Linguistics A Phonology 1 (2008) - asakoh.netasakoh.net/martin-ling-A/pdf/slides_phonology_part1.pdf · Phones, Phonemes and Allophones Defining a set of phonological features, along

Linguistics A

Week 4Phonology (Sound Patterns of Language)Part 1

Phonology (vs. Phonetics)

PhoneticsHow sounds are articulated in the vocal tract;Acoustic/auditory properties of linguistic sounds;

PhonologyDiscrete representation of linguistics sounds in metal grammar;Rules governing the possible arrangements (patterns) of sound segments, how individual sounds are structured into syllables, words, etc.;

Phonological FeaturesWe expect the characterization of sounds in mental grammar (discrete mental representations of sounds) to be substantive in nature. We will see that discrete sound segments are made up of more basic discrete elements called phonological features. Phonological features are substantive insofar as they based on substantive phonetic (acoustic, articulatory, etc.) properties of linguistic sounds.

A sound segment, then, is a set of discrete phonological features. Knowing a sound segment means knowing which features this set contains. This can be most easily captured in terms of a set of " valued features; for example {"F,…}, where +F indicates the presence of a particular feature F and !F indicates the absence of F.

The Major Classes of Linguistic Sounds

Obstruent sounds oral stopsfricatives obstruent consonantsaffricates

Sonorant soundsNasal stopsLiquids sonorant consonantsGlidesVowels vowels

Feature Specification of Major Classes

O N L G Vconsonantal + + + ! !

sonorant ! + + + +syllabic ! !/+ !/+ ! +nasal ! + ! ! !/+

(O = obstruents, N = nasals, L = liquids, G = glide, V = vowels)

Consonant Features

"consonantal"sonorant"syllabic"nasal"voiced"continuant"labial"alveolar"palatal"anterior"velar"coronal"sibilant

A close-up look: Feature Specifications of Labials

p b mconsonantal + + +sonorant ! ! +syllabic ! ! ! or +nasal ! ! +voiced ! + +continuant ! ! !labial + + +alveolar ! ! !palatal ! ! !anterior + + +velar ! ! !coronal ! ! !sibilant ! ! !

Vowel Features

!consonantal+sonorant+syllabic"nasal"voiced"high"low"front"back"round"tense

Feature Specifications of English Vowels

high + + ! ! ! + + ! ! ! !low ! ! ! ! + ! ! ! ! + !front + + + + + ! ! ! ! ! !back ! ! ! ! ! + + + + + !round ! ! ! ! ! + + + + ! !tense + ! + ! ! + ! + ! ! !

[i] [w] [e] [e] [F] [u] [U] [o] []] [a] []

Phones, Phonemes and Allophones Defining a set of phonological features, along with their possible arrangements and values, allows us to specify the set of all possible sounds occurring in human languages, called phones.

The set of phones occurring in a particular language (say, English or Japanese) thus constitutes a subset of the set of all possible phones, almost invariably a proper subset (though clearly some languages make use of a much wider range of sounds than others).

Phonemes are basic abstract sound units that occur in a particular language and are used systematically to distinguish one word from another in that language.

To distinguish the notion phone from the more abstract notion phoneme we will put the former inside of square brackets [ ] and the latter inside of slanted brackets / /. For example, /s/ will be used to indicate a phoneme whereas we will use the notation [s], when referring to the phone.

In general the relation between phones and phonemes is one-to-one. However, in some cases a single phoneme in a language may correspond to more than one phone; in other words, the phonetic realization of the phoneme differs depending on the surrounding phonological context. When two or more different phones correspond to a single phoneme, we call them allophones of the phoneme.

For example, as we will see, the two phones [p] and [ph], as witnessed in words like [pwl] and [sph

wl] respectively, are allophones of a single phoneme /p/.

Identifying Phonemes and Allophones

phonemic level: /P1/ /P2/

allophonic level: [p1] [p2] [p3]

(1) Minimal pairs(2) Phonological similarity(3) Distribution (contextual predictability of variation)

Minimal Pairs

When two different, distinct sounding words (or “possible words”) are otherwise identical except for one phone occurring in exactly the same “position” (phonological context) in each word, the two words are said to form a minimal pair. In such cases, these two distinct phones, by which the two words in question differ minimally, are very likely to constitute distinct phonemes in the language.

[ __ wp ]

[thwp] tip [lwp] lip[dwp] dip [rwp] rip[nwp] nip [±wp] chip[swp] sip [®wp] jip[zwp] zip [wwp] whip[•wp] ship [hwp] hip

[b __ t]

[bit] beet [but] boot

[bwt] bit [bot] boat

[bet] bait [b]t] bought

[bet] bet [bt] but

[bFt] bat [bat] Baht

Phonological Similarity

[phwt] ~ [kh

wt] for a minimal pair differring only with respect to the sounds [ph] ~ [kh]. Similarly, the minimal pair [spwt] ~ [skwt] differ only with respect to the sounds [p] ~ [k]. But we should not jump to assume that there are four distinct phonemes /ph/, /kh/, /p/, /k/ based on these facts alone.

Given the phonological similarity of the pair [p] ~ [ph] and of [k] ~ [kh], we must consider the possibility that these pairs constitute allophones of a single phoneme. We should first check to see if there are any minimal pairs distinguished only by whether a voiceless stop is aspirated or not; if there are no such minimal pairs, we should check the distribution of these sounds to see of they alternate in predictable ways depending on the phonological context.

Complementary Distribution

Nasal vowels Oral vowels

bean [bi#n] bead [bid]bee [bi]

roam [ro#m] robe [rob]row [ro]

pain [pe#n] paid [ped]pay [pe]

Deriving Surface Phones (Allophones) from Underlying Phonemes

Suppose that words have two phonological representations, an Underlying Representation (UR), which contains only phonemic information, and a Surface Representation (SR), which contains all information necessary for actual phonetic articulation of distinct phones.

The UR of a word consists only of sounds that are phonemes in the language. UR is then mapped to SR, possibly involving the application of phonological rules. Thus, the SR of a word may differ systematically from the UR and may include phones that are not even phonemes in the language.

Mapping UR to SR

UR phonemic

(Phonological Rules)

SR (allo)phonic

Context-Sensitive Phonological Rules

X Y (W) (Z)

rule context

input output target = an occurrence of X

a b / ___ eaf ebf

a b / c ___ eaf eafcaf cbf

a b / c ___ d eaf eafcaf cafcad cbd

Nasal vs. Oral Vowels

Nasal and oral vowels are in complementary distribution.Nasal vowels occur before a nasal consonant; oral vowels occur elsewhere. The phonological context in which nasal vowels appear is narrower (more specific) than the context in which oral vowels appear. Strategy: take the allophone that occurs in the wider (more general) context to correspond to the underlying phoneme, and derive all other allophones by way of context-sensitive phonological rules.

Nasalization Rule

⋮ ⋮

!consonantal ⋮ +consonantal+syllabic +nasal !syllabic!nasal ⋮ +nasal

⋮ ⋮

Nasalization Rule (simplified notation)

V V: / __ C[+nasal]

Derivations of the SRs of the words bead and bean:

UR: /bid/ /bin/↓ ↓

Nasalization Rule NA bi:n↓ ↓

SR: [bid] [bi:n]