Historical Linguistics II

1
Historical Linguistics II
2
Counting numbers of the world, 3000 y.a.

• Here are the numbers 1 10 in
• Latin (modern Italy)
• Una, duo, tres, quattuor, qinque, seks, septem,
  octo, novem, decem
• Sanskrit (modern India)
• Sunya, eka, tri, catur, panca, sas, sapta, asta,
  nava, dasa
• Proto Celtic (modern UK)
• Oinos, dwosu, treis, kwetwars, kwenkwe, sueks,
  sektnz, okto, nauin, deknz
• Proto German (modern Germany)
• Ainaz, dwai, oriyiz, fidwors, fimfi, sehs, sibum,
  ahtos, niwun, tehan

3
Protolanguages

• In the late 1500s, missionaries to India began to
  notice similarities between Latin and Sanskrit.
• Later, European philologists noticed similarities
  between Latin, Sanskrit, Italian, Dutch, Persian,
  old Norse, Romanian an other languages
• In the 18th century William Jones suggested that
  there may be an Indo-European protolanguage
• In 1861, August Schliecher attempted to
  reconstruct proto-Indo-European
• In 1871, Schliecher formulated the Indo-European
  family tree

4
The family tree

• Schliecher hypothesized that sounds changed in
  regular and predictable ways over time
• The regularity hypothesis
• And that (therefore) languages that have
  systematic similarities at the phonological level
  may be related
• The relatedness hypothesis

5
Dead languages can be reconstructed

• You can reconstruct the older form of a language
  internally, by looking at the present forms and
  applying rules of sound change to work backward
• internal reconstruction
• You can reconstruct a completely dead
  proto-language comparatively, by looking at
  related words in the languages that descended
  from it and then applying rules of sound change
• comparative reconstruction

6
Internal reconstruction

• Where does the silent g come from?
• In English words like sign, paradigm, and
  deign
• If we compare a long list of words, we find that
  the g is always silent when it comes before a
  nasal vowel at the end of a word
• We also find that when it is before a nasal vowel
  but not at the end of a word, it is pronounced
• signal, dignity, etc.

7
Conditioned sound change

• This coherent distributional pattern is evidence
  of a systematic, conditioned sound change.
• We can postulate that, at one time, the g was
  pronounced in all of the words in which it is
  orthographically represented.
• And thereby reconstruct an earlier form of
  English
• (because English is a well documented language,
  these hypotheses can often be checked against
  historical data)

8
Comparative Reconstruction

• How do we reconstruct languages that are dead
  and gone?
• How do linguists know, for example, that the
  proto-Indo-European (PIE) word for to carry was
  bheroh?
• PIE was spoken more than 6,000 y.a.
• There was no writing then
• It ceased to exist about 5,500 y.a.

9
Comparative reconstruction

• Begin with a list of similar words from related
  langauges
• Sanskrit bharami
• Anc. Greek phero
• Latin fero
• Old Norse bero
• Old Irish biru
• Old Slavic bero

10

• Remember that sound changes are regular
• look for other correspondences
• Formulate laws of sound change
• E.g. Grimms law
• Proto-Indo-European voiceless stops change into
  voiceless fricatives.
• Proto-Indo-European voiced stops become voiceless
  stops.
• Proto-Indo-European voiced aspirated stops become
  voiced fricatives
• Work backwards, using these laws, to reconstruct
  what must have been the actual word

11
Comparative method procedure

1. Compile cognates (no borrowings)
2. Determine sound correspondences
3. Total? Then the original is obvious
4. Partial? Then the original is the product of know
   sound change rules.
5. NB when positing sound changes, employ Occams
   Razor (i.e., KISS rule)
6. Check for regularity

12
e.g., 1

• Step one
• Start with a cognate set

Lang A Lang B Lang C Gloss
siza sesa siza strawberry
13
2

• Step two
• Determine sound correspondences

Position Lang A Lang B Lang C
1 s s s
2 i e i
3 z s z
4 a a a
14

• Step three
• Reconstruct a sound for each position, using
  knowledge of sound change rules

Position Lang A Lang B Lang C rule Proto form
1 s s s none s
2 i e i egti Fronting e
3 z s z sgtz/ V_V Voice assim. s
4 a a a none a
15
Theories of language change

• Wave theory
• Language evolution

16
Wave theory

• Actually, as we saw in the study of dialects in
  sociolinguistics, languages are not discrete
  entities.
• Dialect continuum? Sprachbund? Language contact?
• The difference between German and Dutch, for
  example, is a matter of degree. There is no
  linguistic boundary that demarcates German
  speakers from Dutch speakers. One language
  gradually turns into the other as you move from
  Denmark to Germany
• The family tree theory cannot account for this
  gradualness
• Thus, the wave theory of language change
• Schmidt, 1872

17
Wave theory
Proto-Germanic
Dutch
German
18
Language evolution

• Do languages evolve? Like animals?
• Recent research supports the idea that languages
  should be thought of as being similar to computer
  viruses
• During the process of learning, languages
  colonize our minds.

19
Language evolution

• The three qualities that drive evolution are
• Variation
• All members of species vary from one another.
• This variation causes differential fitness
• Heredity
• Animals pass on their features through their DNA
• Differential reproduction
• More successful (i.e., more fit) individuals
  reproduce more than less fit individuals

20
Language evolution

• Languages also have these three features
• Variation
• At all levels of structure, idiolectal and
  dialectal
• Heredity
• Languages are passed down from generation to
  generation
• Differential reproduction
• Some languages are passed down to more people,
  while others are passed down to fewer people

21
Cultural selection for learnability

• Languages do not have to fight each other for
  berries and girlfriends (like we do), but it
  turns out that variants of languages that are
  easier to learn are passed down more often than
  variants that are more difficult to learn.

22
For English

• English in 1100 ACE
• Free word order
• Very complex verb and noun morphology
• OE verbs had 10 classes
• OE nouns had 5 declensions
• English in 2000 ACE
• Strict SVO word order
• Very simple verb and noun morphology
• ModE verbs have two classes
• ModE nouns have two declensions