1036: 4P

HISTORICAL LINGUISTICS

Outline

1 Genetic relations between languages

2 Historical linguistics

3

1/36

Phonological change in the history of ‘English’

Genetic relations between languages

Language classification

Typological classification

How living languages differ from each other Comparing grammatical properties of languages and grouping them according to the properties they share e.g. word order (SVO, SOV etc.) or morphological type (analytic, synthetic, agglutinative, etc.)

Genetic classification How languages change over time Tracing lineages of language families Comparing languages related genetically to each other Drawing conclusions about what specific changes took place Reconstructing dead languages that we don’t have records of

2/36

Genetic relations between languages

Genetic classification

Organizing language into language families A language, or a group of languages, evolve from another language

(1) The family tree of Romance languages (simplified)

Latin

Italian Spanish Portuguese French Romanian

3/36

Genetic relations between languages

What causes language birth?

How does this happen?

The driving force of language evolution is innovation A population speaking a language develops a linguistic innovation (a small change in their language)

"All languages have evolved from earlier forms of the same language as innovations have arisen, and persisted as new generations of children learned the somewhat changed language".

(Lyovin et. al. 1997)

We group descendant languages with their ancestors

4/36

Genetic relations between languages

Anagenesis

Middle English

Some lineages involve relations between a single ancestor and a single descendant

Old English Old Japanese

Middle Japanese

Modern English Modern Japanese

Anagenesis – evolution of one language from another in a straight, non-branching line

5/36

Genetic relations between languages

Innovations and retentions

An example of innovation SOUND CHANGE

E.g. at the end of Middle English, the sound [u:] changed into [aU].

6/36

Middle English Modern English present-day spelling [hu:s] [haUs] [nu:] [naU] [u:t] [aUt]

house now out

Some northern English dialect did not participate in the [u:] > [aU] change. They retained the original [u:].

(2) a. Innovation is a product of change b. Retention is the ancestral state without any change

Genetic relations between languages

The arbitrary divisions into Old, Middle and Modern stages of a language

Even though Old, Middle and Modern English are related by a straight, non-branching line, they are considered different languages. Just like Latin is considered a different language from Italian or French. How do we know exactly when Old English became Middle English?

(3) Approximate dates: a. Old English: c.600–1100 b. Middle English: c.1100–1540

These days are largely arbitrary and correlate with historical events, rather than reflect the actual time of change The actual changes were gradual

7/36

Genetic relations between languages

Cladogenesis

Part of the population speaking a language develops a linguistic innovation (a small change in their language) If that part of the population is (or becomes) isolated from the rest, that innovation will not spread to the rest of the population

Cladogenesis – the result of incomplete spread of innovations The ancestral languages thus splits into two (or more) daughters, at least one of which is characterized by innovations. Examples of cladogenesis:

Latin splitting into a number of dialects that gave rise to modern Romance languages The Germanic family tree

8/36

Italian (and the other Romance languages) developed from Latin:

Latin (tree incomplete)

Spanish Portuguese Catalan French Sardinian Italian Romanian

The Romance clade The Roman empire The Romance languages

Genetic relations between languages

The (Proto-)Germanic clade

9/36

Genetic relations between languages

Reconstructed languages

We don’t have records of the ancestor of all Germanic languages, i.e. the Proto-Germanic language

We indicate that by calling it a proto-language

For the same reason, its three daughters are also proto-languages: Proto- Western Germanic, Proto-Northern Germanic and Proto-Eastern Germanic Proto-languages are reconstructed, that is inferred from what we know about language change

We need to know how language change works ⇒ historical linguistics

10/36

Historical linguistics

Historical linguistics

11/36

Historical Linguistics studies how languages change.

This includes how languages are changing right now, and also how they changed in the past, which allows us to reconstruct earlier languages.

Historical linguistics

Synchronic vs. diachronic studies

12/36

Synchronic: single “slice” in time.

Diachronic: comparing multiple slices across time.

Historical linguistics studies diachrony.

Historical linguistics

From English to PIE

13/36

We saw that English, German, and other languages are Germanic, i.e. descedants of Proto-Germanic.

In turn, Proto-Germanic, Proto-Slavic, and others are Indo-European, i.e. descendants of Proto-Indo-European.

Historical linguistics

Proto-Indo-European (PIE)

PIE cladogenesis (simplified)

14/36

Historical linguistics

From English to PIE

Questions that historical linguists ask:

How did this happen?

More importantly, how do we know this happened?

15/36

These changes weren’t documented as they happened. We don’t even have records of the earlier languages.

Historical linguistics

Language change

16/36

This is all the result of language change, which is constant and inevitable.

It affects all levels of structure, although we’ll mostly talk about sound change.

Historical linguistics

Sound change

17/36

Sound change usually goes unnoticed by speakers.

An individual’s speech patterns can (and often do) change over their lifetime.

Historical linguistics

Does the Queen speak the Queen’s English?

Harrington et al. 2000. Journal of the International Phonetic Association 20:63-78.

Compared Christmas broadcasts from 1950s with those from the 1980s.

Vowel production had shifted towards that of standard southern British accent.

18/36

Historical linguistics

British English vowels

19/36

Historical linguistics

The changes

20/36

Historical linguistics

The uniformitarian principle

21/36

Although these changes are small, they can accumulate and result in more significant changes over time.

The same mechanisms which operated to produce the large-scale changes of the past may be observed in current changes.

That is, findings about present-day changes inform our understanding of historical changes.

Historical linguistics

The uniformitarian principle

22/36

The earlier stages of languages accessible to historical linguistics were just languages, i.e. not fundamentally different from the kinds of languages that exist now.

Earlier languages were subject to the same kinds of changes as modern languages, because they were just languages.

Historical linguistics

A corollary

23/36

Change is not decay, degeneration, or corruption.

Nor is it progress!

Historical linguistics

Next: Sound changes in the history of English

24/36

Both recent and directly observable,

as well as very old ones that ultimately allow us to relate English to other languages.

Phonological change

The Great Vowel Shift

25/36

English used to sound more similar to the other Germanic languages.

In the 1400s, the pronunciation of vowels began to change. The GVS was more or less complete by 1800.

It’s one of the reasons English spelling is so bad, and why e.g. the vowels spelled as a in sane (mid) and sanity (low) are so different.

Phonological change

From Middle English to Modern English

ME (1400s) Early ModE ModE 1400s 1500s Present [tid] [təjd] [tajd] ‘tide’ [hus] [həws] [haws] ‘house’ [fet] [fit] [fit] ‘feet’ [fol] [ful] [ful] ‘fool’ [tak] [tæk] [tejk] ‘take’ [bɛt] [bet] [bit] ‘beat’

26/36

Phonological change

Never ending change

27/36

Vowel shifts were not only a thing of the past.

They are going on right now!!

Northern Cities Vowel Shift: change in vowels that hadn’t been affected by the GVS.

Phonological change

Northern Cities Vowel Shift

28/36

Phonological change

Can you identify the word?

29/36

➤ ➤ ➤ ➤ ➤ ➤ ➤ ➤ ➤ ➤ ➤ ➤

Phonological change

The facts

30/36

word phrase sentence block➤ living on one block➤ old senior citizens➤

living on one block desk➤ behind the desk➤ so I like that, you know,➤

sitting behind the desk socks➤ had to wear socks➤ You had to wear socks,➤

no sandals. busses➤ busses with the➤ I can remember vaguely➤

antennas when we had the busses with the antennas on top.

Phonological change

Northern Cities Vowel Shift

31/36

A shift in vowel quality of several North American vowels.

Extensively studied by William Labov (U Penn).

http://www.youtube.com/watch?v=9UoJ1-ZGb1w

Phonological change

Some of the steps in the shift

Raising and lengthening of [æ].

32/36

word unshifted shifted stack [stæk] [steək]

Phonological change

Some of the steps in the shift

Raising and lengthening of [æ]. Fronting of [ɒ].

32/36

word unshifted shifted stack [stæk] [steək] stock [stɒk] [stak], [stæk]

Phonological change

Some of the steps in the shift

Raising and lengthening of [æ]. Fronting of [ɒ]. Lowering of [ɔ].

32/36

word unshifted shifted stack [stæk] [steək] stock [stɒk] [stak], [stæk] stalk [stɔk] [stɒk]

Phonological change

Some of the steps in the shift

Raising and lengthening of [æ]. Fronting of [ɒ]. Lowering of [ɔ]. Backing of [ʌ].

32/36

word unshifted shifted stack [stæk] [steək] stock [stɒk] [stak], [stæk] stalk [stɔk] [stɒk] stuck [stʌk] [stɔk]

Phonological change

NCS in action

33/36

Phonological change

NCS in action

34/36

Phonological change

NCS in action

35/36

Phonological change

Where does this happen?

36/36

Understanding language change

11/59

So if changes are basically arbitrary, how can we say anything about language relatedness?

Luckily, changes tend to be patterned and regular.

English through the ages, about 6,000 years ago

12/59

gIhdyestronom kImtom pihonto¯m ˚

egwment agusihms ˚

bherontes, podmis gIhonghontes, wogIhnomis-kwe porontes.

rēgm tosyo-kwe wirōm dekImt ˚ ˚

egwhnent.

English through the ages, about 6,000 years ago

13/59

domoms edhegwhent, eghrebhont monoghom gIhltom ˚

grhnom-kwe, gwō s, porkIoms, ˚

plhu-kwe alyosyo kweturpodos pekIwos. ˚

Fast forward 4,000 years . . .

14/59

English through the ages, about 2,000 years ago

15/59

gIhdyestronom gestrana

kImtom hu˚nda

pihontōm fiyandō

egwment kw̄ amun

agusihms okusō ˚

bherontes, podmis gIhonghontes, wogIhnomis-kwe porontes. birandya, fō tum gangandya, wagnum-uhw farandya.

rēgm tosyo-kwe wirōm dekImt egwhnent. kun i̊nga Tais-uhw wirō tehun̊ T slo¯gun.

English through the ages, about 2,000 years ago

16/59

domoms edhegwhent, eghrebhont hū sō burnun, grubun

monoghom managa

gIhltom gul̊ Ta

grhnom-kwe, gwōs, k˚urna-hw, kwō ,

porkIoms, farhans

plhu-kwe alyosyo fi̊ lu-hw alyes

kweturpodos fidurfo t̄iya

pekIwos. fihwō .

Another 1,000 years . . .

17/59

English through the ages, about 1,000 years ago

18/59

gIhdyestronom kImtom gestrana giestron

hůnda hund

pihontōm fiyandō féonda

egwment kw̄ amun cō mon

agusihms okusō ̊ æxa

bherontes, podmis gIhonghontes, wogIhnomis-kwe porontes. birandya, fō tum gangandya, wagnum-uhw farandya. berende, fō tum gangende, and on wægnum farende.

r̄ egm kun i̊nga Tonecyning

tosyo-kwe Tais-uhw

and Tæs

wirōm wirō weora

dekImt egwhnent. tehun̊ T slo¯gun. tígen slo¯gon.

English through the ages, about 1,000 years ago

19/59

domoms edhegwhent, eghrebhont hū sō T̄a hū s

burnun, hie burnon,

grubun grubon

monoghom managa manig

gIhltom gul̊ Ta gold

grhnom-kwe, gwōs, k˚urna-hw, kwō , and corn, cȳ,

porkIoms, farhans fearas,

plhu-kwe fi̊ lu-hw

and fela

alyosyo alyes elles

kweturpodos fidurfo t̄iya fyderf¯etes

pekIwos. fihwō . féos.

Yet another 1,000 years . ..

20/59

English through the ages, present time

21/59

gIhdyestronom kImtom gestrana giestron

hůnda hund

pihontōm fiyandō féonda

egwment kw̄ amun cō mon

agusihms okusō ̊ æxa

bherontes, podmis gIhonghontes, wogIhnomis-kwe porontes. birandya, fō tum gangandya, wagnum-uhw farandya. berende, fō tum gangende, and on wægnum farende.

r̄ egm kun i̊nga Tonecyning

tosyo-kwe Tais-uhw

and Tæs

wirōm wirō weora

dekImt egwhnent. tehun̊ T slo¯gun. tígen slo¯gon.

English through the ages, present time

21/59

gIhdyestronom gestrana giestron Yesterday

kImtom hu˚nda hund 100

pihontōm fiyandō féonda enemies

egwment kw̄ amun cō mon came

agusihms okusō ̊ æxa axes

bherontes, podmis gIhonghontes, wogIhnomis-kwe porontes. birandya, fō tum gangandya, wagnum-uhw farandya. berende, fō tum gangende, and on wægnum farende. carrying, on foot going, and in wagons riding.

wirōm dekImt egwhnent. wirō tehun̊ T slo¯gun.

r̄ egm kun i̊nga Tonecyning The king

tosyo-kwe Tais-uhw

and Tæs and of his

weora men

tígen ten

slo¯gon. they killed.

English through the ages, present time

22/59

domoms hū sō T̄a hū s

edhegwhent, bunun hie burnon,

eghrebhont grubun grubon

monoghom managa manig

gIhltom gul̊ Ta gold

grhnom-kwe, k˚urna-hw, and corn,

gwōs, kwō , cȳ,

porkIoms, farhans fearas,

plhu-kwe fi̊ lu-hw

and fela

alyosyo alyes elles

kweturpodos fidurfo t̄iya fyderf¯etes

pekIwos. fihwō . féos.

English through the ages, present time

22/59

edhegwhent, eghrebhont burnun, grubun

domoms hū sō T̄a hū s The houses

hie burnon, grubon they burned, took

monoghom managa manig much

gIhltom gul̊ Ta gold gold

grhnom-kwe, k˚urna-hw, and corn, and corn,

gwō s, kwō , cȳ , cows,

porkIoms, farhans fearas, farrows,

plhu-kwe fi̊ lu-hw

and fela and much

alyosyo alyes elles else

kweturpodos fidurfo t̄iya fyderf¯etes four-footed

pekIwos. fihwō . féos. property.

Legend

23/59

Indo-European

Proto-Germanic

Old English

Modern English

Indo-European

24/59

Proto-Indo-European: a real language.

Spoken ∼ 6,000 years ago (± 1,000 years).

Some of its grammar and lexicon can be inferred from its daughter languages.

Jones

Sir William Jones (1746–1794)

British judge, ended up in Calcutta.

Noticed that Sanskrit bore a curious resemblance to Greek and Latin . ..

25/59

A common source

26/59

“The Sanscrit language, whatever be its antiquity, is of a wonderful structure; more perfect than the Greek, more copious than the Latin, and more exquisitely refined than either, yet bearing to both of them a stronger affinity, both in the roots of verbs and the forms of grammar, than could possibly have been produced by accident; so strong indeed, that no philologer could examine them all three, without believing them to have sprung from some common source, which, perhaps, no longer exists; there is a similar reason . . . for supposing that both the Gothic and the Celtic . . . had the same origin with the Sanscrit; and the old Persian might be added to the same family.” Address at the third meeting of the Asiatic Society, 1786

PIE cladogenesis (simplified)

27/59

Rask

Rasmus Rask (1787–1832)

Philologist, PIE scholar, formidable polyglot.

Did a lot of documentation of European cognates.

No access to Sanskrit

28/59

Cognates

29/59

“When agreement is found in [the most essential] words in two languages, and so frequently that rules may be drawn up for the shift in letters [sounds] from one to the other, then there is a fundamental relationship between the two languages; especially when the similarities in the inflectional system and in the general make-up of the languages correspond with them.”

Investigation on the Origin of the Old Norse or Icelandic Language, 1818

The IE family

30/59

English, French, Greek, Armenian, Urdu, Russian, . . . all descended from a single language.

They are daughter languages of a single proto-language.

Intermediate stages (subgroups) may be established based on shared non-trivial innovations.

Proto-Germanic

31/59

Spoken ∼ 2,000 years ago

No surviving texts, but Gothic & early Scandinavian runic inscriptions come close.

IE cognates

32/59

What Rask noticed:

Greek Latin English patēr pater father treis tres three hekatón centum [k] hundred

IE cognates

32/59

What Rask noticed:

Greek Latin English patēr pater father treis tres three hekatón centum [k] hundred

IE cognates

32/59

What Rask noticed:

Greek Latin English patēr pater father treis tres three hekatón centum [k] hundred

Grimm

Jacob Grimm (1785–1863)

Did a lot of linguistics when he wasn’t writing fairy tales.

In spite of the fact that Rask discovered the PIE > PGmc correspondences first, the ‘law’ bears Grimm’s name.

33/59

Grimm’s Law

34/59

p t k → f θ h

b d g → p t k

bh dh gh → b d g

How can we tell?

35/59

Shift in Shift in

Sanskrit Sanskrit

Greek Greek

Latin Latin

English English

Germanic p > f pad- pod- ped- foot t > θ tanu- tanaós tenuis thin k > h çatam hekatón centum hundred b > p — — lubricus slippery d > t daça déka decem ten g > k ajras agrós ager acre

bh dh gh

> b bhrātā > d vidhavā > g (g)hansas

phr´̄atēr ēítheos khḗn

frāter vidua (h)ānser

brother widow goose

Legend

36/59

Indo-European

Proto-Germanic

Old English

Modern English

English through the ages, present time

37/59

gIhdyestronom kImtom gestrana giestron

hůnda hund

pihontōm fiyandō féonda

egwment kw̄ amun cō mon

agusihms okusō ̊ æxa

bherontes, podmis gIhonghontes, wogIhnomis-kwe porontes. birandya, fō tum gangandya, wagnum-uhw farandya. berende, fō tum gangende, and on wægnum farende.

r̄ egm kun i̊nga Tonecyning

tosyo-kwe Tais-uhw

and Tæs

wirōm wirō weora

dekImt egwhnent. tehun̊ T slo¯gun. tígen slo¯gon.

English through the ages, present time

37/59

gIhdyestronom gestrana giestron Yesterday

kImtom hu˚nda hund 100

pihontōm fiyandō féonda enemies

egwment kw̄ amun cō mon came

agusihms okusō ̊ æxa axes

bherontes, podmis gIhonghontes, wogIhnomis-kwe porontes. birandya, fō tum gangandya, wagnum-uhw farandya. berende, fō tum gangende, and on wægnum farende. carrying, on foot going, and in wagons riding.

wirōm dekImt egwhnent. wirō tehun̊ T slo¯gun.

r̄ egm kun i̊nga Tonecyning The king

tosyo-kwe Tais-uhw

and Tæs and of his

weora men

tígen ten

slo¯gon. they killed.

English through the ages, present time

38/59

domoms hū sō T̄a hū s

edhegwhent, burnum, hie burnon,

eghrebhont grubun grubon

monoghom managa manig

gIhltom gul̊ Ta gold

grhnom-kwe, k˚urna-hw, and corn,

gwōs, kwō , cȳ,

porkIoms, farhans fearas,

plhu-kwe fi̊ lu-hw

and fela

alyosyo alyes elles

kweturpodos fidurfo t̄iya fyderf¯etes

pekIwos. fihwō . féos.

English through the ages, present time

38/59

edhegwhent, eghrebhont burnun, grubun

domoms hū sō T̄a hū s The houses

hie burnon, grubon they burned, took

monoghom managa manig much

gIhltom gul̊ Ta gold gold

grhnom-kwe, k˚urna-hw, and corn, and corn,

gwō s, kwō , cȳ , cows,

porkIoms, farhans fearas, farrows,

plhu-kwe fi̊ lu-hw

and fela and much

alyosyo alyes elles else

kweturpodos fidurfo t̄iya fyderf¯etes four-footed

pekIwos. fihwō . féos. property.

Question words in English and Romance

39/59

English: what, where, when, why, . . .

Modern pronunciation: [w-]

Archaic pronunciation: [hw-] ([w]) (and how some speakers still pronounce them) ˚

Spanish: qué ‘what’, quién ‘who’, cuándo ‘when’, . . . Pronunciation: [k-] or [kw-], but the orthography always has a trace of the Latin pronunciation: [kw-].

Archaic English: [hw-]

Spanish: [k(w)-]

40/59

(e.g. when)

(e.g. cuando)

It’s Grimm’s Law!

k > h

It’s the same rule that turned the Latin centum [kentum] into hundred.

How?

41/59

What about English how? Is it a wh-word?

How was it pronounced before the Great Vowel Shift?

The systematicity of sound change

42/59

Grimm’s law is basically exceptionless: it’s confirmed by many other cognate series in the daughter languages.

This entails it’s not a change in the lexicon, but in the phonological system, like the GVS and the NCVS.

The systematicity of sound change

But it was systematic in another sense: it can be described as three simple changes in voicing or manner of articulation (as opposed to nine):

voiceless stop → fricative unaspirated voiced stop → voiceless

voiced aspirated stop→ unaspirated

Importantly, place is maintained.

43/59

Grimm’s Law

44/59

p t k → f θ h

b d g → p t k

bh dh gh → b d g

A golden standard

45/59

This and other discoveries in the study of IE are a gold standard.

We expect sound change to be systematic in this sense: supported by many cognates, and expressable in terms of simple phonological processes.

The comparative method

46/59

The comparative method used to make these discoveries has been used to group other languages and reconstruct their proto-language ancestors.

The comparative method

The comparative method

47/59

Two languages are related (= used to be the same language) if they exhibit recurring correspondences in basic vocabulary.

The comparative method

Recurring correspondences in basic vocabulary

48/59

Recurrence ensures that similarities aren’t accidents.

Correspondences: not only identities and similarities, but any kind of regular alignment.

Basic vocabulary: words that are most likely to show the highest level of lexical continuity over time.

The comparative method

Recurrent correspondences

49/59

The objective is to figure out the (phonological) grammar of the parent language, as well as systematic changes to the daughter languages.

Accidental similarities don’t tell us anything about this.

The comparative method

Recurrent correspondences

50/59

We ensure the correspondences are not accidental by:

Finding many cognate series illustrating the same correspondence.

Being able to express the changes to the daughter languages in terms of simple rules (or series of simple rules).

The comparative method

Correspondences in the Romance language family

51/59

Italian Spanish Portuguese French Gloss 1. corpo cuerpo corpo corps body 2. crudo crudo cru cru raw 3. catena cadena cadeia chaîne chain 4. cacciare cazar caçar chasser to hunt

Italian Spanish Portuguese French environment k k k k elsewhere k k k ʃ before [a]

The original Latin words had [k] everywhere.

In French, there was a sound change in one specific context.

https://www.youtube.com/watch?v=_enn7NIo-S0

The comparative method

Polynesian

52/59

A group within the larger family of Austronesian languages.

It includes Hawaiian, Maori, Samoan, Takuu, Tongan, . . . .

Diversification started about 2,000 years ago.

The comparative method

Proto-Polynesian nasals

53/59

Can you fill in the blanks for Proto-Polynesian?

PPn Haw Mao Sam Tak Ton * ofo noho noho nofo nofo nofo ‘sit, dwell’ *ma u manu manu manu manu manu ‘bird’ * ifo niho niho nifo nifo nifo ‘tooth’ *la i lani raŋi laŋi rani laŋi ‘sky’

The comparative method

Proto-Polynesian nasals

54/59

PPn Haw Mao Sam Tak Ton *nofo noho noho nofo nofo nofo ‘sit, dwell’ *manu manu manu manu manu manu ‘bird’ *nifo niho niho nifo nifo nifo ‘tooth’ *laŋi lani raŋi laŋi rani laŋi ‘sky’

Proto-Polynesian had both [n] and [N]. [N] changed to [n] in Hawaiian and Takoo; [n] remained unchanged in all the daughter languages.

The comparative method

What Polynesian nasals tell us

55/59

We need to look at the whole system (or as much of it as we can).

Conclusions about one recurrent correspondence can be the key to understanding some other recurrent correspondence.

The comparative method

Are some daughters more ‘faithful’?

56/59

Not necessarily.

It’s even possible for the mother language to be different from all the daughter languages.

The comparative method

Proto-Polynesian l and r

57/59

What did Proto-Polynesian have in places where its daughters have [r] or [l]?

Haw Mao Sam Tak Ton lau rau lau rau lau ‘leaf’ ʔulu kuru ʔulu kuru kulu ‘breadfruit’ — uru ulu uru huu ‘enter’ lani raŋi laŋi rani laŋi ‘sky’ walu waru valu varu vau ‘scrape’ ala ara ala ara ʔaa ‘awake’ hale whare fale fare fale ‘house’ hala whara fala fara faa ‘screw pine’

The comparative method

Proto-Polynesian l and r

58/59

Polynesian had both [r] and [l]; the key is Tongan:

[r] changed to [l] in Hawaiian, Samoan.

[l] changed to [r] in Maori, Takuu.

[r] was deleted in Tongan.

The comparative method

Proto-Polynesian l and r

58/59

Polynesian had both [r] and [l]; the key is Tongan:

[r] changed to [l] in Hawaiian, Samoan.

[l] changed to [r] in Maori, Takuu.

[r] was deleted in Tongan.

All the daughters underwent changes.

The comparative method

What l and r in Polynesian tells us

59/59

It’s not about finding the ‘oldest’ language.

The mother language is a different language!