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

Language as a Clue to Prehistory

Found this one:

() - Indo_Uralic.pdf by Fredrik Kortlandt

"Károly Rédei (1986) lists 64 words which were supposedly borrowed from Indo-European into Uralic at an early date. The material is divided into three groups: 7 Proto-Uralic (PU) etymologies, 18 Finno-Ugric (FU) etymologies, and 39 Finno-Permian (FP) and Finno-Volgaic (FV) etymologies. "

But PIE and PU speakers lived too far apart: Ukraine and the northern Ural Mountains.

"Secondly, the number of verbs in the oldest material is too large to support the hypothesis that they were borrowed: 3 out of 7 (43%) in the first group, 5 out of 18 (28%) in the second group, and 2 out of 39 (5%) in the third group."

"Thirdly, the derivation of the Proto-Uralic forms from their alleged Indo-European sources involves considerable formal difficulties." Then discussing the four nouns in that group.
  • PU *nime and PIE *h3neh3mn > *nômn "name"
  • PU *sene (*sone) and PIE *sneh1wr, gen. *snh1wens "sinew, tendon"
  • PU *waske "some metal" FK considers Tocharian A wäs B yasa "gold" to be a Samoyedic borrowing
  • PU *wete and PIE *wodr, gen. *wedns "water"
Then the eight verbs:
  • PU *mighe- "to give, sell" ~ PIE *mey- "to exchange" > *meytH- > Latin mûtâre "to move, change"
  • PU *muske- "to wash" ~ PIE *mesg- "to dip, sink" > Lat. mergere
  • PU *toghe- "to bring, get, give" ~ PIE *deh3- "to give" > Lat. dâre
  • PFU *aya- "to drive, hunt" ~ PIE *h2eg- "to drive" > Lat. agere
  • PFU *kane- "to sprinkle, pour, throw, dig" ~ PIE *kh2en- "to dig" > Sanskrit khánati "he/she/it digs"
  • PFU *teke- "to do, make" ~ PIE *dheh1- "to put, place, do, make" > Greek tithênai "to put, place", English "do"
  • PU *wetä- "to lead, direct, pull" ~ PIE *wedh- "to lead, bind, pledge" > English "to wed"
  • PFU *wighe- "to take, carry" ~ PIE *wegh- "to bring, transport" > Latin vehere "to carry, transport", English "way", "wagon"
Not much, but it's something.
 
Found this one:

() - Indo_Uralic.pdf by Fredrik Kortlandt

"Károly Rédei (1986) lists 64 words which were supposedly borrowed from Indo-European into Uralic at an early date. The material is divided into three groups: 7 Proto-Uralic (PU) etymologies, 18 Finno-Ugric (FU) etymologies, and 39 Finno-Permian (FP) and Finno-Volgaic (FV) etymologies. "

But PIE and PU speakers lived too far apart: Ukraine and the northern Ural Mountains.

"Secondly, the number of verbs in the oldest material is too large to support the hypothesis that they were borrowed: 3 out of 7 (43%) in the first group, 5 out of 18 (28%) in the second group, and 2 out of 39 (5%) in the third group."

I'm no statistician, but 3 out of 7 seems a bit shallow for a statistical argument, no?
 
Merritt Ruhlen pursued an interesting relationship pattern for child/son/daughter and sibling/brother/sister; he found it in EVERY language family of the Americas and was able to guess its ancestral form:

A very long Chapter from Ruhlen's books finds that certain relationship words "represent[] a diagnostic trait of Amerind comparable in value to the Amerind pronominal pattern na ‘I’/ma ‘thou,’ whose importance Greenberg and others have stressed."

Merritt Ruhlen said:
I would like to suggest that these diverse forms have all evolved from a system that consisted of three terms in Proto-Amerind: *t’ina ‘son, brother,’ *t’una ‘daughter, sister,’ and *t’ana ‘child, sibling.'
. . .
Berman (1986: 421) concludes that Yurok “ts¯an- ‘young’ is related to tsin ‘young man’ cited above. I believe that one of these is an old changed form of the other, but I do not know which is which.” If one considers only Yurok, then no explanation of these forms is possible. However, when Yurok is placed in the wider context of Amerind, the source of these related forms is a trivial consequence of the Proto-Amerind system of gender ablaut outlined in this chapter.

Speaking of Omaha kinship system -- one of several such systems -- Thai uses a hybrid system. Four words are used for the eight cases
{mother,father}'s {older,younger} {brother,sister}​
BUT when parent is older than his/her sibling, it is parent's gender that determines the proper word. When parent is younger than his/her sibling, it is the sibling's gender that determines the proper word.

Living in low-mobility rural Thailand and interested in genealogy, I discovered various relationships. For example, my father-in-law's cousin was the grandfather of a woman who lived 17 km from us -- we knew her only because she had a Farang boyfriend. So they were first cousins twice-removed. That's very cumbersome to say in English, but in principle, Thai has an elegant way to express that relationship.

Thai has, supposedly, separate words -- all single-syllable! -- for child, grandchild, great-grandchild, great-great-grandchild and great-great-great-grandchild!
lûuk lǎan lěen lʉ̂ʉ lʉ̂ʉp lʉ̂ʉt

I write "supposedly" because I've heard only "lûuk lǎan lěen" in ordinary conversation, with "lʉ̂ʉn" volunteered when I show interest in the words. The word for "first cousin" in Thai is "lûuk pîi lûuk nɔ́ɔng" ("child older (sibling) child younger (sibling))." Shouldn't the first cousin twice-removed be "great-grandchild older child younger"? A second cousin should be simply "grandchild older grandchild younger." Yet when I proposed such phrases I was met with blank stares or the comment "We just say 'not a relative'." :cool:

By the way, the word for niece/nephew (by a younger sibling) means "subordinate; underling; minion; henchman."
 
"Károly Rédei (1986) lists 64 words which were supposedly borrowed from Indo-European into Uralic at an early date. The material is divided into three groups: 7 Proto-Uralic (PU) etymologies, 18 Finno-Ugric (FU) etymologies, and 39 Finno-Permian (FP) and Finno-Volgaic (FV) etymologies. "

...
"Secondly, the number of verbs in the oldest material is too large to support the hypothesis that they were borrowed: 3 out of 7 (43%) in the first group, 5 out of 18 (28%) in the second group, and 2 out of 39 (5%) in the third group."

I'm no statistician, but 3 out of 7 seems a bit shallow for a statistical argument, no?
I agree on small-number statistics, though this number is for all of Uralic, anything present in its two major branches, Finno-Ugric and Samoyedic. The latter is rather divergent in vocabulary, giving a risk of something in Proto-Uralic only being preserved in Proto-Finno-Ugric.

Adding the numbers gives 3+5=8 out of 7+18 = 25, or 32%. Better numbers, and still a high fraction of verbs.

The Uralic family has 9 primary branches, from east to west: Samoyedic, Hungarian, Mansi, Khanty, Permic, Mari, Mordvin, Permic, Saamic. Traditional subgrouping:
  • Samoyedic
  • Finno-Ugric
    • Ugric
      • Hungarian
      • Ob-Ugric
        • Mansi
        • Khanty
    • Permic
    • Mari
    • Western Uralic
      • Mordvin
      • Finnic
      • Saamic

As to verbs being less easily borrowed, that is supported by research: Borrowability and the notion of basic vocabulary and for a PDF, (PDF) Borrowability and the Notion of Basic Vocabulary -- in the authors' sample, and using words with recognized heritage, about 31.2% of nouns were borrowed, but only 14.0% of verbs, 15.2% of adjectives and adverbs, and 12.1% of function words. Verbs were only 1/4 of their sample, with most of the rest being nouns.
 
Merritt Ruhlen pursued an interesting relationship pattern for child/son/daughter and sibling/brother/sister; he found it in EVERY language family of the Americas and was able to guess its ancestral form:
...
*t’a/na ‘child, sibling.’
*t’i/na ‘son, brother’
*t’u/na ‘daughter, sister’
Notice the vowel alternation: a (both) ~ i (male) ~ u (female)
Is this gender-ablaut system present in other words? MR doesn't mention that in his article.

Speaking of Omaha kinship system -- one of several such systems -- Thai uses a hybrid system. Four words are used for the eight cases
{mother,father}'s {older,younger} {brother,sister}​
BUT when parent is older than his/her sibling, it is parent's gender that determines the proper word. When parent is younger than his/her sibling, it is the sibling's gender that determines the proper word.
As preparation, I considered brother and sister, then father and mother.

The primary distinction is age:
pîi "elder sibling"
nɔ́ɔng "younger sibling"

with gendered
pîi-chaai "elder brother: male elder sibling"
nɔ́ɔng-chaai "younger brother: male younger sibling"
pîi-sǎao "elder sister: female older sibling"
nɔ́ɔng-sǎao "younger sister: female younger sibling"

Father and mother are pɔ̂ɔ and mɛ̂ɛ coexisting with borrowings from Sanskrit or Pali bì-daa (< pitar-) and maan-daa (< mâtar-)

Turning to uncles and aunts, I find
lung "parent's elder brother"
bpâa "parent's elder sister"
aa "father's younger sibling"
náa "mother's younger sibling"

So it's a system where the older one has a gender distinction while the younger one doesn't.

Turning to grandparents,
bpùu "father's father"
dtaa "mother's father"
yâa "father's mother"
yaai "mother's mother"

Living in low-mobility rural Thailand and interested in genealogy, I discovered various relationships. For example, my father-in-law's cousin was the grandfather of a woman who lived 17 km from us -- we knew her only because she had a Farang boyfriend. So they were first cousins twice-removed. That's very cumbersome to say in English, but in principle, Thai has an elegant way to express that relationship.

Thai has, supposedly, separate words -- all single-syllable! -- for child, grandchild, great-grandchild, great-great-grandchild and great-great-great-grandchild!
lûuk lǎan lěen lʉ̂ʉ lʉ̂ʉp lʉ̂ʉt
Those words look similar, so it seems like they were derived from each other, like with compounds like "child-child" > "grandchild". Thai uses noun-adjective order, so the second member of the compound could have been reduced to the final consonant.

The words for "son" and "daughter" are lûuk-chaai "male child" and lûuk-sǎao "female child".

I write "supposedly" because I've heard only "lûuk lǎan lěen" in ordinary conversation, with "lʉ̂ʉn" volunteered when I show interest in the words. The word for "first cousin" in Thai is "lûuk pîi lûuk nɔ́ɔng" ("child older (sibling) child younger (sibling))." Shouldn't the first cousin twice-removed be "great-grandchild older child younger"? A second cousin should be simply "grandchild older grandchild younger." Yet when I proposed such phrases I was met with blank stares or the comment "We just say 'not a relative'." :cool:
 
Thus, in Thai, older sorts of relatives are more likely to have lexically-separate gender distinctions than younger ones, which more likely have <gender> <generic-term>. So I decided to check more generally.
  • Indo-European: all separate
  • Latin: children: same
  • Spanish, Portuguese, Greek: siblings: same
  • Turkic: parents, elder siblings: different, younger siblings, children: same
  • Mongolian: all separate, though with elder brother, elder sister, younger siblings
  • Tungusic: parents, elder siblings: separate, younger siblings, children: same
  • Dravidian: parents, siblings: separate, children: same
  • Nakh: parents: separate, siblings, children: same
  • Abkhaz-Abaza: parents: separate, siblings, children: same
  • Semitic: parents: separate, siblings, children: same
  • Egyptian: parents: separate, siblings, children: same
  • Austronesian: parents: separate, siblings, children: same
  • Malayo-Polynesian: parents, siblings: separate, chidren: same
  • Polynesian: all same
There is sometimes a lot of variation within a family. But overall, one finds paterns
  • All separate
  • Parents, siblings: separate, children: same
  • Parents, older siblings: separate, younger siblings, children: same
  • Parents: separate, siblings, children: same
  • All same
with only a few exceptions, like Greek and Mongolian.

Is the gendering of elder relatives more psychologically salient than the gendering of younger relatives?
 
Thus, in Thai, older sorts of relatives are more likely to have lexically-separate gender distinctions than younger ones, which more likely have <gender> <generic-term>. So I decided to check more generally.
  • Indo-European: all separate
  • Latin: children: same
  • Spanish, Portuguese, Greek: siblings: same
  • Turkic: parents, elder siblings: different, younger siblings, children: same
  • Mongolian: all separate, though with elder brother, elder sister, younger siblings
  • Tungusic: parents, elder siblings: separate, younger siblings, children: same
  • Dravidian: parents, siblings: separate, children: same
  • Nakh: parents: separate, siblings, children: same
  • Abkhaz-Abaza: parents: separate, siblings, children: same
  • Semitic: parents: separate, siblings, children: same
  • Egyptian: parents: separate, siblings, children: same
  • Austronesian: parents: separate, siblings, children: same
  • Malayo-Polynesian: parents, siblings: separate, chidren: same
  • Polynesian: all same
There is sometimes a lot of variation within a family. But overall, one finds paterns
  • All separate
  • Parents, siblings: separate, children: same
  • Parents, older siblings: separate, younger siblings, children: same
  • Parents: separate, siblings, children: same
  • All same
with only a few exceptions, like Greek and Mongolian.

Is the gendering of elder relatives more psychologically salient than the gendering of younger relatives?
In many cultures babies are effectively sexless beings, fashion only starts to diverge around ages 4 to 8 and the very word for baby/newborn is often neuter in languages with three-way grammatical gender. Arguably, living in large families, younger relatives are folks you've known as babies while older relatives were already being gendered when you came around.
 
Looking back at the work of Alain Matthey de l'Etang and Pierre J. Bancel and comparing it to kinship types, they find a distribution of words most consistent with Iroquois-like kinship, where mother's sister = mother and father's brother = father, and mother's brother and father's sister have words separate from these. In particular, mama papa baba nana tata dada for parents and related elders the same sex as them, and kaka and aya for related elders the opposite sex from them. 31. Why Kaka and Aya? by Merritt Ruhlen, who proposed both forms as "global etymologies".

Here again, the explanation is likely psychological saliency. A mother's-side female elder may seem like another mother, and a father's-side male elder may seem like another father. A mother's side male elder and a father's side female elder would not fit in very well, thus explaining Iroquois-like kinship and the Etang-Bancel pattern.
 
Some examples of this pattern:
  • English: father mother -- brother sister -- son daughter
  • Italian: padre madre - fratello sorella - figlio figlia
  • Spanish: padre madre - hermano hermana - hijo hija
The Italian words for brother and sister have a diminutive suffix: "little brother", "little sister"

Psychological saliency can also explain the origin of color words, though it's rather clearly perceptual saliency.
This research involves finding color terms that are
  • Generic, used for everything with that color. No "blond".
  • Not a special sort of word. No "crimson" or "scarlet".
  • Not recently borrowed.
  • Not a descriptive word. No "salmon-colored".
  • Lexically simple. No compounds, like "blue-green", or derivations, like "light red" or "reddish".
and then checking their meanings by having native-speaker informants identify which colors on a color chart have which color names. Berlin's and Kay's research involved using Munsell color chips.

This research found a sequence of adding color words, starting with light and dark and repeatedly splitting colors to get the large generic-color vocabularies that some langs now have.
 
 Color term

The first generic colors are light and dark. Our ancestors started with such words, and if they wanted to be more specific, they'd say berry-colored or straw-colored or leaf-colored or sky-colored or similar sorts of words.

The first set of word additions is of white, red, yellow, green, blue, and black.
  • Light: W-R-Y . dark: G-Bu-Bk
  • W . R-Y . G-Bu-Bk
  • W . R . Y. G-Bu-Bk ... W . R-Y . G-Bu . Bk ... W . R . Y-Gr-Bu - Bk
  • W . R . Y . G . Bu-Bk ... W . R . Y . G-Bu . Bk ... W . R . Y-Gr . Bu . Bk
  • W . R . Y . G . Bu . Bk
The next color added is usually brown, and that one is followed by such additional ones as gray, orange, purple, pink, and light blue.

Distinguishing green and blue is usually done late in this sequence.

Without a lot of generic color terms, those terms in use are often broad and imprecise by present-day standards. For example, "red" hair is typically orange or orange-brown or light brown. This usage goes back a long way:  Red hair noting redhead, n. & adj. meanings, etymology and more | Oxford English Dictionary - "OED's earliest evidence for redhead is from 1510, in the writing of J. Stanbridge." Such people are sometimes called "carrot tops", after that similarly-colored vegetable.

Related to that is the very limited color vocabulary of the Homeric epics, the Iliad and the Odyssey. That led to speculation that their putative author, Homer, was color-blind. Leaving aside the issue of whether there was some well-defined historical "Homer", a more plausible solution is a shortage of generic color words.
Gladstone notes that, compared to modern writers, Homer rarely mentions color, and what is mentioned is mostly limited to shades of black and white, with red, yellow, and green making only occasional appearances. Black is mentioned almost 200 times, white about 100. Red is mentioned fewer than 15 times, and yellow and green fewer than 10. Moreover, Homer’s descriptions of color can be, at times, completely bizarre: skies the color of bronze, stars are an iron or copper hue, sheep wool and ox skin appear purple, horses and lions are red, and honey glows green. Most conspicuous, however, Gladstone noted the complete absence of the color blue. Nothing is ever described as “blue.”

But Homer’s blindness could not be an explanation for the strange use of color in the Iliad and Odyssey. The existing texts record stories from a longstanding oral tradition. Moreover, once Gladstone sifted through Homer’s texts, he also analyzed the descriptions in other ancient Greek texts and found they too had a conspicuous lack of color terms, limited to mostly shades of black and white—and again, a total absence of the color “blue.” The word didn’t even exist. Did the Homeric Greeks have defective color vision? Was there something physically different about their eyes? Indeed, that was Gladstone’s conclusion: “[The] organ of colour and its impressions were but partially developed among the Greeks of the heroic age.” The ancient Greeks, according to him, were color blind.
That seems very unlikely, because they shared the genetics of color vision. In fact, this genetics goes back to some early Old World simian, some 30 million years ago, with the duplication of a color-receptor gene.  Evolution of color vision
Gladstone was on to something with his statistical compilation of color use in ancient Greek literature, but as it turns out, his study was a bit narrow: the ancient Greeks were not alone in their limited color descriptions, nor in the conspicuous absence of the color “blue.” Expanding upon Gladstone’s research, philosopher and philologist Lazarus Geiger found the same phenomenon in ancient Hebrew literature, Assyrian texts, Icelandic sagas, the Koran, ancient Chinese stories, Hindu Vedic hymns and Indian epics such as the Mahabharata. It is as though the entire ancient world were living in murky world of black and white, basking under heavy, brazen bronze skies, interrupted on rare occasion by flashes of red or yellow. The only ancient culture to have a word for blue was the Egyptians, as they developed the first synthetic pigment, Egyptian blue (the secret of its manufacture was lost in Roman times, but is thought to have been derived from heating together a quartz sand, a copper compound, calcium carbonate, and a small amount of alkali).
The page has a graph of counts of hue and lightness words. Lightness/hue word count: Iliad: 3.6, Odyssey 2.7, Bible 2.1, Shakespeare 1.5, 1800 average 1.2, 1858 average 1.2, 2008 average 0.9.

What happened? Increasing technology, making possible many different colors of the same sort of object. One would thus have to have lots of color words for distinguishing them.
 
 Indo-European vocabulary - PIE had words for light-colored and dark-colored, as one would expect. For red, it had *h1rewdh- for yellow and green *ghelh3- for brown possibly *bherH-

Turkic: white *ak, black *kara, red *kïrïl, yellow, white *siarïg, green *yal'ïl (< *yal' " young, fresh, green, raw, moist"), blue, green *kök

Central Asia has deserts Kara Kum "black sand" and Kyzyl Kum "red sand".

Not much success with "red" in Uralic, Sino-Tibetan, Semitic, Austronesian however.

So one can't look very far back with color words.

"White", "black", "red", "yellow", and "green" are part of the Swadesh list and similar lists, though no other color words.

The Swadesh wordlist. An attempt at semantic specification is very careful.

White: (1,1,1), snow, paper, babies' teeth, ... black: (0,0,0), coal, red: (1,0,0), blood, yellow: (1,1,0), egg yolk, green: (0,1,0), new grass.

The numbers are (red,green,blue) color-channel values with maximum 1. Not quite Munsell chips - Munsell Color System; Color Matching from Munsell Color Company and  Albert Henry Munsell and  Munsell color system

Absent from these lists is "blue". Checking on Wiktionary, I find it to be very unstable.

The Germanic langs' forms are all derived from Proto-Germanic *blêwaz but that's about as good as it gets in Indo-Europeandom. It has cognates like Latin flâvus "yellow" and Proto-Slavic *belu "white", from PIE *bhel- "shiny, white", colors that don't seem anything like blue in general, but they are bright like the sky. These words have additional cognates PGmc *blankaz "bright, white" > E "blank", *blakaz "burnt" > E "black", Latin fulgêre "to blaze", flagrâre "to burn", flamma "flame".

Latin had caeruleus < caelum "sky", but its descendants all have different word forms. French has bleu, a borrowing from a Germanic language during the Migration Period, one that got borrowed into English after the Norman Conquest. Spanish and Portuguese have azul, from the source of "azure": Arabic lāzuward < Persian lâjvard "lapis lazuli". Italian has (dark) blu, (light) azzurro, and Romanian has albastru < Latin albus "white". Sky-blue first also?

Turning to Slavic, we find a variety of forms.
  • Bulgarian sin, Slovak siny, Belarusian sini (dark), Russian sinij (dark), Ukrainian synij (dark) < Proto-Slavic *sinu "blue, gray"
  • Belarusian blakitny (light), Ukrainian blakytnyj (light) < Polish blekitny (light) < Middle High German blankhît "bright"
  • Czech modry, Slovak modry, Slovenian moder, Serbo-Croatian modar < Proto-Slavic *modru "blue, pale"
  • Polish niebieski < niebo "sky"
  • Russian goluboj (light), Ukrainian holubyj (light) < golub' "dove, pigeon"
  • Serbo-Croatian plav < Proto-Slavic *polvu "pale" < Proto-Balto-Slavic *palwas < PIE *pelH- > Latin pallidus, Proto-Germanic *falwaz > English fallow "left unworked (of farmland), is inactive"

So it looks like this word was unstable at first, from blue not being very distinctive at first. But as painting and dyeing became better, then the words became much more stable. It's like words for numbers: very stable once one has a lot to count, not very stable before that.
 
Last edited:
Checking on "green", it's well-preserved over the last few millennia, even if not as much before.
  • Proto-Germanic *grôniz < PIE *ghreh1- "to grow (of plants?)"
  • Latin viridis < virêre "to be growing, to flourish" < PIE *weys- "to increase"
  • Proto-Slavic *zelenu < PIE *ghelh3- "green, yellow" > Greek khloros "bright green, yellow"
Turning to "yellow", I find
  • Proto-West-Germanic *gelwaz, Proto-North-Germanic *gulaz < PIE *ghelh3- "green, yellow"
  • Latin flâvus < PIE *bhel- "shiny, white"
    • Catalan groc < Latin crocus (the flower)
    • Spanish amarillo, Portuguese amarelo < Latin amârus "bitter"
    • French jaune, Italian giallo, Romanian galben < Latin galbinus
  • Proto-Slavic *zhiltu < PIE *ghelh3- "green, yellow"
So "yellow" was not very stable in Latin's descendants.
 
Supposedly colors are introduced to primitive languages in a relatively fixed order:
Dark, Light, Red, Green, Yellow, Blue, Brown, ...


I cannot claim to be an expert on general Thai color names, but for many years I witnessed some color naming in rural Thailand, especially by my wife and wife's father. My wife's brother sometimes joked that they were "color-blind."
/kǐao/ 'green' was routinely used for hues that were obviously blue. (I now see that Wiktionary shows "1. green. 2. (archaic) blue; indigo." for /kǐao/.) Even dark skin was sometimes described as /kǐao/.

When describing hair color /dam/ 'black' was used for the 100% jet-black normal among ethnic Thais, while /dɛɛng/ 'red' was used for any color except jet black. My father-in-law used /dɛɛng/ for any hair color from very dark brown to pure yellow! (I suppose this weird nomenclature for hair color is obsolescent now, with a very large percentage of young or middlish Thai women dyeing their hair!)

When I go to the store to buy brown sugar, I must say 'sugar red.' (Adding confusion is that /náam-dtaan/ is the word for BOTH 'sugar' and 'brown', so 'brown sugar' would become the nonsense phrase 'sugar color sugar.' This word also denotes (as its compound constituents imply) the sweet (brown!) juice of the sugar-palm tree's fruit.)

With the use of 'red' and 'green' to describe colors that we would call 'brown' or 'blue' I see vestige of the primitive four-color model.

It's not that the Thais lack words for blue or brown. In fact, 'color rind mangosteen', 'color blood pig' and 'color smoke cigarette' were in frequent use to denote different shades of brown!
 
Drastic demographic events triggered the Uralic spread | John Benjamins

In particular, the  4.2-kiloyear even - a big drought in much of the world for as much as a century around 4,200 BP / 2200 BCE The drought was in N & C Africa, S Europe, the Middle East, N China, E & C North America, but there was increased rainfall in N Eurasia, S & SE Asia, W North America, South America.

This drought likely caused the ends of the Old Kingdom of Egypt and the Akkadian Empire of Mesopotamia.

The authors propose a connection with the Uralic speakers' initial dispersal from their homeland, which they propose is W Siberia near the Ural Mountains.

A millennium before this dispersal was another dispersal, the Proto-Indo-European one, the Yamnaya one, from what is now Ukraine. Yamnaya people went west into the Balkans (Illyrian? Albanian? Greek? Thracian? Phrygian?), east into Central Asia, settling down in the Tarim Basin (Tocharian), and north, where it spread westward and eastward, making the Corded Ware culture zone, from Belgium to Moscow (Celtic, Italic, Germanic, Balto-Slavic) A branch of it moved further eastward, becoming the Sintashta culture between the south end of the Ural Mountains, the Caspian Sea, and the Aral Sea. This was around 4,000 BP / 2,000 BCE. The Sintashta people spoke Proto-Indo-Iranian, they invented war chariots, and they then spread south and southwest from there, overrunning their old homeland as the Scythians and Sarmatians, and going into Iran, Afghanistan, Pakistan, and northern India.

Indo-Iranian borrowings in Uralic : Critical overview of sound substitutions and distribution criterion
A notable problem is the parallel borrowing of same Indo-Iranian words to various branches of the Uralic family. It is not always easy to distinguish the parallel borrowings from the earlier loanwords into the common proto-language, and in earlier research the parallel loans have not received enough attention, despite their key importance to the chronology of the loanwords.

Shrikant G Talageri: Uralic (Finno-Ugrian) and Indo-Iranian Connections
“The earliest layer of Indo-Iranian borrowing consists of common Indo-Iranian, Proto-Indo-Aryan and Proto-Iranian words relating to three cultural spheres: economic production, social relations and religious beliefs. Economic terms comprise words for domestic animals (sheep, ram, Bactrian camel, stallion, colt, piglet, calf), pastoral processes and products (udder, skin, wool, cloth, spinner), farming (grain, awn, beer, sickle), tools (awl, whip, horn, hammer or mace), metal (ore) and, probably, ladder (or bridge). A large group of loanwords reflects social relations (man, sister, orphan, name) and includes such important Indo-Iranian terms like dāsa ‘non-Aryan, alien, slave’ and asura ‘god, master, hero’. Finally a considerable number of the borrowed words reflect religious beliefs and practices: heaven, below (the nether world), god/happiness, vajra/‘Indra’s weapon’, dead/mortal, kidney (organ of the body used in the Aryan burial ceremony). There are also terms related to ecstatic drinks used by Indo-Iranian priests as well as Finno-Ugric shamans: honey, hemp and fly-agaric” (KUZMINA 2001:290-291).
from "KUZMINA 2001: Contacts Between Finno-Ugric and Indo-Iranian Speakers in the light of Archaeological, Linguistic and Mythological Data. Kuzmina E. E. in “Early Contacts between Uralic and Indo-European: Linguistic and Archaeological Considerations”. Ed. Carpelan, Parpola, Koskikallio. Suomalais-Ugrilainen Seura, Helsinki, 2001."

Also borrowed, words for 10, 100, 1000. For 100: Finnish sata, Hungarian száz -- like Sanskrit satám < PIE *kmtóm

As to kinship terms being borrowed, that happened to English, where uncle, aunt, cousin, nephew, and niece are all borrowed from Old French, and grand-<relatives> are imitations of similar Old French terms.
 
In the same time frame came the Seima-Turbino Transcultural Phenomenon (henceforth ST), an archaeological complex marked by distinctive bronze artifacts, especially symbols of power such as spearheads and axe heads, found across many archaeological cultures and along major waterways from the Altai to Scandinavia. In particular, tin from Altai mines made possible large-scale bronze production, with the forging and casting done there as well as in the southeastern Ural area and Southwest Asia. The most recent radiocarbon dates place ST between c. 4,200 and 3,900 BP, and somewhat later west of the Urals along the Volga (Marchenko et al. 2017; Krause et al. 2019; Chernykh 2008).

ST is evidently the archaeological signature of a waterborne trade network (Barfield 2009; Nichols & Rhodes 2018) that brought metal from Ural and Sayan mines and metal artifacts from Ural forgeries westward to Europe. Its heyday coincides rather closely with the 4.2 ka event. The trade network itself probably existed long before the Bronze Age and demonstrably continued into the Middle Ages as the Bulgar and Viking trade routes.
Not long afterward, Uralic speakers came into contact with Indo-Iranian ones, though by then, the Proto-Uralic community had fragmented into 9 main branches. From east to west: Samoyedic, Hungarian, Mansi, Khanty, Permic, Mari, Mordvin, Finnic, Saamic, with ages of around 1,000 to 2,500 years. Of these, Proto-Samoyedic speakers had the least contact with Indo-Iranian speakers, and the rest varying amounts of contact.

After these contacts, they spread northward, sometimes in relatively recent centuries.
In most places, local northern toponymy and vocabulary for tundra flora and fauna include words of non-Uralic origin, showing that today’s northernmost languages were the frontier languages in the spread (Aikio 2012; Helimski 2001a; Saarikivi 2006, in press). Additionally, the fact that this vocabulary is borrowed shows that not only PU and CU speakers, but also branch ancestor speakers, were unfamiliar with the tundra ecology and needed to borrow terms for it.
But the initial dispersion was different.
In contrast to the northward spreads, the initial Uralic spread was almost entirely east-west in direction, with daughter branches taking root along most of the east-west extent of the Volga and probably along the middle and upper Tobol, Irtysh, and Ob and the upper Yenisei (Figure 1). This spread appears to have been rapid, largely without substratum, and with minimal evidence of frontier expansion and isolation by distance. It was westward overall; only the Samoyedic branch probably did not take part in the westward spread and may have moved eastward instead.

The paper page links to Drastic demographic events triggered the Uralic spread: Supplements with details like what they used as clues for the Uralic homeland, and also various linguistic features.
 
How well can scientists map linguistic, archaeological and genetic events to each other?

By "genetic event" I'm especially thinking of tests on DNA from ancient skeletons; I'm especially interested in Y-chromosome haplogroups. Jean Manco used to have a wonderful website where she posted the results of every DNA test performed on ancient skeletons. She passed away; her final version is available on the Wayback machine. With DNA testing increasingly common it would be nice to have an updated list, but is there one? (Some of the best websites have been by amateurs, and often disappear. Perhaps with Search an inferior "professional" site would turn up . . . but probably behind a pay-wall.)

What's the relationship between languages and cultures in Scandinavia and the Baltic area? I'll show what I've gleaned, mostly from Wikipedia, listing relevant cultures in roughly chronological order.

Mesolithic Central Europe had mostly I2 Y-haplogroup, with G2a2 Y-haplogroup often found among the earliest farmers, whether Linear Ware or Impressed Ware. In many locales these Y-chromosomes have now been replaced by R1b or R1a; but this does NOT mean there was general gene replacement. A small band of strong male invaders can have high procreative and survival success and have a dramatic effect on Y-chromosome frequencies even when autosomal genes change little! The dominance is retained with an explicit or implicit caste system with caste inherited from father.

The Funnel-Beaker Culture (4300 BC - 2800 BC) was the first neolithic culture of northern Germany, noted because it lacked orderly garbage dumps compared with other sedentary cultures. Among several firsts are the earliest preserved wagon wheels (although the wagon was probably invented elsewhere). Their language is unknown but probably was NOT Indo-European, Uralic or proto-Basque. Their Y-chromosome is unknown but was probably I2 -- the chief haplogroup of mesolithic Central Europe. Funnel Beaker diminished when Corded Ware arrived, persisting for a few centuries in Sweden.

The Comb Ceramic Culture (4200 BC - 2000 BC) lived between the Baltic Sea and the Urals (and may have originated east of the Urals). They had little or no farming but used pottery apparently influenced by the Far East. In Finland they lived in teepees and hunted seals. Their Y-chromosome was R1a-YP1272, a branch of R1a that split off before the very early Indo-Europeans. Their trajectory was similar to that of Uralic people, but Wikipedia insists that their migrations were long before those of the Uralic speakers. Their language is unknown.

The Pitted Ware Culture (3500 - 2300 BC) were seafarers who thrived along the coasts of Scandinavia. They borrowed heavily from Comb Ceramic, but were very different genetically. While Comb Ceramic had Eastern European genes, Pitted Ware descends from the hunter-gatherers of Sweden; their Y-chromosome was I2. They hunted seals and land animals like elk, wild boar and beaver; fished and gathered plants; but didn't need to farm: they could trade with neighbors who did farm. Because of their seafaring skills, I think they were raiders (proto-Vikings?) who sometimes just took what they wanted. They came into conflict with (and tended to dominate?) other Scandinavians beginning with Funnel-Beaker.

The Corded Ware Culture (3000 BC - 2350 BC) dominated North Central Europe from the Rhine to the Volga and probably(?) spoke Indo-European language(s). Today R1b dominates the western part of the Corded Ware region, and R1a dominates in the East, but ancient skeletons tell a very different story, and various Y-chromosomes have turned up, including R1a in the West. Wikipedia sources use this genetic fact to argue that Corded Ware -- despite its strong similarities to Yamnaya culture -- was a "parallel" culture. The R1b Y-chromosome common in Corded Ware skeletons is R1b-Z2103 which, while possibly ancestral to Hittites and other very early Indo-European speakers, split from the main Western Indo-European clade (Yamnaya) about 4000 BC. Some centuries before Corded Ware there was the related Globular Amphora culture. Its major Y-chromosome seems to be I2. Its language is unknown although Gimbutas treats it as part of the "Kurgan horizon" which culminated in Yamnaya.

The Battle Axe Culture and Single Grave Culture basically just refer to Corded Ware of Scandinavia

The Bell Beaker Culture (2800 BC - 2000 BC) quickly dominated almost all of Western Europe. They spoke proto-Celtic or Italo-Celtic. Their R1b-P312 Y-chromosome originated in 2700 BC and now completely dominates Western Europe. About 3000 BC, R1b-U106 split off and became common in Germany and Sweden (although as mentioned this is a different clade than the R1b's found in ancient Corded Ware skeletons).

Balto-Slavic speakers moved into the Baltic area perhaps about 2000 BC? They had R1a Y-chromosome.

Uralic speakers arrived near the Baltic even later than the Balto-Slavic people? Many had N Y-chromosome.

The Nordic Bronze Age began about 2000 BC and was centered in Denmark and southern Sweden. It had strong contacts with the Myceneans of Greece!
Wikipedia said:
the Corded Ware constructed a series of defensive palisades ... which may be a sign of violent conflict between them and the Pitted Ware. [Funnel-beaker also built palisades to defend from Pitted Ware.] Though cultural influences of the Battle Axe culture are detectable in Pitted Ware burials, its peoples do not appear to have mixed with each other. By ca. 2,300 BC, the Pitted Ware culture had merged with the Battle Axe culture. The subsequent Nordic Bronze Age represents a fusion of elements from the Pitted Ware culture and the Battle Axe culture.

The Jastorf Iron Age began about 550 BC. While the Nordic Bronze Age is usually described as the origin of the Germanic languages, in fact proto-Germanic's alleged date is AFTER Jastorf.

- - - - - - - - -

Whew! I'd intended a much briefer summary, but once started I had to be thorough. And even so, I show more questions than answers.

The prehistoric cultures of the Scandinavian area interest me because of the light they may shed on the  Germanic substrate hypothesis. Germanic is unique among Indo-European language families in that it shows much divergence in lexicon, phonology and grammar; and is difficult to place into a "genetic" language tree.

Other Indo-Europeans conquered easily. Proto-Celtic speakers quickly dominated Western Europe; early Greek and Iranian warriors seemed invincible; and so on. BUT the Pitted Ware seafarers put up a fight, and Wikipedia concludes that the "Nordic Bronze Age represents a fusion of elements from the Pitted Ware culture and the Battle Axe culture." I think we must look to Pitted Ware to solve the mystery of the origin of the Germanic language family. "Sea, ship, sail, keel, eel, ice" are words that were probably borrowed into Germanic from a sea-faring people. Yet Wiki's  Germanic substrate hypothesis doesn't even mention Pitted Ware, arguably the most likely source of such words.

I think a pidgin language would have developed between Pitted Ware and Battle-Axe: the two cultures had much contact but remained too aloof from each other for bilingualism or (at least initially) creole. That pidgin language influenced the development of proto-Germanic.
 
This came out recently:
Calibrated weighted permutation test detects ancient language connections in the Circumpolar area (Chukotian-Nivkh and Yukaghir-Samoyedic)* | John Benjamins
by Alexei S. Kassian, George Starostin, Mikhail Zhivlov, Sergey A. Spirin

Some of the authors had previously done Indo-Uralic and Altaic, using the same method as in this paper, and they are taking on more: Samoyedic, Yukaghir, Chukchi-Kamchatkan, Eskimo-Aleut, Nivkh, Burushaski, Yeniseian, Na-Dene (Athabaskan, Eyak, Tlingit, Haida)

There is a version on BiorXiv: Circumpolar peoples and their languages: lexical and genomic data suggest ancient Chukotko-Kamchatkan–Nivkh and Yukaghir-Samoyedic connections | bioRxiv

The supplementary material contains details on the methods. The authors used the same consonant classes as previously, grouping by points of articulation, and they did the same scramble test as previously, with one difference. They used weights for the meanings of the unscrambled ones, weights equal to (maximum number of cognates) / (total members), and then averaged over families. If weighted by (say) the number of languages in each family, the order would be somewhat different. For each pair of langs, they scrambled each one relative to the other one, and they found the fraction of scrambled ones that does better than the unscrambled one.

Their weights: "I" 0.805, "you sg." 0.797, "2" 0.769, "eye" 0.738, "we" 0.725, "tongue" 0.706, "name" 0.667, "1" 0.634, approximately linear the rest of the way, to "many" at 0.219.
 
Oops, the scramble-test probability should be the maximum of the probabilities for each direction.

For some language families, they had difficulty in reconstructing the protolanguages' list members. So they added the lists for subfamilies with better-reconstructed ones. They tested their scramble test on these familes' subfamilies.

Chukchi-Kamchatkan (Chukotko-Kamchatkan)
Proto-Chukotkian - Proto-Itelmen (Kamchatkan) - < 10^(-7)

Eskimo-Aleut
Proto-Eskimoan (Inuit-Yupik) - Proto-Aleutian - 2.02*10^(-3)

Na-Dene: (((Athabaskan, Eyak), Tlingit), Haida) -- Haida treated separately in this paper
Athabaskan - Eyak - < 10^(-7)
Athabaskan - Tlingit - 0.01
Eyak - Tlingit - 1.17*10^(-4)
Haida - (Athabaskan-Eyak-Tlingit: AET) - 0.997


They found very low scramble-test probabilities for Chukchi-Kamchatkan and Nivkh - 4.2*10^(-5) - and Samoyedic and Yukaghir - 5.12*10^(-4).

They checked on Dene-Burusho-Yeniseian:
Yeniseian - Burushaski - 0.009
YB - Athabaskan - 0.013 - 0.066
YB - Eyak - 0.023 - 0.847
YB - AE - 0.006 - 0.289
YB - Tlingit - 0.557 - 0.364
YB - AET - 0.014 - 0.247

The numbers jump around rather oddly, and that was also evident in their overall table.
 
The Chukchi-Kamchatkan-Nivkh relationship supports a hypothesis in  Uralo-Siberian languages

The authors note some oddities of Uralic-Yukaghir. On the Uralic side, most of the parallels are present in Samoyedic, in one count, only 8 out of 43 are not present in Samoyedic, and 2 of those have Samoyedic sound shifts.

Turning to Indo-European, there are U-Y matches without IE parallels, but almost no IE-Y matches without U parallels.

Borrowings? Most of the vocabulary matches are of basic sorts of words, which does not fit.

The authors propose a scenario where some group of people end up speaking a neighboring group's language while retaining some basic words of its previous language.

Thus, in the Chukotko-Kamchatkan–Nivkh language pair the most likely scenario is genealogical relationship, implying that the Chukotko-Kamchatkan and Nivkh languages originate from a common proto-language. On the contrary, for the Samoyedic-Yukaghir language pairs, a contact scenario (possibly with an ancient language shift) is more probable.
 
Returning again to *puluga

How reconstructible is Proto-Trans-New-Guinean? by Andew Pawley
*(m,mb)elak

(a) Lightning, lightning flash' (N.), or 'to flash, lighten' (V.)
The gloss for forms below should be read as referring to a noun lightning, lightning flash' unless otherwise indicated by use of '(V.)', when the gloss should be 'to flash (of lightning)'.

...
(b) 'light, brightness' (N.), 'be light (as of fire or sun) (V.)'
Glosses should be read as referring to noun 'light, brightness' unless otherwise indicated.

...
A complicating factor is the likelihood of interference from sound symbolic associations. Suter (p.c.) points out that the sequences p-l and b-l appear to be sound symbolic for flashing, e.g. Proto Malayo-Polynesian *bilak *lightning', German Blitz, Latin fulgur, English flash. Some languages show a proliferation of look-alike terms in this semantic domain, e.g. Ono (Huon Peninsula) gbilap 'lightning', mapalak 'twinkle', walatak 'flash'. Some of the look-alikes in the *(m,mb)elak cognate set may be independent developments or reshaped by sound symbolism. p-l, b-l sound sound symbolism would favour the independent change of m-initial to b-initial forms rather than the converse.
Sound symbolism? How would that be?

Some words are sound-imitation or sound-symbolism words,. "Cuckoo", for instance, is likely from imitation of a cuckoo call. Such words are rare, however. Dogs are not usually called woof-woof or arf-arf, for instance.
 
Back
Top Bottom