Historical Morphology / Phonology Flashcards

1
Q

Elements of Language Change

A
  1. Lexical Change (words added / lost)
  2. Semantic Change (meaning)
  3. Orthographic Change (letters + letter shapes)
  4. Phonological Change (sounds)
  5. Morphological Change (grammatical marking, inflection)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Lord’s Prayer: Phon Change

A

fæder > father
* medial consonant has changed from OE to PDE
* change can be described as d > ð / V_Vr

N.B. [>] is used to describe diachronic change

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Lord’s Prayer: Morph Change

A

OE: heofonum
PDE: heavens

  • OE had morphological case: -um marked the PL.DAT case.
  • Development is not from -um to -s, but the entire collapse of the case system in English
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Case System Collapse

A

OE noun declension:
* 2 numbers (SG, PL)
* 4 cases (NOM, ACC, GEN, DAT)

PDE noun declension:
* 2 numbers (SG, PL)
* 2 cases (Base form, Possessive)
*GEN can be seen to be retained as the possessive

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Comparison and Correspondence

A

The instances of Eng /t/ seem to correspond with instances of GER /t͡s/

These correspondences are regular (applicable everywhere).

Change can be seen as t > t͡s / # _
Part of the OHG Consonant Shift

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Correspondence

A

If correspondence in comparative evidences are due to phonological changes, we expect them to be regular.

In fact, we expect them to be exceptionless

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Exceptionlessness Case Study: German

A

OHG > PDGerman
θ > d (No phon conditioning, all instances of θ changed)

  • /θ/ is still present in PDG from borrowed words from after the original phonological change occurred.
  • N.B. every occurrence of /θ/ that existed in OHG became /d/ in PDG. (the change was exceptionless)

Learning point: Exceptionless change should lead to exceptionless correspondences

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Notation: Diachronic and Synchronic Arrows

A

[>] represents phonological changes
[–>] represents phonological rules

Synchronic: Language states at one point in time
Diachronic: Language change over time

Combining both ideas:
* x –> y: phoneme x is realised as the allophone y
* x > y: what was x at TIME1 became y at TIME2

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Mechanisms of Language Change

A
  1. Innovation: the alteration in phon segments or structures (the structural aspect of change)
  2. Propagation: the way in which the alteration is taken up by speakers (the social aspect of change)

NB: No diachronic event in the history of languages can occur unless both occur.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Understanding Language Change

A

Related to:
1. Sociolinguistic Variation
2. Dialectological Variation

cf: Lang Var leads to Lang change
* Innovations may not always propagate through a speech community

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Dialectology

A

Isogloss: line marking dialect differences

Key mechanism behind accent / dialect diversity
* changes innovated in a particular area
* not propagated throughout the entire speech community

eg: uː-diphthongisation
* some areas of the UK still retain the monophthong qualities
* related to the GVS

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Comparative Approach

A

The comparative approach to historical linguistics hinges on a few factors:
* Language families, which allow a comparative reconstruction of the proto-language
* Family trees and cladistry
* Reconstruction of phonological and morphological structures from daughter languages

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Morph comparison and reconstruction

A

Case study: Nominal Case

In Gmc NPs, case is realised as inflectional suffixes on DET, N and A, etc.
* syncretism (one form multiple functions) was common

Daughter Languages:
* Eng (few cases)
* Ger (4 cases)
* Gothic (6 cases)
* Latin (6/7 cases)
* Sanskrit (8 cases)

Evidence suggests that Sanskrit retains all IE cases, while in other daughter languages there have been varying degrees of case merger

IE languages with extensive case merger become more reliant on expressing precise meanings via word order.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Phon comparison and reconstruction

A

PIE obstruent phonemic inventory:
* stops: (-) p, t, k (+) b, d, g
* affricates: (-) t͡ʃ (+) -dʒ
* fricatives: (-) f, θ, s, ʃ, h (+) v, ð, z, ʒ

The segments display high levels of systematicity
* Two series of stops and affricates at all places of articulation
* two series of fricatives at allmost all places of articulation
* Voiceless vs Voiced distinction

Phonologists expect segments to form systems like this in languages, but it is not always the case.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Germanic Stop systems

A

OE:
(-): /p/ /t/ /k/
(+): /b/ /d/ /g/

NGmc:
(-): /p/ /t/ /k/
(+): /b/ /d/ /g/

EGmc: Additional Place of Articulation (Labiovelar)
(-): /p/ /t/ /k/ [/kʷ/ (labialised)]
(+): /b/ /d/ /g/

PGmc:
(-): /p/ /t/ /k/ [/kʷ/ ]
(+): /b/ /d/ /g/ [
/gʷ/
]

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

IE stop systems

A

AGk:
(-voice)(-aspiration): /p/ /t/ /k/
(-voice)(+aspiration): /pʰ/ tʰ/ /kʰ/
(+voice): /b/ /d/ /g/

Sanskrit:
(-voice)(-aspiration): /p/ /t/ /k/
(-voice)(+aspiration): /pʰ/ tʰ/ /kʰ/
(+voice)(-aspiration): /b/ /d/ /g/
(+voice)(+aspiration): /bʰ/ /dʰ/ /gʰ/

NB: Sanskrit -V+A series is an innovation that combines PIE stops with neighbouring laryngeal segments. Only the first 3 are inherited from PIE.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Reconstructing PIE stop system

A

Correspondences between AGk and Sanskrit:
1. Both have (-v)(-a)
2. Both have (+v)(-a)
3. AGk has (-v)(+a). Sanskrit has (+v)(+a)

It would be an impossible coincidence for AGk and Sk to have these regular correspondences unless they were related by regular phon change.

In most IE languages, the descendants of the third series of stops are voiced. Hence can conclude that AGk underwent devoicing in that series.

Reconstructed PIE stop system:
(-voice)(-aspiration): /p/ /t/ /k/
(+voice)(-aspiration): /b/ /d/ /g/
(+voice)(+aspiration): /bʰ/ /dʰ/ /gʰ/

18
Q

PIE > PGmc Stop system

A

Recall: PIE stop system
(-voice)(-aspiration): /p/ /t/ /k/
(+voice)(-aspiration): /b/ /d/ /g/
(+voice)(+aspiration): /bʰ/ /dʰ/ /gʰ/

Recall: PGmc Stop system
(-voice)(-aspiration): /p/ /t/ /k/
(+voice)(-aspiration): /b/ /d/ /g/

Hypotheses:
1. Loss of aspiration in the third series (merger)
2. Devoicing of second series (merger)
3. What happens to the voiceless series then?

19
Q

Grimm’s Laws (Germanic Consonant Shift)

A
  1. Spirantisation
  2. Devoicing
  3. Loss of Aspiration
20
Q

Grimm’s Law 1: Spirantisation

A

Voiceless stops become voiceless fricatives:
p,t,k > f,θ,x

21
Q

Grimm’s Law 2: Devoicing

A

Voiced stops become voiceless stops:
b,d,g > p,t,k

22
Q

Grimm’s Law 3: Loss of aspiration

A

Voiced aspirated stops become voiced unaspirated stops:
bʰ,dʰ,gʰ > b,d,g

23
Q

Relative chronology of the GL changes

A

If GL3 occurs before GL2, the outputs of GL3 would become inputs for GL2.

This would result in the expected correspondences of Sankrit and OE being:
bʰ - p
dʰ - t
gʰ - k
But this is not the case, hence the changes must have happened in the other order (GL2 prior to GL3).
Same reasoning holds for establishing the relative chronology of GL2 and GL1.

24
Q

Chain Shift

A

Two types: (The names are quite clear)
1. Drag chain
2. Push chain

eg: GL as drag chain (one change being the phonological motivation for another)
1. GL1 happens, leaving the -v stop series ‘empty’
2. GL2 happens to fill the -v stop series. +v stop series now empty.
3. GL3 happens to fill the +v stop series. +v+a stop series now lost.

25
GL1 and exceptionless hypothesis
**Exceptionless / Regularity Hypothesis**: Phonological changes occur without lexical exemption. Problematic correspondences in Latin (retains PIE stops) and OE (underwent GL). /t/ - /t/ (no change) /t/ - /d/ (expected θ) These are **not random failure of GL1** (not exceptions) Moving forth: * may not have properly described the change involved * other changes may have intervened
26
GL1 Observations
When the segment is preceded by a stop, the segment does not change as per GL1. This suggests that GL1 was a conditioned change: p,t,k > f,θ,x / ¬C__ (¬C meaning 'not an obstruent') This seems to account for the first "exception": /t/ - /t/ (no change)
27
Verner's Law
Recall: If a word is retained in two daughter languages, then we expect the segments to be inherited as the same in both languages, unless there has been change. Second hypothesis for /t/ - /d/: Some other changes must have intervened. VL: [obstruent -voice] > [obstruent +voice] /σ(unstressed) _ VL fundamentally affect medial segments, but not if the **preceding syllable was stressed in PIE** Discovery of VL led to the development that **exceptionless hypothesis is not false**
28
Rule-related phon changes
1. Rule addition 2. Rule change 3. Simple rule loss 4. Rule loss with lexicalisation (=restructuring)
29
Rule addition
Case study: Rhoticity in EngEng Original state -> r-ful Rule added: r --> Ø/ __(rhyme) Result: r-lessness When the rule of r-deletion was first added to Eng, it was **variable** * rules can remain variable indefinitely * over time rules stop being variable and become obligatory Even with the rule of r-deletion, the linking r can still arise in some phonological conditions.
30
Rule change
Case study: l-vocalisation Original rule: /l/ --> [ɫ] / __ (rhyme) Rule change: /l/ -->[ɫ] / __ (rhyme) > /l/ --> [w] / __ (rhyme) Output of the rule has been changed (loss of occlusion)
31
Rule Loss (simple)
Rule disappears, removing all its effects. eg: NYC r-fulness If the increase in r-ful forms continues, it could become categorical (i.e., only r-ful forms are used) which would mean that the rule would be lost
32
Rule loss with lexicalisation (Restucturing)
A rule can disappear from a language but the **derived allophones it created remain** * involves a change at the phonemic level * segments that were allophones "become" phonemes Case study: French Nasal Vowels Stage 1: Nasalisation Rule V --> ṽ / __ N Stage 2: Rule Addition of Nasal-Deletion rule V --> ṽ / __ N N --> Ø / __ # Stage 3: Restructuring * Stage 2 rule is **variable** * Learners assume there is no nasal in the phonemic form * Both rules are thus lost + **phonemic form changes** * **Restructuring** of phonemic form Something that is predictable (allophonic) has now become unpredictable (phonemic)
33
Case Study: GL1
Stage 1: PIE /pods/ --> [pods] 'foot' Stage 2: Rule addition of spirantisation p, t, k --> f, θ, x / ¬C __ (variable: learners know there is a stop in the phonemic form) /pods/ --> [fods] 'foot' Stage 3: Later pre-PGmc GL1 stops being variable, learners assume there is no stop in the phonemic form * **lexicalisation + restructuring**: rule is lost and the phonemic form changes /fods/ --> [fods] 'foot'
34
Emergence of morphology via phonological changes
cf: Rule loss with lexicalisation > What happens if the lexicalisation happenes differently to different members of a morphological paradigm? Leads to **morphologisation** --> emergence of **irregular morphology**
35
Irregular Plurals
Most PDE plurals are created by the affixing an <-s> or some variety. In irregular plurals, plurality is indicated by the use of a different vowel to the singular. * Plurality used to be expressed by suffixation in OE. * Phonological rule was innovated: i-umlaut a --> e/ __N i,j This environment was not common in nouns (eg mann had it, but only in the plural). * Another rule innovated: apocope (deleted /i/ and /u/ following heavy syllables) i,u --> Ø / σ[H] __# apocope stops being variable, learners assume vowel alternation is the phonemic form. **Lexicalisation + Restructuring** leads to the rise of irregular morphology
36
Morphologisation at rule death
Morphologisation (structures that arose in other domains of linguistics enacting morphological changes) * the rule only applied to the plural morphology * when the rule disappears, vowel change becomes the morphological marker of plurality * this leaves a morpheme with more than one phonemic / underlying form *PDE also shows vowel alternation in verb paradigms
37
Analogical Levelling
cf: morphemes with >1 underlying form is difficult for learners. Analogical levelling simplifies the paradigm to remove the extra stored information. eg: help used to be an irregular verb but has now been levelled. * INF, PST and PST-PRT are all (was assumed by learners to be the same vowel in all forms) * began using regular -ed tense inflection as well
38
Analogical Levelling 2
Analogy typically functions to **remove morphological irregularity** cf: phonological changes cann create morphological irregularity Levelling is a kind of morphological change. > Unlike phonological change, morph change is **not expected to be exceptionless** > analogy can be **unpredictable** (eg: consonant in INF 'freeze' levelled to sibilant in OE but levelled to the rhotic in OHG)
39
Reduction and Loss of Inflection
Reduction > tends to affect segments at the end of words > change is quite common in the 'final position' > OE stress patterns led to final syllables being unstressed (change is common in unstressed positions) NB: Even if change is common in any given environment, that does not mean that change is guaranteed to happen Changes that count as reduction: 1. weak positions (final + unstressed) 2. involve some kind of loss (loss of difference between segments, loss of segments and even entire loss of syllables) Loss of difference between segments can occur in some forms of **mergers** >in weak positions segments can merge to become unmarked segments > schwa claimed to be unmarked vowel, coronal claimed to be unmarked place of articulation Examples of reduction in weak positions: 1. vowels > schwa 2. loss of final vowels 3. labials / velars > coronals 4. loss of final consonants
40
Morphological Erosion in history of English
eg: collapse of case system > led to erosion of inflection > inflections were mostly vowel differences and consonants in **final unstressed syllables** eg: collapse of adjectival inflection > presence of syncretism (conducive for merger) > many of the distinctions that are expressed clearly are carried by a final vowel > these final syllables were unstressed
41
Phonological reduction in history of English
Stage 1: ME unstressed codas m>n * loss of difference due to a **merger to the unmarked coronal /n/** Stage 2: ME unstressed syllables i,e,u,o,a>ə * merger to unmarked vowel * loss of differences --> fewer vowels can occur in unstressed syllables Stage 3: ME unstressed final syllables ə>Ø * Segmental loss
42
Morphological effects of phonological reduction
Phonological reduction led to mergers in the forms of the case inflections. Effected **erosion** of the inflections