benefits can be drawn from this. Does learning Hungarian become easier when we know that it is remotely related to English? Probably not. Certainly, IE is big enough already not to warrant making it even bigger by demonstrating that even more languages are related to it. So it is a matter of improving our general knowledge of language history, and there are areas where even "mid-range" language relationships (i.e., ones of a time depth like IE or Uralic) are probably waiting to be discovered, such as New Guinea, maybe even Africa or the indigenous languages of the Americas. Indeed, it seems a bit like wrongly set priorities when we try to trace the languages of Eurasia back 10,000 years when some other parts of the world are still waiting to be traced back to the 5,000 years already achieved in Eurasia.

The ancient Near East was a crazy quilt of small language families - Sumerian, Elamite, Hurrian-Urartian, Hattic, etc. - usually not claimed for Nostratic.

At least Elamite has been seriously compared with both Afrasian and Dravidian. Quite feasibly one of these can be a layer of contact influence of course, maybe the former in particular, since I think it is likely that the spread of Semitic wiped out a few other Afrasian groups. Compare IE and the Balkans, where Greek/Latin/Slavic wiped out several smaller languages like Dacian, Thracian and Phrygian. Or just the known history of Mesopotamia itself, which has clear mini-spread-zone traits: Akkadian wiped out Sumerian, then Aramaic wiped out Akkadian, and now Aramaic is just about being wiped out by Arabic. There were probably a few other waves of this sort between the rise of agriculture and the invention of writing. And the lesson is not just that there is turnover, but also that new languages mostly come in from the southwest — and are therefore likely to have been Afrasian. So an "Eteo-Sumerian" that was a sister group of Semitic and left some influence also in (pre-)Elamite would be a fairly natural hypothesis.

Yes, Elamite has been connected to Dravidian; what I have seen doesn't look all that bad, but I know so little about these matters that I can't make any judgment. If one conjectures that Proto-Dravidian was the language of the Indus Valley Civilization, the geographical separation of the two becomes easily surmountable. There may have been many more Elamo-Dravidian languages all over what is now Iran. After all, there must have been something there before Iranian moved in, so why not Elamo-Dravidian?

And certainly, Semitic wiped out a number of languages when it spread across the Near East. The question of course is, where was Proto-Afrasian spoken? Here, an African homeland seems likelier to me than a Near Eastern one, as all but one of the Afrasian families are spoken exclusively in Africa, and Proto-Semitic could easily have moved into the Levant from Egypt around 8,000 BC or so. So the Neolthic Near East probably was more diverse than the Bronze Age one.

On the other hand, the inclusion of Greenberg's Eurasiatic into Nostratic brings some languages of peoples who never had agriculture into the family.

Eurasiatic relates to wider Nostratic much in the same way Nostratic relates to other macrofamily hypotheses: it's made up of more thoroughly researched language families, and yet, it does not appear to be too much stronger at all. If the wider version of Nostratic is onto anything, I don't think it is going to have a clear Eurasiatic subgroup versus all other members splitting off earlier. A lot of the binary pairings within Nostratic have never been firmly explored, and perhaps there is actually rather e.g. a "Southwestern Nostratic" family that comprises Kartvelian + Afrasian.

Greenberg's Eurasiatic is based on mass lexical comparison, and therefore highly questionable, hardly better than Amerind. However, most (not all) of the language families in this group share those famous pronouns which led to the nickname "Mitian", and five of them (IE, Uralic, Turkic, Mongolic and Tungusic) belong to the "core" of the Nostratic hypothesis. Yet, even if Nostratic is true, that still doesn't mean that Eurasiatic or "Mitian" forms a valid node within it! The evidence simply is not strong enough to say that these languages are related, but inviting enough to looking closer at them.

These discussions on where and when exactly Homo sapiens came into being have very little to do with the question of reconstructable macrofamilies.

They are, however, very useful in bringing out of the shadows the broader political-ideological motivations that underpin some people's linguistic theories.

Sure - some relationship hypotheses, and also some rejections of them, are motivated by racism and nationalism. Also, as most people speak the language of their germane parents, some correlation between linguistic and genetic affinities is to be expected, Alas, language shifts happen often enough to weaken such correlation almost beyond significance.

A popular hypothesis is that Proto-Nostratic was the language of the first Neolithic farmers of the Near East and the family spread with agriculture, but this has problems.

Notably, there wasn't a big genetic expansion across the area with the development of agriculture - it seems to have been spread culturally, rather than demically, at least at first (unlike in Europe). The two big expansions into the middle-east - Afro-asiatic speakers out of Africa, and the mysterious expansions out of somewhere in the caucasus/armenia area (which spread as far as the Minoans), occured later, though presumably (given their scale) must have been related to some sort of 'technological' (/sociological, etc) development.

Just that. The cultural spread of agriculture in the Near East would probably mean that people continued speaking their Mesolithic languages, and I am of the opinion that language families in Mesolithic times weren't great except in areas that had only recently been populated after the glaciers receded. And we indeed see a high linguistic diversity in the Bronze Age Near East (but see below on that).

The ancient Near East was a crazy quilt of small language families - Sumerian, Elamite, Hurrian-Urartian, Hattic, etc.

You're very keen on crazy quilts, aren't you? It's not an expression I've ever heard before. Is a crazy quilt actually a thing, or are the quilters in your area just notoriously erratic?

But seriously, we maybe should point out that Sumerian, Elamite and Hurian-Urartian do not come from the ancient Near East, although Hurrian did end up there for a while. That doesn't undermine your broader point, however.

Indeed not. Elamite may have been related to Dravidian (which of course doesn't mean that it came from India - Elamite and Dravidian may be the last remnant of an ancient family of the Iranian highland that was eclipsed by Iranian in the Bronze Age, see above); Hurrian-Urartian may be related to Nakh-Daghestanian and Hattic to Abkhaz-Adyghean (there are claimants to both, but I don't know how good the evidence is); and nobody knows where Sumerian came from. In fact, everything we know from written sources in the Ancient Near East may have come from somewhere else. So in the end, we know as much about the languages of the Neolithic Near East as about those of Neolithic Europe - close to nothing.

[Salmoneus]

some correlation between linguistic and genetic affinities is to be expected, Alas, language shifts happen often enough to weaken such correlation almost beyond significance.

With respect, that's nonsense. Language shifts not at least significantly motivated by demic expansion occur occasionally. That certainly means genes (and archaeology, etc) are not 100% certain guides to language. But it's ridiculous to say the correlation is "almost insignificant" just because it's 95% instead of 100%.

Indeed not. Elamite may have been related to Dravidian (which of course doesn't mean that it came from India - Elamite and Dravidian may be the last remnant of an ancient family of the Iranian highland that was eclipsed by Iranian in the Bronze Age, see above); Hurrian-Urartian may be related to Nakh-Daghestanian and Hattic to Abkhaz-Adyghean (there are claimants to both, but I don't know how good the evidence is); and nobody knows where Sumerian came from. In fact, everything we know from written sources in the Ancient Near East may have come from somewhere else. So in the end, we know as much about the languages of the Neolithic Near East as about those of Neolithic Europe - close to nothing.

My point was more a geographical one: Elamite was never, so far as we know, spoken anywhere even vaguely in the vicinity of the Near East. But, looking it up, it seems I'm out of step with the times, and everywhere is now the Near East (and also the Middle East).

[I think of the Near East as being the Balkans, the Levant, and Anatolia. I could just about extend it to Mesopotamia and Armenia, but Iran would be right out, and firmly Middle Eastern (along with Arabia and Egypt). But *shrug* who am I to argue with the tide of usage...]

[WeepingElf]

some correlation between linguistic and genetic affinities is to be expected, Alas, language shifts happen often enough to weaken such correlation almost beyond significance.

With respect, that's nonsense. Language shifts not at least significantly motivated by demic expansion occur occasionally. That certainly means genes (and archaeology, etc) are not 100% certain guides to language. But it's ridiculous to say the correlation is "almost insignificant" just because it's 95% instead of 100%.

OK - I may just have overestimated the significance of language shifts, perhaps because I have grown up in a northern German village that was undergoing a language shift (from Low German to High German) myself, so language shifts do not seem out of the ordinary to me. But actually, they seem to be rather uncommon. Genes and archaeology are not perfectly certain guides to language, but where both coincide, language probably does so as well.

Indeed not. Elamite may have been related to Dravidian (which of course doesn't mean that it came from India - Elamite and Dravidian may be the last remnant of an ancient family of the Iranian highland that was eclipsed by Iranian in the Bronze Age, see above); Hurrian-Urartian may be related to Nakh-Daghestanian and Hattic to Abkhaz-Adyghean (there are claimants to both, but I don't know how good the evidence is); and nobody knows where Sumerian came from. In fact, everything we know from written sources in the Ancient Near East may have come from somewhere else. So in the end, we know as much about the languages of the Neolithic Near East as about those of Neolithic Europe - close to nothing.

My point was more a geographical one: Elamite was never, so far as we know, spoken anywhere even vaguely in the vicinity of the Near East. But, looking it up, it seems I'm out of step with the times, and everywhere is now the Near East (and also the Middle East).

[I think of the Near East as being the Balkans, the Levant, and Anatolia. I could just about extend it to Mesopotamia and Armenia, but Iran would be right out, and firmly Middle Eastern (along with Arabia and Egypt). But *shrug* who am I to argue with the tide of usage...]

What I meant with the "Near East" was the region consisting of Anatolia, the Levant and Mesopotamia. I apologize for the misunderstanding. And even to this region, Elamite is marginal, originating in the southwestern part of the Iranian highland. Where Sumerian came from, is uncertain. The notion that it must come from inner Asia because of its agglutinating structure is misguided; such conclusions are invalid. At least, Sumerian shows no clear signs of affinity to Uralic or Altaic, nor to Dravidian (or, for that matter, Elamite).

[gach]

With respect, that's nonsense. Language shifts not at least significantly motivated by demic expansion occur occasionally. That certainly means genes (and archaeology, etc) are not 100% certain guides to language. But it's ridiculous to say the correlation is "almost insignificant" just because it's 95% instead of 100%.

I'm no expert in population genetics, but 95% is still a hugely tight correlation. Can you back that up with a reference? I wouldn't expect nearly as tight a relation between language and genetics.

For instance, the understanding I have of the expansion of Uralic is that you can find individual genetic markers that trace the family to a certain extent. There is, however, no single Uralic genetic type that you could identify. The interpretation you can make from this is that there was a movement of people associated with the spreading of language, but these immigrants wouldn't form large enough populations in their new homelands so that they could have dominated the local genetics. Most of the spread must then have happened as a language shift by the original population.

[mèþru]

As someone from the Near East: until classical antiquity, Iran and even India were more important and familiar to my ancestors than Greece.

I think that Proto-Dravidian was spoken by a peoples neighbouring and likely trading with the Indus Valley Civilisation. I think that the population of the civilisation originally spoke an unrelated language (probably related to either Burushaski or Nihali) and may have continued to do so in later stages, but that Dravidian peoples replaced/mixed in with the original population over time.

[WeepingElf]

With respect, that's nonsense. Language shifts not at least significantly motivated by demic expansion occur occasionally. That certainly means genes (and archaeology, etc) are not 100% certain guides to language. But it's ridiculous to say the correlation is "almost insignificant" just because it's 95% instead of 100%.

I'm no expert in population genetics, but 95% is still a hugely tight correlation. Can you back that up with a reference? I wouldn't expect nearly as tight relation between language and genetics.

Nor would I. Perhaps 80 to 90%, but not much more; maybe even less. There are some conditions which are conducive to language shifts, most of which, however, are almost certainly rare in "primitive" societies. The language shift in my home village, for instance, was of course towards the official language of a modern nation state. Language shift happened massively in the Roman Empire, of course; for instance, the French have more Gaulish than Roman ancestry, yet they speak a descendant of the latter's language. Not because the Romans actively eradicated Gaulish and other languages of their subjects - they couldn't care less what the people spoke at home - but because they simply didn't bother learning their languages (except Greek), so their subjects had to learn Latin in order to deal with their new rulers and participate in their sophisticated culture. Also, Latin was useful in dealing with people of a different language. So a Gaulish merchant ordering wine from an Iberian vintner would write his order in Latin, which he could expect his business partner to know.

But I am digressing. The question is this: to which degree did language shift happen in Neolithic Europe, for instance? In some ways, the situation will have been similar to colonial-era North America, even if the technology gap was smaller - the farmers move in and displace the hunter-gatherers into marginal territories the farmers do not covet because they are less suitable to agriculture than the loess-filled valleys they settled. Eventually, the farmers were the vast majority, and the remaining foragers gradually acculturated and probably also took over the farmers' languages. The end result is a community which is genetically mostly Neolithic and to a much smaller part Mesolithic, and speaks a Neolithic language. Same thing happened with the advent of the Indo-Europeans, who had advantages such as horses, draught animals, the plough, the wheel and copper, and soon dominated because they could work much more land per capita than the Neolithics and thus reproduced faster. (Today's Central Europeans are genetically more Yamnaya than LBK.) So we have a good, but not perfect agreement between genes and languages.

For instance, the understanding I have of the expansion of Uralic is that you can find individual genetic markers that trace the family to a certain extent. There is, however, no single Uralic genetic type that you could identify. The interpretation you can make from this is that there was a movement of people associated with the spreading of language, but these immigrants wouldn't form large enough populations in their new homelands so that they could have dominated the local genetics. Most of the spread must then have happened as a language shift by the original population.

Uralic seems to be loosely connected to the N3 (or whatever its designation now is) Y-DNA haplogroup, but I concur with you that, for instance, Finns have not really that much in common with Samoyeds genetically. The Saami are, it appears, also genetically quite distinct, and a language shift has been proposed here. Turkic also involved a lot of language shift, which we can ascribe to élite dominance.

So at any case, one must be careful there.

[Salmoneus]

No, I don't have a citation for that '95%' - I don't even know how you'd operationalise these terms for a rigorous study. It was impressionistic, a way of saying 'not universal, but close to it'.

And since people seem to be drawing an unintended binary out of it - either there's 100% population replacement through genocide or else there's an unmotivated language shift - let's narrow the terms a little here.

By, let's call it, "anomolous" language-shift, I mean:
- a shift from one language to a language that is not closely related, so that the shift cannot be mediated through intermediary dialects, the spread of individual features one-by-one, or sociolectical continua, but must require some relatively clean break between two languages that are not mutually intelligible and have no intermediary forms. Shift from one dialect to a dominant neighbouring dialect, I'm not counting - nor would the spread of high german features northward into low german through upper registers, resulting in eventual replacement of the latter, count in this case.

- a shift that is not accompanied by a clear genetic shift, that leaves clear markers of the intrusive population in the resulting population. [often, though not always, disproportionately found in the y-dna, which tends to indicate political dominance]

- furthermore, I'll accept that in modern times, when the intrusive population is a global empire and the native population is a remote tribe, the power imbalance (and the technology of transport and communication) may sometimes be enough to result in language shift without a big direct genetic contribution, in a way that's not directly comparable to pre-imperial times.

Within these parameters, anomolous language shifts I'm sure do occur, but they are very rare. So, I'm not saying the language always matches the majority genetic heritage of a group; just that it's very unusual for the language of a group to shift without a visible genetic influx. In other words: language adoption rarely happens at a distance!


In the case of Uralic: the big obvious marker is N1 paternal lineage, which is common throughout Uralic-speaking populations, and to a lesser extent some neighbouring groups (Indo-Europeans in the Baltic or along the Volga, for instance), but very rare or non-existent elsewhere. It turns out, the same is also true autosomally. Now, sure, Uralic speakers west of the Urals are all genetically majority-IE, and also show remnants of the mesolithic and neolithic european populations. But they also show the distinct marks of an intrusive Siberian population - 8% in Finns, 5% in Estonians, but only 1% in Lithuanians and 0% in Latvians, according to Tambets et al this year, who conclude "we find that most of the Uralic speakers and some of their neighbours share a genetic component of possibly Siberian origin. Additionally, we show that most Uralic speakers share significantly more genomic segments identity-by-descent with each other than with geographically equidistant speakers of other languages. We find that correlated genome-wide genetic and lexical distances among Uralic speakers suggest co-dispersion of genes and languages." Crucially, so far as we can see neither the y-dna nor the autosomal component were to be found anywhere in europe until after about 500BC, which, indeed, ties in well with the young ages of some of these y-dna lineages. We can be extremely confident that there was a mass population movement into currently Uralic areas by a closely-connected group of people from Siberia within the last 3000 years, and that population movement does seem to align with current Uralic-speaking groups. Particularly because in at least some cases that introgressive population clearly adopted a socially dominant position - as demonstrated by the fact that the Finns are only about 8% Siberian in their total genetics, but at least 70% of Finns are descended in the paternal line from male Siberian immigrants, which tends to men the immigrant men were having a lot more sex, which tends to mean they were socially dominant.

So sure, you can't say that all Uralic-speakers form a monolithic racial entity entirely pure and uncontaminated by non-Uralics; but that's not really the point.

[I know, Tropylium's going to come in and yell at me for impugning the white european credentials of the finns. Whatever.]

The one big exception here is apparently the Hungarians, who, it seems, do not have any noticeable Uralic heritage other than their language. It's possible this was originally diluted alongside their cultural Altaicisation, but the main culprit is probably the late invasion of Hungary, when the small invading population was able to use the mediaeval state institutions to impose their language more widely than their semen (and where at the same time the invaders were soon culturally dominated by and genetically pervaded by the wider European cultural elite). As I say, this sort of process isn't impossible, it's just very unusual, and much more likely to happen in eras with greater institutional and technological development.


Regarding Rome, incidentally: yes, lots of people adopted Latin. But at the same time, they also massively intermarried with Romans.

[Vijay]

I've heard (in passing) of Dravidian being compared to all kinds of other language families/languages and have always thought this was either because of (a) typology or (b) chance phonetic similarities noted by certain overenthusiastic Tamils with little to no actual knowledge of linguistics. (Such people invoke option (b) less often than others invoke (a), IIRC). It doesn't look like there's ever that much evidence in favor of a relationship, does it?

[Salmoneus]

I think that Proto-Dravidian was spoken by a peoples neighbouring and likely trading with the Indus Valley Civilisation. I think that the population of the civilisation originally spoke an unrelated language (probably related to either Burushaski or Nihali) and may have continued to do so in later stages, but that Dravidian peoples replaced/mixed in with the original population over time.

Do you have any reason for thinking these things?

The Indus Valley civilisation was apparently primarily a mix of the native Indian population (now mostly seen in various scheduled tribes) and farmers from Iran. Modern Dravidians have elements of that native population, as well as of Indo-Europeans, and apparently also some heritage from a mixed group similar to the Indus Valley civilisation (though the same mixture could have happens elsewhere - it may be that the iranian farmers brought farming further into india, but just happened only to develop civilisation in the IV).

So there are four big, obvious possible Dravidian-Harappan relationships:
a) Dravidian and Harappan are (/were) both relicts of the ancestral Indian population, and hence are cousins
b) Dravidian is the descendent of Harappan, which was a relict
c) Dravidian is a relict, but Harappan was an intrusive language introduced by the Iranian farmers. Dravidian and Harappan are entirely unrelated
d) Dravidian is the descendent of Harappan, which was intrusive
e) Dravidian and Harappan are (/were) both languages brought by an intrusive population from Iran

b) and d) suffer from the problem that the two languages come from opposite ends of India and there's no sign of Dravidian ever having been spoken up there (Brahui is a much later migration). I don't see how to meaningfully distinguish between the others in terms of probability. There's also a varient of c), let's call it f), in which the Harappans are intrusive, and the Dravidians are natively from somewhere in the Iran-Pakistan-Northwest India area and were displaced by the Harappans, but that would really be speculative.

I don't see how you could know about a shift from a Burushaskoid language to a macro-Dravidian language in ancient Harappa, given the lack of interpreted written materials?

[Anyway, regarding Elamo-Dravidian: Elamo-Harappan is reasonably plausible historically under scenarios c-f, and possible but less likely under a-b (a back-migration of the mixed indus population back into Iran). But, let's not, geographically/historically not THAT much more plausible that Sumero-Harappan or Hurrio-Harappan, since those languages were also spoken not far from Iran. A broader Elamo-Dravidian presumably relies on a further Harappo-Dravidian ink (i.e. a-b or d-e). If you DO want to find anothe language to link Dravidian too, Elamite may be about the most prima facie plausible - Dravidians ARE the descendents of an intrusive population that DOES seem to have come at least to a large extent from Iran, where Elamite WAS spoken, so no miracles need to be assumed there. The time depth would not be inviting, however...]

[mèþru]

I read that the reconstructed core vocabulary of Dravidian reflects an environment similar to that of ancient forest east and south of the IVC. That, and the IVC is likely the source of many non-Dravidian substratum words found in Indo-Aryan but not in Iranian

[gach]

No, I don't have a citation for that '95%' - I don't even know how you'd operationalise these terms for a rigorous study. It was impressionistic, a way of saying 'not universal, but close to it'.

The point being that even without explicit numerical values, stating that something is "close to universal" is a strong claim and requires sturdy evidence to back it. Otherwise it doesn't count as solid scientific reasoning.

In the case of Uralic: the big obvious marker is N1 paternal lineage, which is common throughout Uralic-speaking populations, and to a lesser extent some neighbouring groups (Indo-Europeans in the Baltic or along the Volga, for instance), but very rare or non-existent elsewhere. It turns out, the same is also true autosomally. Now, sure, Uralic speakers west of the Urals are all genetically majority-IE, and also show remnants of the mesolithic and neolithic european populations. But they also show the distinct marks of an intrusive Siberian population - 8% in Finns, 5% in Estonians, but only 1% in Lithuanians and 0% in Latvians, according to Tambets et al this year, who conclude "we find that most of the Uralic speakers and some of their neighbours share a genetic component of possibly Siberian origin. Additionally, we show that most Uralic speakers share significantly more genomic segments identity-by-descent with each other than with geographically equidistant speakers of other languages. We find that correlated genome-wide genetic and lexical distances among Uralic speakers suggest co-dispersion of genes and languages." Crucially, so far as we can see neither the y-dna nor the autosomal component were to be found anywhere in europe until after about 500BC, which, indeed, ties in well with the young ages of some of these y-dna lineages. We can be extremely confident that there was a mass population movement into currently Uralic areas by a closely-connected group of people from Siberia within the last 3000 years, and that population movement does seem to align with current Uralic-speaking groups. Particularly because in at least some cases that introgressive population clearly adopted a socially dominant position - as demonstrated by the fact that the Finns are only about 8% Siberian in their total genetics, but at least 70% of Finns are descended in the paternal line from male Siberian immigrants, which tends to men the immigrant men were having a lot more sex, which tends to mean they were socially dominant.

The 8% and 5% proportions of eastern genes are clearly there but at the same time they don't form nearly the majority of the Finnic genetic ancestry. They are the part which you can identify with the people who brought the Uralic languages to the shores of the Baltic Sea. The language certainly came from the east. The rest of the genetic ancestry of the people points to other routes into the country, though. These people must have spoken other languages which their descendants, including the present population, have since lost. That right there is language shift, a large proportion of the population acquiring a first language which was foreign for their ancestors.

I don't think that anyone is denying the importance of migrations to the spread of languages. The real data, as I'm aware, simply suggests that you can't blindly identify your genetic and linguistic ancestors as the same people. In some cases that might work well but in others it might lead you to grossly erroneous results that fail to acknowledge the complexities of real history. It's just not a reliable initial assumption to make.

So sure, you can't say that all Uralic-speakers form a monolithic racial entity entirely pure and uncontaminated by non-Uralics; but that's not really the point.

[I know, Tropylium's going to come in and yell at me for impugning the white european credentials of the finns. Whatever.]

I think it's quite unnecessary to invoke ideas such as "pure and uncontaminated" genes or "white european credentials". Nothing good or productive lies that way and surely no one else has brought them up.