Phonological Processes - Introduction to Linguistics - Lecture Slides, Slides of Linguistics

Phonological Processes, Speech Communication, Perception of Voiced Sounds, Phonological Features, Patterning of Sounds, Phonotactics, Model these Patterns, Alternations and Underlying Forms represents lecture layout.

Typology: Slides

2011/2012

Uploaded on 11/21/2012

ashakiran
ashakiran šŸ‡®šŸ‡³

4.5

(27)

261 documents

1 / 43

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
LING 40030 - General Linguistics
•Week 12: Phonological Processes: How can
we describe the patterns found in speech?
o Phonological Features
o Patterning of Sounds: Phonotactics
o Alternations and Underlying Forms
o Ordering of rules/constraints
o Phonetics and Phonology Case Study
Docsity.com
pf3
pf4
pf5
pf8
pf9
pfa
pfd
pfe
pff
pf12
pf13
pf14
pf15
pf16
pf17
pf18
pf19
pf1a
pf1b
pf1c
pf1d
pf1e
pf1f
pf20
pf21
pf22
pf23
pf24
pf25
pf26
pf27
pf28
pf29
pf2a
pf2b

Partial preview of the text

Download Phonological Processes - Introduction to Linguistics - Lecture Slides and more Slides Linguistics in PDF only on Docsity!

LING 40030 - General Linguistics •^

Week 12: Phonological Processes: How canwe describe the patterns found in speech?^ o Phonological Featureso Patterning of Sounds: Phonotacticso Alternations and Underlying Formso Ordering of rules/constraintso Phonetics and Phonology Case Study

Recap on Speech Perception

•^

Perception involves: Detection of acoustic cues associated witharticulatory configurationsKnowing a lot about acoustics!Deducing of the speaker’s intended message,with reference to acoustic cues.

Speech Perception

•^

What are the acoustic cues for the voicingcontrast in English Aspiration: word (and syllable) initially voicelessstops are accompanied by aspiration.Others? Lisker (1986) lists sixteen, and admitsthere could be more!What cues are there to signal voiced sounds?

Perception of Voiced Sounds

  • Listen to this phrase: what is it?• Yes, you’ve uttered it frequently yourself too:

ā€œSay cab twiceā€

  • Now listen again, this time I’ve removed half of

the vowel (about 80 ms) prior to the /b/

  • Sounds like ā€œSay cap twiceā€ā€¢ A phonetic correlate of voiced sounds is longer

preceding vowel length

Detail Voice : VOT

  • Lisker & Abrambson (1964) proposed Voice Onset Time

Phonological Features

  • A possible solution to the invariance problem in

speech perception.

  • [ - voice ] = { high + VOT value, …}• [ + voice ] = { zero or less VOT, increased

preceding vowel duration, …}

  • Phonological/Distinctive features are an

abstraction…calling them phonetic features is amisnomer

Phonological Features

  • Using phonological features, we can easily

capture Natural Classes

  • [ +voiced ] = {

b, d, d, , v, z,

, l, m, n,

d

  • [ - voiced ] = {

p, t, t, k, f, s, , h, t

  • [ -continuant ] = stops and nasals• [ +continuant ] = everything else!• [ -consonantal, +sonorant, +high, -back] = ???• high front vowels = { i, e }

Phonological Features

  • You may have noticed that features are binary

i.e. have a + or - value

  • This is true of the majority of features• Some, however, are unary (have a single value)• [ LABIAL ] = bilabial + labio-dentals• [ CORONAL ] = sounds made with the tip,

blade or front of tongue. Dentals/Alveolars…

  • [ DORSAL] = sound made using tongue back

velars/uvulars

  • [ GLOTTAL], [RADICAL/GUTTURAL]

Patterning of Sounds

  • Phonotactic restrictions/constraints apply to

many kinds of sequences and in differentdomains

  • English allows sequences up to CCC at the

beginning of a word, and CCCC at the end of aword e.g strengths [

st kts

].

  • Our focus:

nasal-consonant sequences

from a

range of languages.

Ndali (a Bantu language)

The prefix in these data consists of a vowel and anasal consonant (/iN+/. Stem

Prefixed form

p

uno

imb

uno

nose

tunje

ind

unje

banana

k

unda

i^

unda

dove

ale

imb

ale

plate

fuwa

if

uwa

hippopotamus

satu

is

atu

python

*[



] = a voiced bilabial fricative – check out the IPA chart.

We use a

two-level

model, with a derivational approach.

Input

= /prefix + stem/

Output = the actual occurring forms (phoneticrepresentation)

-^

What links the two levels are the strategies for dealingwith e.g. disallowed sequences:either ā€˜repair’ strategies

  • i.e. make the offending sequences

a ā€˜better’ sequence or deletion

  • i.e. delete one of the consonant in the

problematic sequence

How do we model these patterns?

I take

wata

kwata

I send

numa

tuma

I speak

mopja

popja

I cook

seva

seva

I dig

fela

fela

I look

mbanja

vanja

Gloss

Prefixed form

Umbundu (a Bantu language)Stem

Alternations and Underlying Forms

  • Step 1: Identify alternations (if any) in a given

dataset showing morphological paradigm

  • Step 2a: If no alternations, then we can propose

underlying forms. THE END.

  • Step 2b: Determine candidate underlying forms• Step 3: Evaluate candidates, based on

ā€œcomplexityā€, and then propose the underlyingform.

Underlying Forms

-^

There are no alternations between stem (nominativeform) and stem+affix

-^

Therefore, lexical forms are /

ko

/, /

lu

/, /

t'ak

/,

/

t

/ and /

stu(n

/

ā€˜my boat’

[^

stu(n

an

]

ā€˜boat’

[^

stu(n

]

ā€˜my house’

[^

t

an

]

ā€˜house

[^

t

]

ā€˜my table’

[^

t'ak

an

]

ā€˜table’

[^

t'ak

]

ā€˜my coat’

[^

lu

an

]

ā€˜coat’

[^

lu

]

ā€˜my head’

[^

ko

an

]

ā€˜head’

[ ko

^

]

possessive

nominative

  • Consider this dataset: