23-10-2016

Language Change:
History of Language

Sir William Jones

(28 September 1746 –27 April 1794, Asiatic Society Founder)

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,
though not quite so forcible, for supposing that both the
Gothic and the Celtic, though blended with a very different
idiom, had the same origin with the Sanscrit; and the old
Persian might be added to the same family.
(Address to the Asiatic Society of Calcutta 1786, founded
1784)

1
 23-10-2016

Paternoster/Lord’s prayer
Latin liturgical version: Old English (c.1100) Matt. 6.9
Fæder ure þu þe eart on heofonum; Si
Pater noster, qui es in caelis: þin nama gehalgod to becume þin rice
sanctificetur Nomen Tuum; gewurþe ðin willa on eorðan swa swa
adveniat Regnum Tuum; on heofonum.
fiat voluntas Tua, Middle English (c. 1380)
sicut in caelo, et in terra. Oure fadir that art in heuenes, halewid
Northumbrian (c.650AD) be thi name; thi kyndoom come to; be
FADER USÆR ðu arðin heofnu thi wille don in erthe as in heuene:
Sie gehalgad NOMA ÐIN. Early Modern English (c. 1559), Book of
Tocymeð RÍC ÐIN. Common Prayer
Sie WILLO ÐIN Our Father which art in heaven,
suæ is in heofne and in eorðo. hallowed be thy name. Thy kingdom
come. Thy will be done in earth as it is
Late Modern English (1970) in heaven.
Our Father in Heaven,
let your holy name be known, Late Modern English (BCP 1928)
Our Father, who art in heaven,
let your kingdom come, hallowed be thy Name. Thy kingdom
and your will be done, come. Thy will be done, On earth as it is
on earth as in heaven. in heaven.

Languages Change
All the time. How and why?
1. Language learning
2. Language contact
3. Change for social reasons
4. Natural linguistic processes
Is this good or bad? Who cares?
All stages of a language are valid expressions of the language
just as all dialects are valid expressions of it. And none violate
the underlying structural properties

2
 23-10-2016

Language Learning
Language is transformed as it is transmitted from one generation
to the next. (Over-) regularisation (scant was scamt)
The child must construct their language (grammar and lexicon)
based on the input received (from parents, older siblings and
other members of the speech community).
1. This process (of linguistic replication) is imperfect and so the
results are variable
2. Bias towards regularization – learning an irregular form
requires more input, so systematic drift
3. Also random differences may spread, especially through a
small population , and become fixed

Language Contact
Entirely new languages can be created this way: (Pidgins and
Creoles)
More commonly, new ‘bits’ are added to existing languages
which don’t change their identity (immediately)
1. Through migration, conquest, trade
2. Adults may learn the new language as a second language
3. Children may be fully bilingual
4. Results in borrowing of words, sounds, even syntactic
constructions

3
 23-10-2016

Tok Pisin (Papua New Guinea)

The Lord's Prayer
Papa bilong mipela Yu stap long heven.
Nem bilong yu i mas i stap holi.
Kingdom bilong yu i mas i kam.
Strongim mipela long bihainim laik bilong yu long graun,
olsem ol i bihainim long heven.

Tok Pisin pronouns

4
 23-10-2016

Borrowing

Change for social reasons

Social groups adopt distinctive norms of dress,
adornment, gesture and so forth; language is part of the
package.
Linguistic distinctiveness can be achieved through
vocabulary (slang or jargon), pronunciation (usually via
exaggeration of some variants already available in the
environment), morphological processes, syntactic
constructions, and so on.
People change the way they speak according to their
social status or the way they would like it to be seen.
Educational policies.

5
 23-10-2016

Natural linguistic processes

Some examples:
Assimilation: NPR /n. pi. ar/ > [m pi ar]
Dissimilation: /haws/ + [PL] > [hawsǝz]
Syncope (deletion): /lajbrəri/ > [lajbri]
/ʧɪ:z stejk/ > [ʧi:stejk]
rapid or casual speech produces assimilation,
vowel reduction, deletion
this pronunciation can become conventionalized,
and so end up being produced even in slower,
more careful speech

Classifying Languages
Either because of what we know or because of
what we can work out, we can classify
languages according to their origins.
Kinship terminology is used and languages are
put into families (and sub-families) the
relationships between languages are described
using female familial terms: most often
daughter and mother

6
 23-10-2016

Indo-
European

7
 23-10-2016

Dravidian
Languages

8
 23-10-2016

Uto-Aztecan (native American)

9
 23-10-2016

Spread of Indo-European

Classifying Languages: Romance

For a long time we have known that many of the languages
now spoken in Western Europe (and elsewhere) are
descended from Latin:

10
 23-10-2016

Classifying Languages: Germanic

Paternoster revisited
Afrikaans German
Ons Vader wat in die hemel is, Vater unser im Himmel,
laat u Naam geheilig word;
laat u koninkryk kom; Geheiligt werde dein Name.
laat u wil ook op die aarde Dein Reich komme.
geskied, Dein Wille geschehe, wie im Himmel
net soos in die he so auf Erden.
Danish Icelandic
Vor Fader, du som er i Himlene! Fa›ir vor, ßú sem er á himnum.
Helliget vorde dit navn;
komme dit rige; Helgist ßitt nafn, til komi ßitt ríki,
ske din vilje ver›i ßinn vilji, svo á jör›u sem á himni.
på jorden, som den sker i Swedish
Himmelen; mel. Vår fader, du som är i himlen.
Dutch-Flemish Låt ditt namn bli helgat.
Onze Vader in de hemel, Låt ditt rike komma.
uw naam worde geheiligd,
uw koninkrijk kome, Låt din vilja ske,
uw wil geschiede, på jorden så som i himlen.
op aarde zoals in de hemel.

11
 23-10-2016

Cognates: Words in two or more daughter languages that

derive from the same word in the ancestral language
English Dutch German Danish
one een ein en
two twee zwei to
three drie drei tre
four vier vier fire
five vijf fünf fem
six zes sechs seks
seven zeven sieben syv
eight acht acht otte
nine negen neun ni
ten tien zehn ti

Cognates, more distant

12
 23-10-2016

Language Classification: How?

We rely on two things:

the Uniformitarian Principle

the absolute regularity of sound-change

The Uniformitarian Principle

‘knowledge of processes that operated in the past
can be inferred by observing ongoing processes in
the present’

or, for language:

‘Language must work now in the same way as it

ever did’

13
 23-10-2016

Regularity of Sound-Change
Most of historical linguistics relies on the assumption
that
sound-change is regular and without exception
That is, any sound-change will affect all the words that
contain that (combination of) sound(s).
Linguists rely on systematic sound changes to establish
the relationships between languages.
The basic idea is that when a change occurs within a
speech community, it gets diffused across the entire
community of speakers of the language.

Systematic sound change

o If the communities have split and are no longer in
contact, a change that happens in one community
does not get diffused to the other community.
o Thus a change that happened between early and late
Latin would show up in all the 'daughter' languages of
Latin, but once the late Latin speakers of the Iberian
peninsula were no longer in regular contact with other
late Latin speakers, a change that happened there
would not spread to the other communities.
o Languages that share innovations are considered to
have shared a common history apart from other
languages, and are put on the same branch of the
language family tree.

14
 23-10-2016

Romance cognates

English
French Italian Spanish Portuguese Rumanian Catalan
Gloss

mother mer madre madre mae mama mare

father per padre padre pae tata pare

Polynesian languages (W-E)

English Hawai-
Tongan Maori Samoan Tahitian
Gloss 'ian
1. bird manu manu manu manu manu
2. fish ika ika iʔa iʔa iʔa
3. to eat kai kai ʔai ʔai ʔai
4. forbidden tapu tapu tapu tapu kapu
5. eye mata mata mata mata maka
6. blood toto toto toto toto koko

15
 23-10-2016

Differences – sound changes

First, Nasal consonants and vowels remain unchanged
Second, Tongan and Maori have /k/, Samoan, Tahitian and
Hawai'ian appear to have /ʔ/ (glottal stop).
unconditioned change from /k/ to /ʔ/ in the Eastern branch
or a change from / ʔ / to /k/ in the Western branch of this family.
We choose the first as more likely,
1. because /k/ is a more common phoneme in the world's
languages
2. because backing of consonants is more common than fronting
as a phonological process
3. because of what we know about the history: Polynesia was
peopled from west to east, and if the change had occurred in the
Western branch, that would have been at a time when all five
languages were still one speech community.

Another difference
4. forbidden tapu tapu tapu tapu kapu
5. eye mata mata mata mata maka
6. blood toto toto toto toto koko

There is a systematic correspondence between /t/ in the

first four languages and /k/ in the easternmost –
Hawai'ian - systematic, unconditioned sound change,
this time in only one language.
When English borrowed the Polynesian word for
"forbidden", we borrowed it from one of the languages
west of Hawaii -- we say "taboo", not "kaboo".

16
 23-10-2016

This is what a family tree of the five Polynesian languages would

look like, based on the small data set above (the picture is
somewhat more complex when we look at other cognate sets --
Maori in particular is probably not correctly placed in this diagram,
which has been designed as an illustration of the method):

Benefits?
First, a systematic pattern of phonological
correspondence across many words is unlikely to have
arisen by chance, whereas completely unrelated
languages often develop surprising similarities in
particular words [whole], entirely by chance.
Second, given systematic patterns of this type, we can
start to apply the comparative method to reconstruct
the parent language. This in turn allows us to examine
relationships among reconstructed languages at a
greater time depth, even if the process of change
entirely obscures the relationships among the
vocabulary items in the daughter languages.

17
 23-10-2016

Difficulties
oEstablishing patterns of this type is difficult.
oIt requires a large vocabulary in all the languages
being compared, in order to find enough cognates
oIt also requires a deep knowledge of the grammar
of each of the languages, in order to see cognate
relationships that might be obscured and not to be
fooled into seeing false cognates.

How far back can we go?

Our methods for reconstruction will take as only as far back as
about 5000 - 7000 years; after that, the number of cognate sets
available for reconstruction becomes just too low to give results
that can be reliably distinguished from chance relationships.
A minority of scholars, however, argue that this is possible:
Nostraticists, derived from their views that there exists a super-
family of language they have called the "Nostratic", which would
include not only Indo-European languages but also Uralic
languages such as Finnish and Hungarian, Altaic languages such as
Mongolian and Turkish, and Afro-Asiatic languages such as Arabic,
Hebrew, Hausa and Somali
The current Ethnologue listing of “language families” includes 108
members, from Afro-Asiatic to Zaparoan. This does not mean that
human language was developed independently 108 times -- it only
means that generally-accepted methods can't establish any
further relationships among these groupings.

18
 23-10-2016

English-es
5th century Germanic invasions replaced Latin and
the Celtic languages with Germanic Dialects (Anglo-
Saxon: Angles, Saxons and Jutes)
Changes in Anglo-Saxon are usually grouped as:

19
 23-10-2016

English Orthography

Shouldn’t it be…?

20
 23-10-2016

Regularity of Sound-Change
The Great Vowel Shift (primarily between 15-16th c.); the
big difference between Chaucer and Shakespeare

21
 23-10-2016

Conversations across time:

Middle English
Cole: Is that thy child?
Alice: Yea, hir name is Ann.
Cole: A good and holy
name.
Alice: Soon she will be
three years of age.
Cole: Will she speke to
me?

1450 to 1550
Cole: Is that thy child?
Alice: Yea, hir name is
Ann.
Cole: A good and holy
name.
Alice: Soon she will be
three years of age.
Cole: Will she speke to
me?

22
 23-10-2016

1550 to 1650
Cole: Is that thy child?
Alice: Yea, her name is
Ann.
Cole: A good and holy
name.
Alice: Soon she will be
three years of age.
Cole: Will she speak to
me?

1650-1750
Cole: Is that thy child?
Alice: Yea, her name is
Ann.
Cole: A good and holy
name.
Alice: Soon she will be
three years of age.
Cole: Will she speak to
me?
Alice: Yea, she speaks
wonder loud.

23
 23-10-2016

Long vowels of Middle English

Stepwise shift

24
 23-10-2016

High-mid vowels raise

25
23-10-2016

26
23-10-2016

27
23-10-2016

28
23-10-2016

29
23-10-2016

30
23-10-2016

31
 23-10-2016

Long vowels of Middle English

32
 23-10-2016

Other processes
Vowel length was closely tied to syllable structure in ME
(a) Some long vowels shortened in closed syllables
(b) Short vowels commonly lengthened in open syllables

Consequences
Since only long vowels participated in the Great Vowel
Shift, we’re left with some interesting alternations MoE:
OE ke:p-te shortened to kep-te (shortening of vowels in
closed syllables) so it didn’t raise in the GVS
But the vowel in OE ce:-pan raised to /i/ and hence, the
present-past alternation keep [i] ~ kept [ɛ]
OE nɔ-su lengthened in ME to nɔ:se, GVS to give [no:s]
nos.tyrl (< nosu + tyrel ‘nose+hole’) stayed short, hence
today we have nose ~ nostril/nozzle
OE ste-lan > ME ste:le hence steal ~ stealth etc.
Writ (no GVS) < OE writ vs. write (GVS)< OE wri:-tan

33
 23-10-2016

Conditioned sound change

begins with allophonic variation (assimilation)

34
 23-10-2016

Conditioning environment changes

Complementary distribution lost because of deletion

At this point, we have phonemic change.

In this case, two new phonemes, /ã/ and /õ/ have emerged
(actually, other nasalized vowels also emerged, but we’re
keeping it simple)

And this may also create minimal

pairs

35
 23-10-2016

Unconditioned sound change –

across the board

Middle Kannada changes

Palli ⟶Halli, (Village)
Pagalu ⟶ Hagalu, (Day time)
Pandhi ⟶ Handhi, (Pig)
Paalu ⟶ Haalu, (Milk)
Poraata ⟶ Horaata (fight)
Puli ⟶ Huli (tamarind)

36
 23-10-2016

Palatalization of Latin k

37
 23-10-2016

Sound-Syntax
Sound change can have important consequences for
morphology, and ultimately for syntax

Latin singular noun paradigms

38
 23-10-2016

But…

39
 23-10-2016

Sound change is uniform

‘… and exceptionless’
Consider:
OE cnafa [knava] > ModE [nejv] ‘knave’
OE cniht [knixt] > ModE [najt] ‘knight’
So what’s the rule?
Modern English reflex of OE cyning [kyniŋ] > [kiŋ]
Rule: k ⟶Ø / -n

The Comparative Method

If we assume that sound-change is regular and
exceptionless in this way, we can use
systematic comparison of languages to see the
relationships between them.

This is known as the Comparative Method.

40
 23-10-2016

The Comparative Method: Part II

Sanskrit Greek Latin Gothic English PIE
pita pate:r pate:r fadar father *pter-
padam poda pedem fotu foot *ped-
bhratar phrate:r frate:r broþar brother *bhrater-
bharami phero fero baira bear *bher-
sanah hene: senex sinista [senile] *sen-
tris tre:s þri three *trei-
deka dekem taihun ten *dekm-
he-katon kentum hund(raþ) hundred *dkm-tom-

Grimm’s law, Rasmus Rask

Aspirated sounds > fricatives, voiced > voiced stops

41
 23-10-2016

Century, cardiac, canine, cornucopia, cuticle

42
 23-10-2016

The Comparative Method: Part III

So far, so old news (but remember that
this wasn’t old news when the
systematic correspondences were
discovered in the nineteenth century!)
Is there anything that confirms the
validity of the Comparative Method?
Some language that shows a feature
predicted by the Comparative Method,
but wasn’t known when the CM was
worked out?
There is …

The Comparative Method: Hittite

Towards the end of the nineteenth century, there
were a few places in IE reconstruction where the
sound-changes didn’t seem to be as regular as they
should be. Relying on the absolute regularity of
sound-change, Ferdinand de Saussure (in 1879)
posited a set of three segments (based on the
variable vowel quality in the daughter languages)
that must have existed but had died out in all of the
languages that were known up to then.
He said that they were probably laryngeals.

43
 23-10-2016

The Comparative Method:

Hittite
A set of clay tablets in familiar cuneiform script but in an unfamiliar
language from ancient Turkey were deciphered (1600-1300BC) by
Bedřich Hrozný; it was worked out that they were in fact in Hittite
(an Anatolian language), which became the oldest known IE
language.
Hittite: approximately 30,000 cuneiform tablets or fragments of
tablets preserved in the archives of the Hittite capital city, Hattusa
(near the modern town of Boğazkale, formerly Boğazköy, Turkey),
and various provincial centres; the majority of the tablets are from
the period of the Hittite empire (c. 1400–c. 1180 bce). Texts in Old
Hittite, a form of the language used between approximately 1650
and 1500 bce, are mostly preserved in copies made during the
empire period; these texts preserve the earliest examples of an
Indo-European language that have thus far been found.

The Comparative Method:

limitations
Note what the symbols h1, h2 and h3 look like.
They don’t imply anything about the pronunciation of
the sounds, beyond the fact that they are probably
laryngeal (as /h/ is).
That shows us that historical reconstruction is often
like an exercise in logic, like the work on phonemes
that we were doing earlier in the semester.
Historical reconstruction may not tell us anything
about what PIE actually sounded like.

44
 23-10-2016

Sound-Change that isn’t Regular

Analogy
The English ordinal number sixth, for example, goes
back to an ancestral form containing the cluster -kst-
(Latin cognate sextus), with a -t- that should not,
according to the regular conditioning of Grimm's Law,
have shifted to th after the fricative -s-.
But the -th of the present-day English word has nothing
to do with any failure of Grimm's Law. In fact, the Old
English form was siexta, with -t-; the -th of sixth was
introduced under the influence of the other ordinal
numbers, where -th was phonologically regular (fourth,
seventh, etc.).

strive / strove / striven > strive / strived / strived

old / elder / eldest > old / older / oldest
Note: elder remains with a specialized meaning; eldest is
used in archaic language.
nigh / near / next > nigh / nearer/ next >
nigh / nearer / nearest > near / nearer / nearest
Sturtevant’s Paradox: “sound change is regular and
causes irregularity; analogy is irregular and causes
regularity”
Brother – brother-en – brether-en (umlaut)

45
 23-10-2016

The paradigm of the Proto-Germanic verb *keusan ‘to

choose, to test’, became very irregular in Old English due to
the influence of three sound changes: palatalization, voicing
and rhotacism:

Proto-Germanic Old English

Present *kéusan ċēosan /tʃ/ - /z/
Past singular *káus ċēas /tʃ/ - /s/
Past plural *kus-ún- curon /k/ - /r/

Past participle *kus-án- gecoren /k/ - /r/

Old High German, being a North Germanic

language, also showed the same type of alternation
in the medial consonant:

Present kiusan [z]

Past singular ko:s [s]
Past plural kurun [r]
Past participle gikoran [r]

46
 23-10-2016

Levelling of paradigms

Present choose /tʃ/ /z/ küren

Past singular chose /tʃ/ /z/ kor

Past plural chose /tʃ/ /z/ kor

Past
chosen /tʃ/ /z/ ge-koren
participle

Semantic continuity
Old English sorh ‘grief’ > Mod. E. sorrow
Old English sārig ‘in pain’ (from sār, Mod E. sore)
sārig > sorig (not a regular sound change)
> Mod.E. sorry

47
 23-10-2016

Person marking in Latin verbs

Spanish dropped some of the

consonants
But maintained a differentiated paradigm

48
 23-10-2016

Gascon too
But maintained differentiated paradigm

But French?
4/6 forms identical!
Consequences for French syntax?

49
 23-10-2016

Only two identical forms

Morphological changes
Latin again, paradigm leveling

Nominative honos honos honor

Genitive honos-is honor-is honor-is

Accusative honos-em honor-em honor-em

Affected this paradigm but not genus or navis

50
 23-10-2016

Morphological changes
Backformation
Worker~work, burglar ~burgle, editor~edit, peddlar~peddle
Operate~operation, orientation~orientate (orient)
Cherries~cherry
Folk eymology
Garter snake ~ garden snake
Brydeguma (bride-man/suitor) - bridegroom
Chaise longue - chaise lounge
Nepotism - nephew-tism

Syntactic changes
English, do-insertion for question and negation
PIE had no articles (and neither do Latin or Sanskrit). Articles
came later in Greek (not in Homer)
Gender in Indo-Aryan

51
 23-10-2016

Jokes, only a linguist will tell

Giles of Tottenham calls for ale at his favourite pub and is
perplexed when the barmaid tells him that the fishmonger is
next door.

52