Language as a Clue to Prehistory

[lpetrich]

Vaclav Blazhek: "Numerals of the World I: Hurrian Numerals"

Mentioning for comparison, Indo-European (Anatolian (Hittite, Luwian), Mitannian-Indo-Aryan), Semitic (Ugaritic, Eblaite, Akkadian), Sumarian, Elamite

Also North Caucasian: NEC (Nakh, Avar-Andian, Tsezian, Lakian, Dargi, Lezgian, Khinalug), NWC

Conclusions: 1, 2, 3, 4 - NEC counterparts. 5 = Sino-Caucasian "(palm of the) hand". 6 - Akkadian, 7 = (5)+2, 8 = (5)+3, 9 = (5)+4, 10 = NEC "hands, handful"

Vaclav Blazhek: "Numerals of the World I: Nilotic Numerals" - reconstructions for 1, 2, 3, 4, 5, 10


An interesting thing in words for numbers is the lack of explicit words for none of some number in them. 101 is "one hundred and one", not "one hundred, no tens, and one". None of some number is assumed from lack of mention of it.

The word "zero" is from French zéro, from Italian zero, from medieval Latin zephirum, from Arabic sifr "nothing" a translation of Sanskrit sunya "empty". Appropriate, since zero is the nothing number. Some langs use words from Latin nûllus "none", and some use native words, like Modern Greek midén "nothing".

 

[Swammerdami]

Jonathan Morris: "The myth of rapid linguistic change (debunked by the Romance languages)" - "If we analyze the change between the Classical Latin Swadesh list and the modern Romance Swadesh lists, we find that 70% of the change to Romanian, 75% of the change to French and 90% of the change to Spanish, is likely to have occurred by the end of the Roman empire."

At least two problems here.
(1) Am I not correct that by Julius Caesar's time, the Latin written language and the Roman vernacular had already diverged? Classical Latin descends from the former, the Romance languages from the latter.
(2) Assuming a constant "half-life" of 1000 years, 50% of words from 1 AD are preserved at 1000 AD, and 25% at 2000 AD. Two-thirds of the changes come in the first half of the time-span. This arithmetic mismatch is further increased by the fact that half-lives are NOT constant, even on the Swadesh list. As the easy-to-replace words are replaced, the average "half-lives" of the remaining words increase.

He dismisses outright the Kurgan hypothesis of the Indo-European homeland, as it might be called.

If we use the same ratio, we are asked to believe that at most 50,000 Bronze nomads and their descendants succeeded in imposing language change on most of Eurasia in 500-1,000 years, including 3 million people in Europe alone, so completely as to obliterate ail traces of substrate languages, even in marginal areas like Norway, when 300,000 Germanic tribefolk concentrated in an area about a twentieth of the size had had no lasting impact beyond a few borrowings and sound changes. Indeed, where did all their surplus largesse of wagons/cows/horses etc. that they needed to impress the natives actually come from?

Nonsense of course. That the "Kurgan hypothesis" is correct has been known with certainty for two decades or so. Facts should be used to tune theories of language change, not vice versa.

There are plenty of examples of languages being eliminated by newcomers. Two thousand years ago, Continental Celtic languages dominated Western Europe. None survive. Aboriginal Philippine languages were completely replaced by Austronesian, Pygmy languages by Bantu, etc. The "Paleo-European" languages have disappeared, with proto-Basque arriving in Spain circa 5500 BC with the sea-faring farmers of Cardial Ware. (Wikipedia still misleads on this.)

He is right about one thing: It is impressive how fast Bell Beaker culture, R1b Y-haplogroup, and Celtic language overwhelmed France and Spain in the late 3rd millennium BC. (This might be an interesting sub-topic.)

 

[Swammerdami]

A good example of a borrowed feature is classifiers / counter words / count words / measure words, common in eastern and southeastern Asia but rare elsewhere. It's more-or-less treating every noun as uncountable. Classifiers are in Chinese, Korean, Japanese, Vietnamese, Thai, and Burmese, and at least some of them have relatives that have no classifiers. For Chinese and Burmese (Sino-Tibetan), Tibetan, for Vietnamese (Austroasiatic), Khmer, for Thai (Austro-Tai), Austronesian, and for Japanese and Korean (Transeurasian / Broad Altaic), Turkic, Mongolic, Tungusic. Austro-Tai and Altaic are both speculative, but supported by some Swadeshian statistics. Also notable is that the classifier words themselves don't seem to be cognate across langs, indicating that it was the idea of using classifiers that was borrowed, and not the classifiers themselves.

I've been intrigued by Thai classifiers since I started learning Thai 40 years ago. (Time flies? )

Some classifiers describe an object's shape. The classifier for "pencil" is also used for the male sex organ!

English has the equivalent of classifiers for uncountable nouns, or when the unit of counting is greater than one: "Three heads of lettuce," "Two decks of cards." Thai is the same just with different word order: "Lettuce three head," "Card two deck." (Here the Thai classifier 'head' is the usual word for 'head', just as in English. The classifier 'deck (of cards)' is a rare word used in few other cases, if any.)

English has "pride of lions," "exaltation of larks," "gaggle of geese," etc. Were these just manufactured by a whimsical guy a few centuries ago? All but a few ("pride of lions," "school of fish," etc.) have fallen into disuse I think. Similarly many classifiers are now rarish in spoken Thai. Classifiers include words for "thing" and "piece"; these can often be substituted for a more formal classifier.

"Train" as in "railroad train" is implicitly classifier-like in English (since a train is a group), though the modified noun is omitted: "When does the last train depart?" In Thai this becomes "Railroad parade final depart when?" with the noun for "parade" doubling as classifier in this and few other contexts.

Classifiers are often just an ordinary word. In a bar you might order "Bottle beer three bottle." More traditional would be "Bottle beer three leaf." I don't know why the same classifier is used both for flat objects and for bottle-shaped objects. (The first "Bottle" is redundant in the first sentence and would probably be omitted. It is NOT redundant in the second if ordering by the glass is an option. If two bottle sizes are available, you'd say "Beer bottle large three bottle.")

The word for "body" is used as a classifier for many nouns (animal, chair, shirt, trousers, etc.) but NEVER for a human. To say "three elephants", "elephant three body" would be normal, but elephants are special in Thailand. "Elephant three rope" is for domesticated elephants. (This classifier is not used for other roped animals.) "Elephant one elephant" for a royal albino elephant. There's a special classifier meaning "herd of elephants."

 

[lpetrich]

Jonathan Morris: "The myth of rapid linguistic change (debunked by the Romance languages)" - "If we analyze the change between the Classical Latin Swadesh list and the modern Romance Swadesh lists, we find that 70% of the change to Romanian, 75% of the change to French and 90% of the change to Spanish, is likely to have occurred by the end of the Roman empire."

At least two problems here.
(1) Am I not correct that by Julius Caesar's time, the Latin written language and the Roman vernacular had already diverged? Classical Latin descends from the former, the Romance languages from the latter.

Not by much.  History of Latin Classical Latin is the dialect that emerged as the literary language in the late Roman Republic. It has some differences from earlier Latin, like the o-stem declension -os -om becoming -us -um, and (s) between vowels becoming (r), like honos, honosis > honor, honoris.

 Vulgar Latin gradually became different from the literary language.  Appendix Probi ("Probus's Appendix"), written around 300 CE, describes how to say the proper Classical Latin forms of a variety of words, with the erroneous forms being what became the Romance forms. oculus non oclus "eye" ...

 Proto-Romance language - around 300 CE?
 Phonological changes from Classical Latin to Proto-Romance
 Lexical changes from Classical Latin to Proto-Romance
Dating Classical Latin to around 100 BCE, that means around 400 years.

 Romance verbs - much of the Latin verb conjugation survived in the Romance langs, with the Continental Western ones inventing a new future tense: (infinitive)-have. However, the Romance langs have lost the passive personal endings.

 

[lpetrich]

John Bengtson: "Basque and the Other Mediterranean Languages" - comparing Basque to North Caucasian (NC) and Afrasian (Afroasiatic, AA)

JB finds lots of differences between Basque and AA, and a lot of resemblance to NC and other Sino-Caucasian langs.

Bsq has a lot of noun cases with a lot of consonants, something much more like NC, Burushaski, & Yeniseian than like AA. Bsq, NC, Bur even have compound case endings. Classical Arabic: a ~ i ~ u. Bsq has what look like fossilized class prefixes, like the (active) prefixes in the NC langs, while AA has feminine -(a)t. Bsq also has no AA-like ablaut plurals, like what Arabic has a lot of.

As to Basque-AA vocabulary, it's (1) specific resemblances to particular AA languages, pointing to contact and borrowing, (2) possible inheritance from some long-ago ancestor, and (3) coincidence.

Word forms considered diagnostic of AA: "blood", "bone", "tooth", "tongue", "horn" -- body parts, with some of the most stable words -- Bsq resembles NC and DC more generally rather than AA.

JB then notes the argument that Basque is a survivor of Paleolithic Europeans' langs. He proposes an alternative: that it is a survivor of Neolithic Europeans' langs, langs spread by farmers from the Middle East.

On Basque 1follow no less an authority than the great vasconist Rene Lafon, who posited that the people of the Basque Country and Aquitania adopted a foreign language from an immigrant population who brought a technologically superior culture."

Lafon identified this culture with copper-using, megalith-building immigrants near the end of the third millennium BCE. After conferring with an archcologist colleague, Peter Rowley-Conwy, I agree with the latter that a likelier candidate is the much earlier Cardial Culture, which arrived on the eastern Spanish coast around 5500 BCE. Recent archeological evidence suggests that the Cardial people, originally from Anatolia, arrived by boat from Italy by means of 'leapfrog' colonization round the South French coast. The name Cardial refers to Cardium edulis, a mollusk whose shells imprinted their clay artifacts. Besides the characteristic ceramics, the Cardial Culture included what the archeologists call a complete "Neolithic package" of cultural traits, including the use of domesticated plants and animals, and long distance trade of obsidian and other lithic material (Price 2000; Zapata et al. 2004; Peña-Chocarro et al. 2005).

Then a lot of vocabulary.

The linguistic evidence presented here indicates that the western Dene-Caucasian speakers of ca, 7500 years ago (linguistic ancestors of the present-day Basques, North Caucasians, and Burushos) had a well-developed Neolithic pastoral-agricultural culture, including the husbandry of large small cattle and the cultivation and milling of cereal grains and some other crops such as pulses.

How do we know that the Basques did not simply adopt these Dene-Caucasian Neolithic terms as loanwords, while retaining the rest of their original language intact? In fact the Neolithic terms have the same phonology and morphology as the most basic parts of the Basque lexicon.

That means that other Neolithic European farmers likely spoke Dene-Caucasian langs.

 

[lpetrich]

Mother Tongue 15:
Yuri E. Berezkin: "From Africa and back: some areal patterns of mythological motifs"

In the early nineties I began to create an electronic catalogue of world mythology and folklore. It now contains more than 45,000 abstracts of texts arranged according to ethnic groups or areal clusters of groups and to motifs (more than 1600 at the moment). With new publications processed and the number of the selected clusters increasing (813 in January 2011), the system becomes ever more sensitive to tendencies in distribution of motifs.

Then boreal vs. austral, much like Michael Witzel's Laurasia vs. Gondwana, but with austral including Southeast Asia and Central and South America.


Jonathan Sherman Morris: "The myth of rapid linguistic change III: The evidence from Greek"

He notes that Tsakonian is a surviving pre-Hellenistic dialect with unusual features. Its phonology is archaic, keeping a where Attic Greek turned it into ê (> Modern Greek i).

At the same time, Tsakonian has undergone extreme morphological simplification, with minimal case inflection (some nouns have no cases, others a single Nominative/Accusative form and a Genitive form) and the formation of the present and imperfect indicative with participles, like English but unlike the rest of Greek.

Also mentioning Griko in southern Italy.

Even taking the written evidence at face value, however, we indeed find that most of the major phonological changes between Classical and Modem Greek had already occurred well before the end of the Hellenistic period.

...
This Boeotian origin of most of the phonological changes which differentiate Modem Greek from Ancient Greek is intriguing and the explanation for this is not yet clear to me, although attempts to argue away such changes as parallel developments seem thoroughly unconvincing and the fact that there are so many innovations deriving from a single dialect hardly seems to be a random occurrence.

It's hard to see what his point is, however, other than that one should not expect constant rates of change, especially in some society that suffered some big upheaval.

About Latin / Romance, he didn't bother to try to prove that different parts of the Empire were colonized by soldiers from different parts of Italy. Ex-Oscans to Iberia, ex-Umbrians to Gaul, and ex-Etruscans to Dacia, or some permutation of these.

 

[lpetrich]

Vaclav Blazhek: "Yeniseian Numerals"

1 = Burushaski, Sino-Tibetan, 2 = ST, 3 = Bur, ST, North Caucasian, 4 = NC, Na-Dene? 5 = "complete"? "hand"?, 6 = "extra"?, 5+1, 7 = "many"?, 5+2, 8 = 10-2, 5+3, 9 = 10-1, 5+4, 10 = "hand?, 20?, 20 = "extra", 2*10, 30 = 3*10, 40 = 4*10, 50 = (1/2)*100, 5*10, 60 = 6*10, 70 = 7*10, 80 = 100-20, 8*10, 90 = 100-10, 9*10, 100 = "new"?, from Tungusic? Turkic?

Mentions possible IE etymology for 6: *(K)sweks < *gh(e)s- "hand" + *weks- "to grow" (not *h2wegs-) - in effect, 5+1

I've also seen IE *dekm analyzed as *dwoH- + *kont- (> Germanic *handuz > English "hand") - "two hands"

Harold Fleming: "The Eight “Blood” Etymologies in Afrasian: And More" - a lot of effort has been focused on Semitic and Egyptian, in Semitic on Akkadian, Hebrew, and Arabic, because of their long written records. He did not only "blood", but also "bone" and "four" and "tooth" and "stone" and some other words.

HF: "Two recent efforts by a Russian group (Olga Stolbova, et al.), and Christopher Ehret produced such different results that Joseph Greenberg told a colleague that he had never seen two such different reconstructions of the same family!"

"Therefore I presume nothing about previous reconstructions in this paper. Proto-Afrasian has yet to be reconstructed convincingly and thus we are writing on a blank slate here."

 

[lpetrich]

Vaclav Blazhek then reviewed Lyle Campbell's and William J. Poser's book "Language classification. History and method"

Saying that the authors ought to have started with Plato mentioning in "Cratylus" 410a that Phrygian has words very similar to the Greek words for fire, water, dog, etc.

"The survey of history of comparative linguistics presented in chapters 2, 3 and 4 is fascinating, in spite of some omitted scholars. Chapter 5 is also infonnative, although rather brief." - that chapter: “How some languages were shown to belong to Indo-European” mentioning Hittite, Armenian, and Venetic.

"But chapters 6-12, all written by Campbell, are biased against everyone who entertains so-called ‘distant relationship’, although many of these scholars are or were counted among the best specialists in their disciplines ..."

Then a transition of an article byf Holger Pedersen on Indo-Uralic, from 1933. It could have used some introduction for those not familiar with the details that he described.

Then Vaclav Blazhek on "The Chinese primordial giant Pangu and his possible Indo-European origin" - he was cut up to create our Universe from his body parts, just like Ymir in Norse mythology. With the help of similar stories elsewhere in Indo-Europeandom, like in the be Vedas, this mytheme can be extrapolated back to the Proto-Indo-European speakers. Marduk v. Tiamat of Babylonian mythology is likely a borrowing of these mytheme.

Pangu's name likely means "ancient basin", and that name likely refers to the primordial abyss of some mythologies, as Tiamat's name does.

More Vitaly Shevoroshkin on the origin of Milyan nouns - more of what he has been doing.

 

[lpetrich]

Vaclav Blazhek on Surmic numerals - Surmic is a a branch of East Sudanic, a branch of the putative macrofamily Nilo-Saharan. Like in East and South Nilotic, number names for more than 5 were borrowed from East Cushitic, a branch of Cushitic that includes Somali. For 1 to 5, the Surmic langs sometimes also borrow from E Cushitic.

John Bengtson: "Brian Houghton Hodgson: a pioneer of the Sino-Caucasian hypothesis" back in the 19th cy.

Jonathan Sherman Morris - "The myth of rapid linguistic change IV: The evidence from Afroasiatic" - discussing Ancient Egyptian and its descendant Coptic. About his earlier discussion of Latin and Greek, "... it was shown that a) the main phonological changes tended to take place at a relatively early stage or were not really phonological changes at all but merely the adoption of dialectal variants as a standard language, b) the main changes on a Swadesh list were overwhelmingly due to internal borrowing rather than to massive phonological change or external borrowing and that there were very few ‘new’ words which had not been extensively documented in the classical language, albeit with a slightly different meaning."

Claiming that most of the Egyptian-Chadic and Egyptian-Cushitic cognates are from before Egyptian was written down, due to the desertification of the Sahara in the mid-Holocene. Thus making most of the Egyptian changes internal replacement. That's a common process, it must be noted. Like for "to eat", Latin edere was replaced by comedere > Spanish, Portuguese comer and manducare > Italian mangiare, French manger.


In the review of the book "Kinship, Language and Prehistory: Per Hage and the Renaissance in Kinship Studies," is mention of the work of Alain Matthey de I’Etang and Pierre Bancel, work already mentioned in Mother Tongue.

Chapter 4 presents new statistical evidence which does appear to support the extreme polarity in the use of the term for one sex, although the data for PAPA (terms in over 70% of languages mean father/father’s brother and only 5% mean mother/mother’s sister) and NANA (terms in over 70% of languages mean mother/mother’s sister and only 6% mean father/father’s brother) are more convincing than MAMA (where terms in over 50% of languages mean mother/mother’s sister, but 27% mean father/father’s brother) and KAKA (where terms in 34% of languages mean mother’s brother).

...
It is also curious that the authors overlooked AJA ‘mother, older female relative’, which Bengtson & Ruhlen reconstructed as a global etymology in 1994 ...

PAPA including BABA, I'm sure.

Kindersprache - German "child language"
Lallwort - German "babble word"

 

[lpetrich]

Mother Tongue 17 is in memory of Nostraticist Aharon / Aaron / Aron Borisovich Dolgopolsky

Has a translation of Alfredo Trombetti's 1923 paper "Puluga: The Most Widespread Name of the Divinity" -- many of its putative cognates mean something like "brightness" or "lightning".

StarlingDB "Tower of Babel" Databases
Wiktionary, the free dictionary

I started with German Blitz "lightning". It's from Proto-Germanic *blaikaz "pale, white", *blinkanan "to shine, glitter, twinkle; to blink, wink", *blîkanan "to gleam, shine", in turn from PIE *bhleyg- "to shine" with descendants in Balto-Slavic and Indo-Iranian,

Proto-Semitic *baraq- (Wiktionary) or *bariq- "lightning" (StarlingDB) has descendants across the Semitic langs (Hebrew baraq, Arabic barq, ...)

StarlingDB has Proto-Afro-Asiatic *bariq- meaning "lightning" though some of its non-Semitic putative descendants are listed as possible borrowings from Semitic langs like Arabic.

Going further into Borean I find *PVRKV with descendants in Afroasiatic and Sino-Caucasian, like Proto-Sino-Tibetan *p(r)iâk "white"

Indo-European *bhleyg- also fits.

That reminds me of Aharon Dolgopolsky himself once discovering that words for "star" are much more stable than those for "lightning". Looking under "flash of light" translations in lightning - Wiktionary, the free dictionary it is rather obvious. However, "white" is one of the more stable sorts of words.

 

[lpetrich]

Then some back-and-forth about "one". Stephan Hillyer Levitt starts out by saying that there are five word forms with this meaning in the langs classified as Nostratic, Eurasiatic, and "Duraljan".

Hannu Panu Aukusti Hakola has written some books, "Duraljan Vocabulary: Lexical Similarities in the Major Agglutinative Languages" and "Duraljan Hypothesis: Towards the Mother Tongue of Man -- Six Articles and Some Reviews Around the Duraljan Hypothesis" -- Dravidian, Uralic, Altaic, Japanese, and Andean (Quechua, Aymara)

"It is the contention of this paper that the number of these basic words can be reduced to perhaps two basic words."

Vaclav Blazhek and Allan Bomhard are both very skeptical, Philippe Bürgisser discusses Nilo-Saharan and finds a variety there, and Juha Janhunen looks at that and other numbers in Ural-Altaic.

"Gradual completion of numeral systems." - that's evident from what Vaclav Blazhek has written about in MT.

Uralic has two major branches, Finno-Ugric and Samoyedic, and they have only two number roots in common, one for 2, *käkta, and one for 5, *wixti, that means 5 in FU and 10 in Sam. Proto-FU nums go up to 6, while Proto_Sam nums to 10.

All the other groupings of the Ural-Altaic "complex" are much younger, in protolanguage breakup or in attestation, and they have full-scale decimal systems. He notes Proto-Japanese 1 *pitö, 2 *puta, 3 *mi, 4 *yö, 5 *itu, 6 *mu, 7 *nana, 8 *ya, 9 *könönö, 10 *töwo with ablaut (vowel-shift) relationship 1 ~ 2, 3 ~ 6, 4 ~ 8, 5 ~ 10.

FU *ikte, Sam *op, Turkic *bir, Mongolic *niken, Tungusic *ämün, Koreanic *kon-, Japonic *pitö
Indo-European *oy- (oynos, oykos, oywos), *sem-
Dravidian *oru
Semitic *ishte-, *ahhad-

SHL returned with suggesting that Sanskrit eka < Hebrew ehhod other Semitic ? Skt eka < Finnish yksi other Uralic ? He concedes skepticism about both possibilities.


Words for 1 are not as stable as words for 2 because 1 overlaps with meanings like "alone". But 2 also has a lot of variety, and the Bengtson-Ruhlen global etymologies for 1 and 2 use a lot of additional semantics, like "finger" and "part".

But at least "elder m/f relative on m/f side" is somewhat well-defined semantics.

 

[lpetrich]

Since this is Thanksgiving, I'd like to thank Allan Bomhard for uploading some macro-linguistics books to the Internet Archive. Thanx for letting me dump the paper copies that I'd made of those books long ago.


Mother Tongue 18 starts with a tribute to Joseph Greenberg's "The Languages of Africa" (1963). His three principles:
(A) the sole relevance of resemblances involving both sound and meaning,
(B) mass comparison as opposed to isolated (binary) comparison, and
(C) only linguistic evidence is relevant to genetic classification.

About (A) he distinguished by phonetic detail the gender marking of Afroasiatic - (), t - some Eastern Nilotic - l, n - Khoekhoe (Nama) - b, s

Khoekhoe distinguishes three genders and numbers. Gender then number:
f 1 s, 2 ra, + di
m 1 b, 2 kha, + gu
n 1 i, 2 ra, + n

"Simultaneously with his studies in genetic classification Greenberg (e.g., 1954b, 1957b, 1957c) was laying the foundations of modem linguistic typology, and he understood, possibly better than anyone else at the time, that typological classification is distinct from genetic classification."

About (B), mass comparison makes patterns stand out much better, something good for subjective classification.

(C) is rather obvious, but it was violated a lot in the past.

 

[lpetrich]

Pierre J. Bancel noted that the Societe de linguistique de Paris (SLP) added to its bylaws "[t]he Society does not admit any communication concerning ... the origin of language."

Then saying that we must distinguish between the origin of language and the origin of languages. That article did not deliver, sad to say, so I will attempt to do so.

I think that our present species originated in a relatively small population, small enough for genetic innovations to quickly spread through it, but large enough to making dropping out of good genes unlikely. Research into our genes does indeed suggest that some ancestral population went through a genetic bottleneck or bottlenecks at some time or times over the last few hundred thousand years.

Such a population would most likely have had one language, the original language of our species.

These people multiplied and split up into descendant populations, and one of these populations split into populations that stayed distinct populations all the way to the present day. The language of that population became the protolanguage of all our present-day and recorded languages: Proto-World, Proto-Global, Proto-Human, Proto-Sapiens.

But the descendants of that population likely changed their languages so much that this original language is now totally obscured, though there are some linguists who claim that they can recover some superconserved word forms. They are mostly of words for body parts and other commonplace things, words that are usually inherited and not changed very much, words for nose and water and the like.

That's why I can't take seriously Alfredo Trombetti's claim that he had discovered the oldest deity name. The most that he did was discover an early word for "white" or "bright", a word that got shifted to "lightning" more than once.

The oldest recoverable deity names that I know of are Proto-Indo-European ones, though Proto-Semitic ones would have a similar time depth.

The best-known of the Indo-European ones is *dyeus pHter "Father Sky" - rather generic words, I must note. From comparative mythology, it may be possible to track back some Father Sky or Father Heaven farther than that, but we would not be able to get the specific words with any confidence. "Father" would be papa or apa or pa or ap or baba ... or tata ... or dada ... or mama ... or nana ... or different vowels in any of these.

 

[lpetrich]

Kirill Babaev: "Joseph Greenberg and the Current State of Niger-Congo"

JG proposed four macrofamilies: Afrasian, Nilo-Saharan, Niger-Congo, and Khoisan, but more mainstream linguists consider all but Afrasian to be lacking in demonstrated overall relationship, and some have been doubtful about Afrasian also.

Later Greenberg always emphasized that his method of multilateral comparison can only be treated as a proposal, a dotted line which was still to be verified or modified by means of the strict comparative method, of which Greenberg remained a passionate supporter through all his life.

He then talks about his research in "Reconstructing Niger-Congo person marking" - personal pronouns: independent, subject on verb, object on verb, possessor on noun

Finds a somewhat different classification from Wikipedia.

• Benue-Congo: Bantu, Cross River, Central Nigerian (Platoid: Plateau, Jukunoid, Kainji), Edoid (in Volta-Niger), ...
• Volta-Congo: Kwa, Kru (in Niger-Congo), Gur (Sav), Adamawa (Sav), Ubangi (Sav), Atlantic (Atlantic-Congo: West Atlantic)
• More distant: Dogon (in Niger-Congo), Kordofanian (in Niger-Congo)

Sav: in Savannas

The only putative Niger-Congo families without such demonstrated affinity are Mande and Ijoid. KB compared person markers for Chadic, East and Central Sudanic, Songhai, and Kadugli-Krongo, and found a lot of difference from NC.

 

[lpetrich]

Allan Bomhard on Proto-Nostratic Morphology, tending to agree with Aharon Dolgopolsky. Subject-object-verb order, genitive postposition *nu, marked accusative postposition *ma etc.

I'll look at Indo-European, Uralic, Turkic, and Mongolian. All ancestrally have SOV order.

Nominative = subject case, accusative = direct-object case, genitive = of-case. Naming cases after prepositions would make their names less scary for big users of prepositions like English speakers, I think.

 Proto-Indo-European nominals - athematic

• Nominative animate: singular ~s, -, plural -es
• Accusative animate: singular -m, plural -ms, -ns
• N/A neuter: singular -, plural -h2 (-a)
• Genitive: singular -s, -os, -es, plural -ôm

 Proto-Uralic language
Nominative -, accusative -m, genitive -n

Proto-Turkic - Wikibooks, open books for an open world
Nominative -, definite accusative -ni, genitive -ning
For an indefinite direct object, one uses no ending, like the subject.

Mongolian
Nominative -, definite accusative -ig, genitive -in

So that Nostratic reconstruction partially fits.

 

[lpetrich]

Pre-Proto-Indo-European and Proto-Nostratic are both proposed to be active-passive languages, and such langs have some grammatical properties that are very unlike those of most languages that many of us are likely to be familiar with. From Klimov on typical features of such langs:

Lexicon:
1. Binary division of nouns into active vs. inactive (often termed animate and inanimate or the like in the literature).
2. Binary division of verbs into active and inactive.
3. Classificatory verbs or the like (classification based on shape, animacy, etc.).
4. Active verbs require active nouns as subject.
5. Singular-plural lexical suppletion in verbs.
6. The category of number absent or weakly developed.
7. No copula.
8. "Adjectives" are actually intransitive verbs.
9. Inclusive/exclusive pronoun distinction in first person.
10. No infinitive, no verbal nouns.
11. Etymological identity of many body-part and plant-part terms (e.g., “ear” = “leaf’).
12. Doublet verbs, suppletive for animacy of actant.

"ear" ~ "leaf" seems like what some zealous conlanger might want to do, like Esperanto's creator using negation prefix mal- a great amount, like bona "good", malbona "bad", even though such oppositions are often lexically separate.

Singular-plural big variation? Basque has something like that.

Copula = "to be". Adjectives as verbs: "Let's active" for "Let's be active".

Syntax:
13. The clause is structurally dominated by the verb.
14.“Affective” (inverse) sentence construction with verbs of perception, etc.
15. Syntactic categories of nearer or farther object rather than direct or indirect object.
16. No verba habiendi.
17. Word order usually SOV.
18. Direct object incorporation into verb.

Like saying "it pains me" instead of "I feel pain".

Verbum habiendi = verb of possession, like English "to have" -- some langs use constructions like "it is at me" rather than "I have it" -- Russian, for instance.

Morphology:
19. The verb is much more richly inflected than the noun.
20. Two series of personal affixes on the verb: active and inactive.
21. Verbs have aspect or Aktionsarten rather than tense.
22. The noun has possessive affixes.
23. Alienable-inalienable possession distinction.
24. Inalienable possessive affixes and inactive verbal affixes are similar or identical.
25. Third person often has zero affix.
26. No voice opposition ...
27. Aetive verbs have more morphological variation or make more morphological distinctions than inactive verbs.
28. The morphological category of number is absent or weakly developed.
29. There are no noun cases for core grammatical relations (no nominative, accusative, genitive, dative). Sometimes there is an active/inactive case opposition.
30. Postpositions are often lacking or underdeveloped in these languages. Some of them have adpositions inflected like nouns.

Proto-Indo-European verbs are reconstructed as mainly varying in aspect instead of tense, and varying rather irregularly, like in Slavic. Aktionsart = lexical aspect, aspect as part of the meaning of a verb.

 

[lpetrich]

I've been skipping over archeology and genetics, and there's a *lot* of articles on the genetics revolution of human prehistory in Mother Tongue. Articles like "The Early Dispersions of Homo sapiens sapiens and proto-Human from Africa"

But that article mentions the "Borean hypothesis": Afrasian, Nostratic (w/o AA), Dene-Caucasian, Amerind
Also "NOSCAU" - Nostratic (w/ AA?), Sino-Caucasian, Austric


Vitaly Shevoroshkin "On the Origin of Milyan Verbs"

 Anatolian languages - Hittite is the oldest recorded IE lang, the lang of the Hittite Empire, over 1750 - 1180 BCE. At its peak, that empire conquered northern Syria and ran into the Egyptians at Kadesh, where they fought a big and inconclusive battle with chariot armies. That empire was destroyed by the Sea Peoples, the destroyers of Mycenaean and Levantine cities and invaders of the Nile Delta.

Most of the others are known from varying amounts of inscriptions. Milyan from 2 poems of 34 and 71 lines each, and 1 of 9 lines. There are lots of such Trümmersprachen ("ruins languages"), known only from short inscriptions and the like, and I read somewhere that what's been written on them exceeds the amount of survivals of them.

 

[lpetrich]

Mother Tongue 18

Shamil Nafiqoff: "Global Etymologies and Alfredo Trombetti" - a long ranger and Proto-Worlder from a century ago.

"As pointed out on several occasions above many global etymologies by Trombetti have been borne out by eminent long-range researchers of later times."

Peter A. Michalove: "Notes on the Moscow Conference on Long-Range Comparison" in 2000

Then a review of "Desi Words Speak of the Past: Indo-Aryans in the Ancient Near East", by Dr. Liny Srinivasan

LS became interested in "finding out the existence of a massive number of Canaanite and Egyptian words in Bengali, I realized the need to uncover the historical connections." Then collaborating with Cyrus Gordon, a recognized authority on the early history of the eastern Mediterranean, and also someone interested in eccentric theories. But these authors' comparisons were murky hand-waving.

As mentioned above the author’s “goal is to draw the attention of the scholars [in Near Eastern studies] to the ancient Indian literature,” but unfortunately the methods employed in this book make it very difficult to be used effectively. It is not enough to juxtapose similar sounding words and names. One must also show that the lexemes are plausibly connected by employing etymological and morphological analysis. In the case of myths and stories there has to be enough structural and historical analysis to demonstrate that the narratives being compared are credibly connected in origin. This reviewer is an outsider to both fields concerned (Near Eastern studies and Indic studies), and it is possible that experts in these fields might be able to extract some useful information from Dr. Srinivasan’s book, but I am not very hopeful of this.

That reminds me of someone comparing Abraham and Sarah from the Bible to Brahma and Sarasvati from Hindu mythology. With enough names and enough comparison latitude, one can compare anything to anything.

 

[lpetrich]

Mother Tongue 19 has an article on correlating macrofamilies with mitochondrial-DNA haplogroups.

Evidence for correlating some languages to mtDNA haplogroups appears more robust than others; it seems by and large that correlations become weaker the closer we approach recent times. As a nonlinguist, I am aware that there are complex language replacement issues pertaining to mismatches between the languages populations currently speak and those they may have spoken in historic or prehistoric times. The language correlations that I have proposed are more or less tentative. I am open to counterarguments with respect to which language to correlate to particular branches in the mtDNA phylotree.

The strongest correlation is for the branching off of the L0 haplogroup from all the others - it is associated with Khoisan speakers. A big problem with the later ones is overlaying of population expansions and elite expansions. That is likely to produce discrepancies between mitochondria and Y chromosomes.


Then, "The Kinship Term KUKU ~ KOKO ~ KAKA in the American Indigenous Languages, the Amerind Hypothesis, and the Dravidian Kinship System. Part I: Linguistic Study" by Alain Matthey de l'Etang & Pierre J. Bancel

They use abbreviations P ‘parent,’ G ‘grand,’ M ‘mother,’ F ‘father,’ Z ‘sister,’ B ‘brother,’ U ‘uncle,’ A ‘aunt,’ E ‘spouse,’ e/y ‘elder/younger,’ W ‘wife,’ H ‘husband,’ C ‘child,’ S ‘son,’ D‘ daughter,’ inL ‘-in-law,’ Sib ‘sibling.’ -- somewhat different from their earlier ones

"Our first goal in this paper is to make a general appraisal of the distribution in the Americas of a term long and well known to Amerindianists, which can be labeled KOKO and has the general semantic scope of GF, MB, EF for the masculine side and GM, FZ and EM for the feminine side."

Too much for coincidence. Diffusion from marrying out of groups? "But most of the documented cases of linguistic and ethnic contact show that kinship terms are seldom borrowed by one group from another, because language appears determinant in keeping the group’s identity." It's not a distinctive Americanism, because it's also found in other places: "New Guinea, Australia, Africa, Eurasia, Oceania,"

Another hypothesis: K is acquired after M, N, B, P, D, T, and thus, KAKA and similar terms would thus be used for more distant relatives.

Another problem: although this is common, what else is common?

 

[lpetrich]

John Bengtson: "Burushaski and the Western Dene-Caucasian Language Family: Genetic and Cultural Linguistic Links"

WIth this proposed Dene-Caucasian family tree, split time 10,660 BCE (from glottochronology - treat as very approximate)

• East: Sino-Dene - 9000 BCE - Sino-Tibetan, Na-Dene
  • West - 8330 BCE
    • 6570 BCE - Burushaski, Yeniseian
    • Basque, North Caucasian
   Cardium pottery makers were the first farmers in Basque country, and they arrived around 5700 BCE. They got their name from a style of decorating pottery, by pressing cockle shells into it to make heart-shaped imprints. Cockles are a kind of bivalve.
  
  The first cardium-pottery makers lived in NW Greece near Albania around 6400 - 6200 BCE, and they were preceded by Neolithic farmers who reached the Greek islands in 7000 BCE. Their predecessors, Pre-Pottery Neolithic B spread across the Fertile Crescent and southern Anatolia, starting in 8800 BCE, and those ones' predecessors, Pre-Pottery Neolithic A, spread across the Levant, starting in 10,000 BCE, with the invention of agriculture. Their foraging but sedentary predecessors, the Natufians, go back to 13,000 BCE.
  
  So the timing is about right for a Basque - NC split in the northern Fertile Crescent around 8000 - 7000 BCE.
  
  JB finds cognate words in Basque, NC, and Burushaski for some kind of grain and also for threshing and threshing floor.  Threshing - removing grain from the rest of the plant by beating it or crushing it. For beating the grain, a flail was often used, a stick with another stick loosely attached to its end. For crushing it, a hoofed animal stepping on it, dragging a board with embedded stone flakes across it, or using a roller on it. Afterwards is  Winnowing - separating out the now-loosened grain from the rest of the plant, the chaff. Premodern winnowing was often throwing this mixture up into the air on a windy day, with the wind blowing away the chaff.
  
  Machines for doing threshing and winnowing were invented in the 18th cy., and combines, machines that thresh and winnow grain crops as they reap those crops, have been used for well over a century.
  
  This also means that the Basque-NC dispersion also involved the Burusho people, who now live in the mountains in the northern end of Pakistan.
  
  Also cognate across Basque-NC-Burushaski are some words for domestic animals and dairying.
  
  Ancestral Sino-Tibetan speakers invented agriculture separately:
  
  Dated phylogeny suggests early Neolithic origin of Sino-Tibetan languages - PMC - around 6000 BCE in the Yellow River basin in northern China
  
  Addressing Dene-Yeniseian, JB writes
  

  Starostin and I asserted that the binary Dene-Yeniseian model is less adequate than the multilateral Dene-Caucasian one, and that Dene-Yeniseian is taxonomically inaccurate, since, in our view, Na-Dene is closer to Sino-Tibetan, and Yeniseian is closer to Burushaski, than Na-Dene and Yeniseian are to each other.

  So it's binary comparison all over again.
  

  It is plain to see from the cultural lexicon discussed above that the Yeniseian languages share virtually none of the vocabulary pertaining to food production (domestic cattle, dairying, cultivated grain crops and their processing), with the other 'Western Dene-Caucasian' languages, Burushaski, Caucasian, and Basque. This is simply because the climate and landscape in the homeland of the Yeniseian peoples, in the Siberian taiga, are not conducive to food production, and presumably the Proto-Yeniseians either did not participate in the Neolithic revolution, or were forced by a new environment to abandon food production. The only exception (until Soviet rule) was the adoption of domestic reindeer by some Ket groups, from Samoyedic neighbors (Vajda 1998). So to some extent, at least, the apparent similarities between Yeniseian and Na-Dene in lexicon associated with sleds, snowshoes, birch-bark, and the like, are to be attributed not to taxonomic closeness but to similar environmental and cultural conditions.