• Welcome to the new Internet Infidels Discussion Board, formerly Talk Freethought.

Language as a Clue to Prehistory

  • Austric 10 KYA -- 9 kya (S China: one site of invention of agriculture)
  • Austronesian 5 KYA -- 5.5 kya (Neolithic farmers)
  • Tai-Kadai 5 KYA -- Kra-Dai -- 4 kya
  • Austroasiatic 7 KYA -- 4 kya (Neolithic farmers)
  • Miao-Yao 4 KYA -- Hmong-Mien

George Starostin then gets into Borean: (Eurasiatic, Afroasiatic), Dene-Caucasian, Austric, with the most work on that being Sergei Starostin's big list of putative EA - DC cognates.

He proposes an age of 15 – 17 KYA. With EA and DC bumped up by 2 kya, that would be 17 - 19 kya, about right for the end of the Last Glacial Maximum. For 6 kya, from AA, that's 21 - 23 kya, well inside the LGM.

USGS Professional Paper 1386–A: Figure Gallery 2, Figure 34 -- there is a little bit of a warm period between 16 and 18 kya, though not as warm as Bølling–Allerød or the Holocene.

Before the LGM, there were lots of Dansgaard-Oeschger events, brief warm periods.
 
Last edited:
Comparison with the ‘Borean’ data has not produced any conclusive results, suggesting that Khoisan, at least, cannot be included in ‘Borean’, although genetic connections on an even deeper level might be possible.

Khoisan is three generally-recognized families, Khoe, Tuu, and Kx'a, all in southern Africa, and two isolates, Hadza and Sandawe, both in Tanzania. Khoisan langs have click sounds, a very unusual feature, and this "family" is based on having those sounds and not being classified as anything else.

In some other work, GS found groupings Hadza , ( Central: (Sandawe, Khoe), Peripheral: (Tuu, Kx'a) )
The preliminary lexicostatistical study of Sub-Saharan African languages identifies at least 4 other superfamiles: Niger-Congo (not quite identical to the Niger-Congo super-family proposed by J. Greenberg), East Sudanic, Central Sudanic, and Kordofanian, plus a number of smaller branches whose position is so far very unclear (such as Songhay or Atlantic languages)
In other work: Lexicostatistical Studies in East Sudanic I: On the genetic unity of Nubian-Nara-Tama | George Starostin - Academia.edu and The Nilo-Saharan hypothesis tested through lexicostatistics: current state of affairs | George Starostin - Academia.edu finding Macro E Sudanic, Macro C Sudanic, Saharan, Koman-Gumuz, Kuliak, Songhay, and Shabo as separate groupings.

Atlantic and Kordofanian are at least possible members of Niger-Congo -- most of these doubtful cases are at the northwest end of NC, and Kordofanian is an offshoot in the northeast. This suggests that Proto-NC speakers lived in the west end of Africa just south of the Sahara Desert.
 
Looking at Current perspectives and the future of domestication studies | PNAS again, and looking in  Sorghum, it's hard to find any domestication event that was at the putative homeland of Niger-Congo: NW Africa just S of the Sahara Desert, and also at the appropriate time, at least 7,000 years ago.

In the middle of that region is  Mande languages a possible member of NC. On the Problem of the Proto-Mande Homeland
I conclude that speakers of Proto-Mande most probably lived around the second half of the 4th millennium BC in Southern Sahara, somewhere between 3° and 12° Western longitude, to the North of the 16° or even 18° latitude. This hypothesis is supported through analysis of the cultural vocabulary that can presumably be reconstructed for the Proto-Mande language.
That means 3500 BCE or 5,500 years ago.

That's a little bit younger than the 6,000 - 7,000 years ago that one gets from the Benue-Congo subfamily of Niger-Congo:

Automated Dating of the World’s Language Families Based on Lexical Similarity | Current Anthropology: Vol 52, No 6
also at
(PDF) Automated Dating of the World’s Language Families Based on Lexical Similarity
and
Automated dating of the world's language families

Though that is broad Benue-Congo, roughly Wikipedia's Volta-Congo, not narrow BC, Wikipedia's BC article.
... we still find a rather advanced society, one that is well acquainted with agriculture (rice, fonio, probably millet and yams), living in villages with dogs, cats, chickens, small and possibly large cattle.

However hasty and imperfect this survey of cultural vocabulary, hypothetically reconstructed for Proto-Mande, may appear, it still seems to provide more evidence for a Southern Saharan rather than Sahelian homeland
 Fonio is a grain crop of West Africa.
 
Distant relationships among the Papuan (= Non-Austronesian) languages of New Guinea and aboriginal languages of Australia remain to be investigated. It is possible that in that region we could distinguish up to 4 – 6 superfamilies (Trans-New-Guinea, Australian, East Papuan, ¤c.), none being properly reconstructed. Some lexical similarities have also been spotted between Trans-New-Guinea morphemes and some of the alleged 'Borean' roots, but these remain too scarce to establish a firm connection.
Another of the calibration points in that "Automated Dating" paper is for Pama-Nyungan in Australia: 4,500 years ago, when Proto-Pama-Nyungan speakers started spreading out of N Australia.
 
Distant Language Relationship: The Current Perspective | George Starostin - Academia.edu
also
Distant Language Relationship: the Current Perspective

Turning to Joseph Greenberg's Amerind, GS concludes
A very preliminary proposal based on EHL exploratory studies suggests that in the Americas one can find at least three types of grouping:

(i) The Almosan superfamily (Algic, Salishan, Wakashan, and some other languages) might be related to Chukchee and Nivkh languages of North Asia, forming a so-called “Beringian” superfamily; connections with “Borean” have been noticed as well.

(ij) A number of families (Penutian, Hokan, Mayan, Mixe-Zoque, Maipuran, Pano-Tukanoan, &c.) presumably form a different superfamily, also with resemblances to “Borean”.

(iij) At the same time, we were not able to detect any external relations for such well-established families as Siouan, Gulf, or Otomanguean.
I looked for some patterns, looking at  Amerind languages and 2017 - Italian Journal of Linguistics - Raoul Zamponi on "First-person n and second-person m in Native America: a fresh look"
  • Northern:
    • Almosan-Keresiouan
      • Almosan - x Bor
      • Keresiouan - x x
    • Hokan-Penutian N-M Bor
  • Central: N-M x
  • Southern: (N-M and Bor scattered)
The N-M refers to these pronouns.

So N-M and Borean connections are not very well correlated, whether positively or negatively.

Unfortunately, I have not found any followup to these claims.
 
DID INDO-EUROPEAN LANGUAGES STEM FROM A TRANS-EURASIAN ORIGINAL LANGUAGE? AN INTERDISCIPLINARY APPROACH – SCIENTIFIC CULTURE
This interdisciplinary study allowed me to establish, on the basis of linguistic, genetic, archaeological, histor-ical and religious data, that linguistic concordances between Gaulish and Slavic were linked with Neolithic migrations from North-Western India and Pakistan to Iran, Mesopotamia, Anatolia, the Caucasus, the North of the Black Sea, Danubic and Balkan Europe, Gaul and Iberia, where Neolithic farmers contributed to the formation of the megalithic civilisation which developed in Gaul from 5.000 BC and brought an archaic language stemming from a Trans-Eurasian original language. This explains the linguistic concordances I estab-lished between Gaulish and Dravidian languages – 250 common words from the 500 words I studied (and 160 with Burushaski), as well as with Altaic, Uralic, Kartvelian, Anatolian and Middle-Eastern languages.
Except that that is just plain wrong about those migrations. There was likely one from there, yes, but well into the Pleistocene, some 40,000 - 50,000 years ago. That is much older than the oldest well-reconstructed families like Indo-European and Austronesian, and is older than the Afrasian, Nostratic, and Dene-Caucasian.

Furthermore, there is plenty of evidence of later migrations, migrations that often left a strong genetic imprint, migrations that do not have this homeland. Neolithic farmers spreading out from the Fertile Crescent in the Middle East, Proto-Indo-European speakers riding their horses outward from the Pontic-Caspian steppe belt (Ukraine, S European Russia, W Kazakhstan), ...

First bioanthropological evidence for Yamnaya horsemanship | Science Advances
Here, we report five Yamnaya individuals well-dated to 3021 to 2501 calibrated BCE from kurgans in Romania, Bulgaria, and Hungary, displaying changes in bone morphology and distinct pathologies associated with horseback riding. These are the oldest humans identified as riders so far.
 
That paper has an odd diagram on PDF page 12 (document page 26). A diagram that claims to show what prehistorian V. Gordon Childe thought were migration routes into Europe from a homeland in E Iran & NW Indian subcontinent.
  • Mediterranean race: SE then SW coast of Arabian peninsula, then S coast of Mediterranean, then Iberia, then Britain.
  • Alpine race: Mesopotamia, then Anatolia, then Balkans, then the Alps.
  • Nordic race: N coast of Caspian Sea, then N coast of Black Sea, then N Europe.
That's the sort of racial theorizing that was common in the late 19th cy. and the early 20th cy. --  Mediterranean race and  Alpine race and  Nordic race

Such "races" have been totally discredited by gene and genome sequencing. What one finds instead is that we are mutts, mixtures of people who departed from different places. In fact, population geneticists prefer to avoid *all* racial classifications, even ones that have some support in genetics. For example, the Caucasoid "race" has some support, but geneticists prefer to say "Western Eurasians", alongside similar designations, to try to be as noncommittal as possible.

Then trying to claim Celtic-Slavic connections. But such connections could be more general Indo-European ones.

The subgrouping of the Indo-European family has been a contentious issue for decades, but lexicostatistics, compring word lists and doing statistics, has come to our rescue. Balto-Slavic is a real grouping, and it clusters with Indo-Iranian, while Celtic clusters with Italic and Germanic, and Greek clusters with Armenian.
Serafimov (2014) shows that Gaulish is the most similar Celtic language with Slavic languages, ...
Because it's the oldest one, the one that retains the most ancestral Indo-European features.
It shows in particular the proximity of Gaulish with Slavic languages of the Balkans, as Bulgarian and Slovenian, to which I would add Bosnian-Serbian-Croatian-Montenegrin (BSCM) and Czech.
So Gaulish is closer to South Slavic rather than to West Slavic or East Slavic?
Apart from lexical concordances, basing on Piqueron (2015), Gaulish shares with Slavic many grammatical concordances. There are three genders of words in Gaulish, masculine, feminine and neutral, as in Slavic. Both are flexional with similar cases. There are imperfective and perfective verbs in Gaulish as in Slavic, ...
Author Xavier Rouard needs to take a crash course in Indo-European linguistics.

Having three genders is shared by most of the older Indo-European languages. There is nothing special about Gaulish and Slavic when Latin, Greek, Sanskrit, and Germanic also have that three-gender system.

Imperfective and perfective verbs? Proto-Indo-European had that kind of aspect system, and Proto-Slavic speakers invented a new one.
 
Personal pronouns, possessive adjectives and prepositions are similar in Gaulish, Slavic, Dravidian and Burushaski.
It just so happens that I recently made a cheat sheet of language families' pronoun systems.

Gaulish is very obviously Indo-European, though it generalized oblique stems 1s *me-, 1p *nos-, 2p *wos- to the nominative case, as Latin speakers had done with 1p and 2p. But was it Proto-Italo-Celtic speakers who did that with 1p and 2p?

Burushaski? Kirill Babaev's collection of pronouns states:
1s *Za, poss a- -- 1p mi -- 2s un, obl gu/o- -- 2p ma
Wiktionary also has a list, but it is unsourced.

Once Again on the Comparison of Personal Pronouns in Proto-Languages by Kirill Babaev
Appendix: Burushaski Swadesh list - Wiktionary, the free dictionary

Dravidian? I could easily find the pronouns in Wiktionary.

Proto-Dravidian:
1s *yân, *nân -- 1px *yâm -- 1pi *nâm
2s *nîn -- 2p *nîm
The Dravidian langs distinguish between inclusive and exclusive we, whether including the one addressed or not.

this *i- -- that (near) *u- -- that (far) *a- -- who? *yâ-
with suffixes (masculine vs. non-masculine, singular vs. plural)
masc sg *-wantu -- pl *-war
nmsc sg *-tu -- pl *-way

Proto-Dravidian had three-term demonstratives, like Spanish, instead of two-term ones, like English.

Sources: Reconstruction: Proto-Dravidian/yĀn - Wiktionary, the free dictionary and Reconstruction: Proto-South Dravidian/ñān - Wiktionary, the free dictionary

For comparison, here is Proto-Indo-European:
1s *ego-, obl *me- -- 1p *wei, obl *nos- -- 2s *tuH, obl *te- -- 2p *yuH, obl *wos-
this, that: *so, *to- -- *i-
who?, what? *kwi-, adj *kwo-

obl = oblique form (non-nominative)

Source:  Proto-Indo-European pronouns

Much more in common with Uralic, Turkic, Mongolic, and Tungusic than with Dravidian.
 
This author claims that Modern French has words close to Burushaski ones, an assertion I checked with Wiktionary's Burushaski list. French maman is a term of address, not the general term for mother, mère. The Burushaski word for mother is not mama but nani, nana. Etc.

He did find a few coincidences, like French je /Zë/ "I" and Burushaski ja /dZa/, but that's the nominative or subject form in French, and the other forms, oblique and possessive, all have stem m-.

About 50 words as tata, dad, mama, mum, mako, child, viro, man, bena, women, meno, think, gabi, take, da, give, beru, carry, cleu, hear, kahla, speak, edo, eat, itao, go, aro, plough, gabala, head, aedu, fire, bergo, hill, corro, summit, temos, dark, kolo, wheel, bitu, life, maros, big, which can be found in many languages as Gaulish and Dravidian, could come from this original language.
Appendix:Swadesh lists - Wiktionary, the free dictionary
I'll use Proto-Celtic and Proto-Dravidian, with Proto-INdo-European in parens:

father, *phatir (*ph2ter), *appa -- mother, mâtîr (*meh2ter), *amma -- son (child), *makwos, *makantu -- man (male), *wiros (*wiHros), (?) -- woman, *benâ (*gwenâ), *pen -- to think, (variable) -- to take, (variable) -- to give, *dati (*dedeh3ti < *deh3-), *ciy- -- to carry, (?) -- to hear, *klinutor (*klnewti < *klew-), *kuntV -- to speak, (variable) -- to eat, *essi _ *phiteti (*h1ed-) , *uHn _ *tiHn -- to walk (go), (variable) -- plow, *aratrom (*h2erh3trom < *h2erh3-), *namkol -- head, *cwennom, *talay -- fire, *tephnets (*tep- "to be warm, hot"), (?) -- mountain (hill), *moniyo-, (?) -- peak (summit), (?) -- darkness (dark), *temeslos (*temH-), *cînkk- -- wheel, *rotos (*Hreth2- "to run"), (all borrowed from Sanskrit cakra "wheel, disk, circle") -- to live (life), *biwo- (*gweyh3-), *man- -- large (big), *mâros (*meh2- "good"), (?)

For Wiktionary translations, I used Irish and Welsh, and Tamil and Telugu and Kannada and Malayalam.

I found a few matches, but not as many as in that article.
 
A Bayesian phylogenetic study of the Dravidian language family | Royal Society Open Science
and
Novel 4,400-year-old ancestral component in a tribe speaking a Dravidian language | bioRxiv

From the first one,
Our results indicate that the Dravidian language family is approximately 4500 years old, a finding that corresponds well with earlier linguistic and archaeological studies. The main branches of the Dravidian language family (North, Central, South I, South II) are recovered, although the placement of languages within these main branches diverges from previous classifications. We find considerable uncertainty with regard to the relationships between the main branches.
Most Dravidian speakers now live in the southern part of the Indian subcontinent, with scattered populations in the north. But there is plenty of linguistic evidence that Dravidian langs were earlier spoken over much of the north, evidence of Dravidian substrates in the Indo-European langs of the north, all the way back to Vedic Sanskrit, in what is now Pakistan.

What was the Dravidian homeland? The Indus Valley (Pakistan)? The Southern Neolithic Complex of around 4,500 years ago?
 
The closest relative to Dravidian? The most prominent candidate is Elamite, spoken in SW Iran over 2800 - 300 BCE.  Elamo-Dravidian languages - though there are some grammatical parallels, there is not much vocabulary in common.

(PDF) Elam: a bridge between Ancient Near East and Dravidian India?: Language Change and Cultural Transformation
proposing that Elamite is closer to Afroasiatic instead.

On the Genetic Affiliation of the Elamite Language | George Starostin - Academia.edu
Closet to Nostratic and Afroasiatic than to Sino-Caucasian?

This putative relationship suggests this migration scenario:

Eastern Fertile Crescent (Iraq) -> S Iran -> NW Indian Subcontinent (Indus Valley, Pakistan) -> most of the rest of the subcontinent.

Current perspectives and the future of domestication studies | PNAS -- domestication started in S Asia a few millennia later than in the Fertile Crescent, though still a few millennia before the estimated origin of Dravidian.
 
The closest relative to Dravidian? The most prominent candidate is Elamite... Elam: a bridge between Ancient Near East and Dravidian India?: Language Change and Cultural Transformation
propos[es] that Elamite is closer to Afroasiatic instead.

On the Genetic Affiliation of the Elamite Language | George Starostin - Academia.edu
Closet to Nostratic and Afroasiatic than to Sino-Caucasian?

This putative relationship suggests this migration scenario:

Eastern Fertile Crescent (Iraq) -> S Iran -> NW Indian Subcontinent (Indus Valley, Pakistan) -> most of the rest of the subcontinent.
Current perspectives and the future of domestication studies | PNAS -- domestication started in S Asia a few millennia later than in the Fertile Crescent, though still a few millennia before the estimated origin of Dravidian.

I don't know about a Dravidian - Afroasiatic link, but there was ancient contact between Egypt and the Indus River Civiilzation:
In 2750 BCE, the Pharaoh Sahure sent ships to the Land of Punt, which is believed to be peninsular India. This indicates early trade and cultural connections between the two civilizations.
This date is *before* Egypt's Old Kingdom, and *before* the "mature form" of the Harappan civilization. But I recall a conjecture that the Dravidian language arrived (from Africa) AFTER the Indus River culture was already thriving.

Is there some snake motif that strengthens the case for Egypt-India connections? May be hard to identify since snake worship was present in many or most early civilizations.
 
(PDF) Rice and the Austroasiatic and Hmong-Mien homelands
by George van Driem

Arguing for southern China.

Phylogenetic evidence reveals early Kra-Dai divergence and dispersal in the late Holocene | Nature Communications also proposes S China, and the Austronesian homeland, Taiwan, is just offshore from S China. Proto-Tai, Proto-Kra-Dai, and Proto-Austronesian have several reconstructed rice-related words.

So one concludes that the Proto-Austric speakers lived in S China and domesticated rice.
 
Inferring language dispersal patterns with velocity field estimation | Nature Communications - looks like a very interesting paper.

But I struggled to understand its method. Each language's velocity vector is the rate of change of each of its parameters, coded as 0 or 1 for each language feature. I was unable to get a clear picture of what they used as past values, however. It also was not very clear how the authors got physical-space directions of change. They did a principal components analysis of the data, finding its greatest directions of variation in state space. They then translated the langs' vectors from state space to physical space, though it is not very clear to me how they did that. With these directions, they could then look for a point of origin.

They tested their method on some synthetic datasets, and they applied it to four language families: Indo-European, Sino-Tibetan, Bantu, and Arawak (N South America).

Their methods finds:
  • IE: NE Anatolia
  • ST: 200 km SE of Lanzhou on the eastern half of the Yellow River in China
  • Bantu: S Cameroon
  • Arawak: N Bolivia
That is a revival of Colin Renfrew's Neolithic-farmer hypothesis of Indo-European, a hypothesis that has big problems with Wörter und Sachen (words and things): words for horse, wheel, wheel hub (navel), axle, wagon shaft (thill), yoke, wool, ... none of which is in evidence in early-Holocene Anatolia. Indo-European wheel words – revised and Language and Time 2: Wool, Wheels, and Proto-Indo-European - The Horse, the Wheel, and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World and Language Log » Horse and wheel in the early history of Indo-European

This method ought to be subject to real-world tests as well as simulated ones, tests using language families with known homelands and dispersion histories. Romance is an obvious one, because of abundant documentation of the rise and fall of the Roman Empire, complete with abundant documentation of Pre-Proto-Romance: Latin. Other good ones are Germanic, Slavic, and Austronesian.

Returning to the paper, Sino-Tibetan and Bantu are about right, however, and I don't feel confident enough to say anything about Arawak.
 
Back
Top Bottom