Language as a Clue to Prehistory

[lpetrich]

Some biconsonantal roots stay biconsonantal, like /-kh "brother". Some have only one triconsonantal derivative, like k-k with k-k-b "star". Some have several, like r-g "use the feet" with various derivatives with -d, -l, -s, -\, \- with meanings "foot" and foot-related activities like "run", "leap", "trample", "dance", "kick", ...

/ aleph (glottal stop)
\ ayin (voiced hh)

Then mentioning Semitic-language prefixes /-, ya-, m-, t-, sh- and suffixes -(a,o,u)n, -(a,u,i)t, -ya, -awi, -ay, -iy, -iyya

-Vn makes plurals, -Vt makes abstract nouns, and -iy makes belonging to something, like Slavic -ski.

Some triconsonantal roots were formed by inserting a semivowel, -y- or -w-, into a biconsonantal root. Like b-t becoming bayit "house".

As an additional root consonant, /- has no fixed meaning.

Repeating the second consonant often has an intensive meaning, but not always. An alternative is to repeat the whole root.

The main suffixes:

• -d, -z
• -r, -l
• -s, -sh
• -m, -n
• -h, -kh, -hh
• -k, -q
• -b
• -t, -s, -d, -z (emphatic)
• -p (>-f)
• -g (>-j)
• -y, -w

None of them have very evident semantics.

Something not addressed very well in the paper is how consistent the triconsonantal derivatatons were between the dialects. From a rather cursory look, they seem consistent, but one would have to do a detailed analysis. But it does seem that a lot of the specific extensions go back to Proto-Semitic, with them being created in some predecessor.

The paper did not go into detail about other Afro-Asiatic languages, and whether they have anything similar.

 

[lpetrich]

Journal of Language Relationship

has
Once Again on the Comparison of Personal Pronouns in Proto-Languages by Kirill Babaev
The article itself (PDF)

In the absence of honorific effects, personal pronouns are very stable. Honorific effects like using second person plural as a formal or respectful form of the second person singular -  T–V distinction in the world's languages - and having several forms of the personal pronouns, like in Japanese, Korean, and Thai. Japanese and Korean also have honorific effects in their verb conjugations: Japanese formal suffix -masu and Korean formal suffix -mnida.

Personal pronouns can be present in three ways: standalone, as verb-conjugation affixes, and as noun-possession affixes.

If a verb conjugation has a sufficiently-distinct personal conjugation, then one can safely omit standalone subject pronouns, making the langs "pro-drop" langs. But what is especially curious is when a language lacks personal verb conjugations, but its speakers often omit subject pronouns, using context instead. Like in Japanese, Korean, and Thai.

KB has a big table of reconstructed 1s, 2s, 1p, and 2p pronouns, using both standalone pronouns and verb conjugations. Only some of the plural forms are recognizable as plurals of the singular forms.

 

[lpetrich]

Looking at 1s and 2s pronouns, the pattern M-T is evident in northern Eurasia, alongside M-S, which seems like a phonetic variant -- Indo-European, Uralic, Altaic (Turkic, Mongolian, Manchu-Tungus), Chuckchi-Kamchatkan, Kartvelian -- Nostratic/Eurasiatic.

Also present in northern-Eurasian verb conjugations is M-K for the first person: M is relative/transitive, K is stative/intransitive. PIE 1s *H may be derived from this *K -- Hittite -mi, -hi conjugations, other IE *-mi, *-oH.

M-T is absent from elsewhere in the world.

Another pattern, S/Z-W, is evident in North Caucasian, Hurro-Urartian, Yeniseian, and Na-Dene -- Dene-Sino-Caucasian. But Sino-Tibetan, Burushaski, and Basque don't fit very well, however, though those three fit another possible set: *n-/m- and *ghu/ghw-, along with Yeniseian prefix pronouns.

Another one, K-M, is found in Tai-Kadai, Miao-Yao, Mon-Khmer, and partly for Austronesian -- Austric. It is also found in New Guinea as Set II, often considered to be the oldest one there.

In much of the Americas is N-M, though that is not universal there.

Kusunda (central Nepal) and West Papua have T-N

In much of the Niger-Congo family, like the Bantu langs, one can find pattern (N/M) -W for 1s subject N object M.

 

[excreationist]

DOUBT is from the French DOUTE

RECEIPT is from RECETE

Also
DEBT
PLUMBER
AISLE
ISLE
ISLAND
SUBTLE
INDICT (pronounced "indite")

DEBT comes from the Latin DEBITUM (DETTE in Old French and Middle English)

The Renaissance scholars wanted the Latin roots to show through the use of silent letters....

(sorry if someone has already pointed this out)

 

[lpetrich]

That is in addition to all the actual archaisms preserved in spelling, spellings which were closer to earlier forms of the language's pronunciation. For instance, gh was the "kh" fricative sound in Old English, still preserved in Dutch and German. But it dropped out of Middle English. Also dropping out was a lot of final "e" vowels. It was not just sounds dropping out by sound shifts. The  Great Vowel Shift is why English values of the Latin-alphabet vowel letters are sometimes very different from most other languages' values of them -- before the GVS, English was close to these others.

For instance, the Old English ancestor of "house" /haus/ was hûs (I use a hat for a long vowel), and of "bite" /bait/ was bît-

This has not only happened to English. For instance, French spelling is closer to Old French pronunciation than to present-day pronunciation -- lots of final sounds dropped off, both vowels and consonants. Also, Modern Greek collapsed several Ancient Greek vowels into a few present-day ones:

• /a/ - a, â, âi
• /e/ - e, ai
• /i/ - i, ê, ei, êi, oi, u, ui
• /o/ - o, ô, ôi
• /oi/ - ôu
• /u/ - ou
• /av/, /af/ - au, âu
• /ev/, /ef/ - eu
• /iv/, /if/ - êu

Phonetic values - Ancient Greek vowel transcriptions

 

[lpetrich]

Suppletion: inflection that uses word forms from different sources. English has some, like good-better-best, bad-worse-worst, be-am-is-are-was-were-been, go-went-gone, and Latin and the Romance languages have some in their words for "good", "bad", "to be" and in the Romance langs, "to go". Slavic verb aspects also have suppletive forms, like Russian govorit' (imperfective), skazat' (perfective) "to speak, say".

There is a lot of it in the earlier Indo-European langs, and some of it can plausibly be reconstructed in PIE, like in the verb aspects.

In the literature, the aspects are often given the names of associated tenses in Greek and Sanskrit:

imperfective - present, perfective - aorist, stative - perfect

Dawn of verbal suppletion in Indo-European languages | Digital Library of the Faculty of Arts Masaryk University by Dita Frantíková with PDF 130121.pdf
Abstract:

If considering one first and best attested language in each branch, we find almost seventy (so far described) suppletive verbal paradigms. The paper examines their respective Proto-Indo-European roots and concludes about the relationship of their form and semantics. Special attention was given to the verbs of being, for which all branches choose to use the root √h1es- for the present form, while as many as seven stems combine with √h1es- for non-present usage. The range of semantic fields found among the suppletive verbs is discussed with conclusions concerning the relationship of form and semantics.

Especially *h1es- "to be (imperfective)" -- perfective and stative forms that I could track down:

• *bheuH- Germanic, Latin, Celtic, Balto-Slavic, Indo-Iranian
• *h2wes- Germanic
• *genh1- Greek

*h1ey- "to go (imperfective)" (Latin eô "I go") also combines with a lot of other roots, as does *bher- "to bear, carry (imperfective)" though there aren't any widely-shared perfective roots for either of them, not like *bheuH- for *h1es-.

The semantics of these suppletive verbs; verbs that are suppletive in at least some of the dialects.

• Body parts and functions: live, die, to kill
• Food and drink: eat, drink
• Physical acts and materials: do, carry, fall, guard, lead (guide), drive, bring, fence in, pull, strike, (hit)
• Motion and transportation: go, run, come, throw, drive
• Possession and trade: have, take, give, steal, buy, sell
• Sense perceptions: see, watch, look
• Mind and thought: want, appear (seem)
• Language and music: say, speak
• Spatial relations: put, place, sit, lie, lay
• Warfare, hunting: fight
• Expressions of being: be, become

The observation one can make about suppletive verbal paradigms summed up in the Figure 6 is that the suppletive stems seem to express general ideas. We do not find verbs with specific meanings such as e.g. fish, knit, sprinkle or decorate. Also, there does not seem to be a specific semantic field(s) in which we should look for suppletive verbs. Their meaning as such does not seem to unify them.

Author Dita Frantíková is also the author of Verbal suppletion in Indo-European languages | CU Digital Repository - in English, with Czech name Slovesný supletivismus v Indoevropských jazycích

Essentially the same thing, but going into detail about support, like tracking down the various verb forms.

 

[lpetrich]

Aspectual suppletion and paradigm defectiveness in the Proto-Indo-European verbal system - "The existence of suppletion in the Proto-Indo-European language is still a question of debate (García Ramón 2002). While the evidence for such a phenomenon has been widely recognized within the verbal system of most Indo-European languages, some scholars describe it as a recent monoglot development which characterizes the history of each single language without involving a previous common stage (Van der Laar 2000)."

Rather odd opinion, since one has to ask what made it so common in the dialects' early histories. Not much is consistent over them, it must be conceded, mainly "to be": *h1es- (imperfective), *bheuH- (perfective).

"In a typological perspective, the phenomenon of suppletion is another piece of evidence in favor of the active-stative hypothesis for the most ancient PIE stage (Lehmann 2002), reflecting a system where the lexical aspect features of the verbal roots are still more important than tense features."

There are multiple forms for some words that may reflect some animate - inanimate distinction. Like "fire":

• h1ngwnís - masculine - animate and active - Latin ignis, Russian ogon', Sanskrit agni
• péh2wr - neuter - inanimate and passive - English fire, Greek pur, Hittite pahhur

In the Vedas, Agni ("Fire") was a prominent deity.

péh2wr had a r/n athematic-noun declension: nominative, vocative, and accusative in -r, other cases in -n-, like genitive ph2wéns. Several other PIE nouns had the same kind of declension: wódr "water" (>Russ. voda, Greek hudor, Latin unda "wave"), *yókwr "liver" (>Latin iecur, iecinor-, Greek hêpar, hêpat-), ... Latin femur "thigh" had stem femin-, but it has no clear IE derivation. But Latin iter "journey" with stem itiner- is likely descended from *h1ey- + *-tr where the first part means "to go" and the second part is the r/n suffix.

 

[lpetrich]

On the nature of verbal suppletion

Evidence from a case of incipient verbal suppletion in Romanian illustrates that the rise of suppletion involves semantic weakening of lexical items in casual speech with resultant changes in word frequency of synonymous and near synonymous verb forms. Over time, these changes result in the replacement of forms of one verb by forms from another, and either loss of the remaining parts of the other verb, or the analogical reformation of forms to replace those taken over by the now suppletive verb.

So if a word seems too weak, it can be replaced by another, or only partially replaced -- suppletion.

Dita Frantíková thesis: Verbal suppletion in Indo-European Languages


Couldn't find much about suppletion in Uralic or Turkic, at least what is not paywalled.


I decided to look at words for "good" and "bad" in Wiktionary, and there is a lot of variation over Indo-European.

English comparatives, superlatives, and adverb good - better - best - well - bad - worse - worst goes back to

Proto-Germanic *gôdaz - *batizô - *batistaz - *wela - ? - *wirsizô - *wirsistaz

Latin has bonus - melior - optimus - bene - malus - pêior - pessimus - male

These are well-represented in the Romance langs.

I found similar suppletion in Greek and Balto-Slavic, but my patience ran out. These words vary enough to make it difficult to reconstruct Proto-Indo-European words for "good" and "bad". However, PIE did have prefixes *h1su- and *dus- with those meanings (>Greek eu-, dus- borrowed as eu-, dys-; Sanskrit su-, dus-)

 

[Swammerdami]

Thanks for the information, lpetrich.

By now it's completely obvious(!) to me that a major source of the "Germanic substrate" was the Pitted Ware culture. Yet the Wiki article  Germanic substrate hypothesis treats the substrate source as a mystery. It mentions Pitted Ware ZERO times in the text (though it links to it in the "See also" section).

Surely, Germanic words like "Ship, Sail, Sea, Seal, Keel, Eel" come from Pitted Ware. What about "King, Knight, Wife, Bride, Child"? Or "Body, Bone, Back. Brain, Finger, Toe, Neck, Cheek"? Et cetera. Is it possible to study the phonologies of the earliest proto-Germanic forms of these non-IE words and guess which ones came from the same substrate language?

I'm not asking for the substrate language(s) to be identified — I stipulate that the Pitted Ware language has ZERO discernable relatives (except for Proto-Germanic itself). I just wonder if the long list of non-IE words in proto-Germanic has enough internal evidence to guess how many different substrate languages were involved?

For reference, here is a rough chronology of the development of proto-Germanic:

2600 BC - three different cultures were prevalent in an interaction region centered near southern Sweden

• The Pitted Ware culture. Despite being non-farmers, their excellent sea-faring skill and littoral locations provided ample food to support high population. (They also ate pigs: wild or domestic?) They traded with farmers, offering seal skins, amber etc. for farming products. With their command of the sea, they were probably very influential if not dominant.
• The northern-most Funnel Beaker (TRB) culture, its southern parts being overwhelmed by Corded Ware culture. Nothing is known of its language, which might have descended from mesolithic Europe.
• Corded Ware and Battle Axe. The vast extent of Corded Ware suggests that two different I-E languages might have been involved: a sibling to Italic in the West, and a sibling to Baltic in the East. Or these languages might have merged as a koine.

Genetically, Pitted Ware and TRB were of Northern Europe H-G stock, while Corded/Battle-Axe came from Yamnaya.

Although pre-Neolithic, the Pitted Ware culture made ceramics (as its name implies), probably borrowed from the Comb Ceramic culture to its East. That pottery in turn has been linked to that of Eastern Siberia!

2200 BC - The cultures were merging, presumably to speak a language ancestral to proto-Germanic. The language was likely Indo-European, but with a major Pitted Ware substrate as that powerful culture shifted to I-E.

1800 BC - Nordic Bronze Age. "The Nordic Bronze Age maintained close trade links with Mycenaean Greece, with whom it shares several striking similarities. [The Nordics] were actively engaged in the export of amber, and imported metals in return, becoming expert metalworkers. With respect to the number and density of metal deposits, the Nordic Bronze Age became the richest culture in Europe during its existence." (But IF this Nordic culture was so strong AND spoke proto-Germanic, might we not expect to find borrowed Germanic words in Greek or Celtic?)

600 BC - Although the Celts of West Central Europe had entered the Iron Age much earlier, the Germanic people hadn't. It was only after this date that the German-Celt contact led to the Jastorf iron-working culture, which is cited as the date of the proto-Germanic language.

There's a huge 2000-year gap between Pitted Ware and Jastorf. This leaves room for important but unknown transitions missing from this summary.

 

[lpetrich]

Thanx.

I remember from long ago learning of Russian irregular nouns, like these family-member words that had PIE *-ter (nom. sing., gen. sing., nom. pl.):

• mother: mat' ма́ть, ма́тери, ма́тери
• brother: brat бра́т, бра́та, бра́тья
• daughter: doch' до́чь, до́чери, до́чери

The words for "mother" and "daughter" are the only ones with their irregularity. So the Russian words partially or completely lost the final r. That's also true of most of the other Balto-Slavic languages. There is a missing family -ter word: a descendant of PIE *ph2tér- "father". But "sister" and "son" are well-preserved.

Such dropping out of descendants of PIE word forms is something that happened elsewhere in the dialects. In Latin, "son" and "daughter" dropped out, and in Greek, "brother" and "sister" dropped out. In some descendants of Latin, some Romance languages, "brother" and "sister" also dropped out.

 

[lpetrich]

I also recall that there are 10 Russian nouns with a very irregular declension, nouns like и́мя "name" (gen. sg. и́мени, nom. pl. имена́) - list from Russian ‘н’ declension of имя, время, etc. – therussianblog

бре́мя bremya “load”, пла́мя plamya “flame”, вре́мя vremya “time”, пле́мя plemya “tribe”, вы́мя vymya “udder”, се́мя semya “seed”, зна́мя znamya “banner”, стре́мя stremya “stirrup”, и́мя imya “name”, те́мя temya “crown of head”

This n-stem declension is a relic of PIE noun-forming suffix *-mn

• imya < *h1nomn "name"
• vremya < *wert-mn < *wert- "to turn" (Latin vert-, Sanskrit vart-)
• semya < *seh1mn "seed" (Latin sêmen) < *seh1- "to sow" (English "seed")
• bremya < *bhermn "load" < *bher- "to bear, carry"
• znamya < *gneh3-mn "sign" < *gneh3- "to recognize, know"
• vymya < *h1owHdhr "udder"
• plamya < PSl *polmy "flame"
• plemya < PSl *pleme "tribe"
• temya < PSl *teme "crown of head"
• stremya < PSl *streme "stirrup"

PSl = Proto-Slavic

For "udder", the PIE word was an r/n-stem one, like *wodr "water" and *peh2wr "fire". The Proto-Slavic form suggests that *-mn was added to it -- *h1owH-mn ?

 

[lpetrich]

Northern Eurasian languages and possible relatives:  Altaic languages (Transeurasian) and  Proto-Altaic language and  Eurasiatic languages and  Nostratic languages (Mitian) and  Afroasiatic languages and  Dené–Caucasian languages and  Austric languages and  Amerind languages and  Borean languages

• Micro-Altaic: (Turkic, Mongolian), Tungusic
• Macro-Altaic: Micro-Altaic, (Korean, Japanese)
• Eurasiatic: Indo-European, Uralic, Altaic, Nivkh/Gilyak, Chukchi-Kamchatkan, Eskimo-Aleut
• Nostratic: Eurasiatic, Kartvelian, Dravidian
• Dene-Caucasian: Basque, North Caucasian, Burushaski, Sino-Tibetan, Yeniseian, Na-Dene
• Austric: (Austronesian, Kra-Dai: (Tai-Kadai, Daic)), Hmong-Mien / Miao-Yao, Austroasiatic
• Borean: Nostratic, Dene-Caucasian, Sumerian, Ainu, Austric

Something curious about the racial composition of the original speakers of nearly all of these language families: they all have non-Negroid features: light skin and straight or loosely-curled hair. Negroid features: dark skin and tightly-curled hair.

Negroid people are the people south of the Sahara Desert, "Negritos": various people scattered over southern and southeast Asia and nearby islands, and "Australoids" of Australia, Papua New Guinea, and Melanesia.

So the "Out of Africa" migration split in two, one going eastward to south Asia, southeast Asia, New Guinea, Australia, and nearby islands, and one going northward to the Middle East and then to the bulk of Eurasia and the Americas.

So if the Borean family is a sensible concept, then these northward migrants would have spoken Proto-Borean.

 

[excreationist]

....Something curious about the racial composition of the original speakers of nearly all of these language families: they all have non-Negroid features: light skin and straight or loosely-curled hair. Negroid features: dark skin and tightly-curled hair.

Negroid people are the people south of the Sahara Desert, "Negritos": various people scattered over southern and southeast Asia and nearby islands, and "Australoids" of Australia, Papua New Guinea, and Melanesia.....

According to Google:

Negroid
/ˈniːɡrɔɪd/
adjective OFFENSIVE•DATED
relating to the division of humankind represented by the indigenous peoples of central and southern Africa.

Dictionary.com | Meanings & Definitions of English Words

The world's leading online dictionary: English definitions, synonyms, word origins, example sentences, word games, and more. A trusted authority for 25+ years!

www.dictionary.com

adjective
Anthropology. (no longer in technical use) of, relating to, or characteristic of the peoples traditionally classified as the Negro race, especially those who originate in sub-Saharan Africa.
noun
Older Use: Usually Offensive. a member of such peoples.

So Google is saying the adjective is also dated and offensive.

Though my Aussie dad sometimes uses the word "negro". I think some black people "own" it by using it amongst themselves.

 

[lpetrich]

 Afroasiatic languages - lots of different classifications. So I'll take a somewhat unorthodox approach to combining them - try to see how common various proposed close relationships are.

They agree the most on Berber - Semitic. Then BS, Egyptian, and Chadic. Then BSEC and Cushitic:

(((Berber, Semitic), Egyptian, Chadic), Cushitic), Omotic

 

[lpetrich]

At the World Atlas of Language Structures:

WALS Online - Feature 136A: M-T Pronouns - a Eurasiatic feature: IE, Uralic, Turkic, Mongolian, Tungusic, Chukchi-Kamchatkan, Eskimo-Aleut. It occurs in a few places out of the Eurasiatic macrofamily.

WALS Online - Feature 137A: N-M Pronouns - while common in the Americas, other phonetic values are also common. N-M is mainly on the west coasts of the Americas, and it's mixed in with non-N-M languages there.

 

[lpetrich]

From  Semitic root I found Materials and Language: Pre-Semitic Root Structure Change Concomitant with Transition to Agriculture

Consider "water": *mây- often plural. Akkadian: mû, Arabic mâ, Aramaic mayyâ, Hebrew mayim, Ge'ez may - obviously 2-consonant. It is from Afro-Asiatic *ma/- (Berber aman, Egyptian maw, ...) also 2-consonant.

Words associated with materials:

• 2-consonant: water, fire (light), rock, flint, wet clay, mud, sand, dirt, wood, pole, stick, leather, tendon, cloth(es), lime, reed
• Both: pebble, stone, copper / bronze
• 3-consonant: leather, bitumen, straw, brick, dust, soil, ore, lead (metal), copper / bronze, salt (saltpeter), sulfur, pottery, clay vessel, wool, antimony, silver, gold, charcoal, to smelt, to refine, bellows

The 2c words are all pre-agricultural, while the 3c words reflect an agricultural economy with lots of mining and refining.

The authors propose that Semitic speakers had lived in the Middle East from the beginning of agriculture, but from Northeast-African Afro-Asiatic homeland hypotheses, Semitic speakers must have moved into the Middle East from NE Africa some 6,000 years ago. So they must have developed 3c roots before they settled down in the Middle East.

So original-agriculturalist languages families have been supplanted by at least two others: Indo-European and Semitic.

 

[lpetrich]

An oddity associated with Semitic root structure:  Broken plural - plural by vowel shift - very common in Arabic. Regular plurals: (masc) -ûn (nom), -în (gen, acc), (fem) -ât.

• Book: kitâb - kutub
• Writer, scribe: kâtib - kuttâb
• Letter: maktûb - makâtîb
• Desk, office: maktab - makâtîb

From root ktb "to write".

Arabic has several patterns of broken plurals, and it's hard for me to recognize any patterns in them.

 

[lpetrich]

The ancient Greek word for parrot and variations (Liddell & Scott):
psittakos
bittakos
sittakos
psittakê

Psittakos Psittake Bittakos Byttakos Sittake Sittakos Sittas - Aristophanes Birds

L) Psittakos was the commonest name in ancient Greek for a Parakeet (i.e. small Parrot), although variant forms coexisted: Byttakos (Ctesias 688F45 p.488 Jacobi) and Bittakos (Eubulus fr. 120.4 Kassel-Austin) before 335 BC, Psittake in Aristotle (HA 597b 27), in the post-classical period Sittakos (Philodemus On Poems 2.20.3 Hausrath, Aelian NA 16.2, Arrian Indica 15.8-9 citing Nearchos 133F9 Jacobi) and Sittake (Philostorgius Church History 3.11 p. 42.17 Bidez), possibly also Sittas (Hesychius c 772). These were all Hellenisations of an Indian name (siptace, according to Pliny NH 10.117) for Indian birds first mentioned by Ctesias around 500 BC (cf. also e.g. Aristotle HA 597b27-29, Pausanias 2.28.1, Aelian NA 13.18, Ovid Amores 2.6.1, Apuleius Florida 12), and later imported as pets into Greece and Rome.

The variation in the initial sound suggests some sound that did not in exist in Greek. *f- ? Did it alternate with *v- ?


That could also explain spongos "sponge" ~ Latin fungus "fungus, mushroom". Stated to have alternate form sfungus. Also Old Armenian sungn, Georgian sok.o, Tsez zik'u, Bezshta sak'o.

A Germanic cognate? English "swamp" and "sump" are from Proto-Germanic *sumpaz, *swammaz "sponge, mushroom, fungus, swamp, marsh"

A Slavic one also? Proto-Slavic *goba "fungus, mushroom, sponge, lip, mouth" - would have to be metathesis (exchange of sounds) - from *boga

The initial sound: Greek sp-, Latin f-, Germanic *s-, Slavic *b-

Premodern sponges were all sea sponges, because plastic sponges depend on the ability to make plastic.

 

[Bomb#20]

Borean: Nostratic, Dene-Caucasian, Sumerian, Ainu, Austric
...
So the "Out of Africa" migration split in two, one going eastward to south Asia, southeast Asia, New Guinea, Australia, and nearby islands, and one going northward to the Middle East and then to the bulk of Eurasia and the Americas.

So if the Borean family is a sensible concept, then these northward migrants would have spoken Proto-Borean.

Only if the time-frames line up. That seems dubious to me. The "Out of Africa" migrations happened presumably something like 50-70,000 years ago. Assuming the Borean family is valid linguistics and not just seeing patterns in clouds, what reason is there to think Proto-Borean is anywhere near that old? Languages change so quickly that traces from that far back would have decayed to invisibility by now; whereas, say, 20,000 years would probably have been enough time for some local language to expand and break up into the families now called Borean.

 

[Swammerdami]

Languages have a tendency to CONQUER.
In most of these examples, language replacement was accompanied by wholesale population replacement.

• The Philippines were inhabited (by "Negritos") before the Austronesians arrived, but AFAIK the Negrito language has completely disappeared, despite many islands and mountains where the language might have "hid out."
• The Bantu language "conquered" most of sub-Saharan Africa in historic times. AFAIK many of the languages of the Pygmys became extinct and unknowable.
• The Celtic languages dominated Europe in the time of early Rome. They are now extinct except for the Insular branches.
• A language family which may have dominated the Neolithic of Southwest Asia exists today only as the isolates Burushaski and Basque, and NE Caucasian languages like Chechen and Dagestan.

These examples suggest to me that macro-families like Eurasiatic or Dene-Caucasian (or even perhaps Borean) may have originated rather recently, as Mr. Bomb points out:

Only if the time-frames line up. That seems dubious to me. The "Out of Africa" migrations happened presumably something like 50-70,000 years ago. Assuming the Borean family is valid linguistics and not just seeing patterns in clouds, what reason is there to think Proto-Borean is anywhere near that old? Languages change so quickly that traces from that far back would have decayed to invisibility by now; whereas, say, 20,000 years would probably have been enough time for some local language to expand and break up into the families now called Borean.

There is a HUGE amount of information available from Y-chromosome haplogroups. For example, although the C haplogroup may have moved through India and into East Asia 55,000 years ago (or thereabouts), the common agnatic ancestor of Genghis the Mongol Khan and Geronimo the Apache lived about 15,000 years ago. (In fact the languages of Mongols and Apaches are NOT related; I just mention this recency to depict the recency of some common ancestries.)

One mystery is the R1b-V88 haplogroup. It separated from the rest of R1b 16,000 years ago, long before the "Kurgan people." It's rather rare today, but is found in ancient European skeletons (and in a few living Europeans especially in isolated Sardinia). The highest concentration of R1b-V88 is in northern Cameroon where it is associated with the Chadic language family. But the relationship, if any, between R1b-V88 migrations and Chadic language is probably unknowable.