LI 752 - Seminar in Generative LinguisticsW2: Introduction to Phonetics

1 Introduction

Last Week: A language consists of

(i) A structured collection of sounds Phonetic Inventory

(ii) A repository of meaning Semantic Ontology

(iii) Rules about how these elements combine (i.e., the ”GRAMMAR”)

• sounds into complex sounds phonology

• sound and meanings into 〈sound,meaning〉 pairs lexicon

• morphemes into words morphology

• words into phrases and sentences syntax

• simple morpheme meanings into complex meanings semantics

• complex meanings with context pragmatics

Phonetics is the study of speech sounds

(i) How do you make these sounds?

(ii) What properties do these sounds have?

(iii) etc.

Q: What are the properties of speech sounds?

2 Basic Phonetics

• There are three main subfields of phonetics:

1 articulatory phonetics: how speech sounds are made. (eg., articulator placement, vocal fold configuration, etc.)

2 acoustic phonetics: the physical properties of speech sounds. (i.e., frequency, wavelength, etc.)

Time (s)5.613 6.5480

4000

Frequency(H

z)

5.73319396 6.35870804

. Spectrogram of the word “food” [fud]

3 perceptual phonetics: how speech sounds are perceived. (eg., pitch vs frequency, the McGurk Effect)

• The international phonetic alphabet (IPA) is a transcription system thatdescribes speech sounds according to their articulatory phonetic features

. → So we’ll focus on articulatory phonetics

• IPA symbols (see IPA chart)

1

. - represent a single sound segment (with multiple features)

. - are usually enclosed in square brackets, eg., [m]

3 Articulatory Phonetics: Consonants

• Speech sounds are made by pushing air through your vocal tract

They can be categorized according to their

1. Air source: lungs (pulmonic)1

2. Air flow direction: out (egressive), in (ingressive)

. These factors define a sound’s airstream mechanism

For now, we’ll only consider one airstream mechanism. - pulmonic egressive sounds

• Speech sounds can also be categorized according to whether they involve aconstriction in the vocal tract

– vowels: mostly unconstricted vocal tract - air flows freely

– consonants: constricted vocal tract - air is obstructed

• Further categories of constriction(s)

1. Place of Articulation (PoA). i.e., which articulators (eg., tongue, lips, etc.) are used. (see vocal tract diagram on the right)

2. Manner of Articulation (MoA). i.e., type of constriction (eg., full closure, near closure, etc.)

3. Laryngeal Properties. i.e., the configuration of the glottis and vocal folds/chords

1The other types of air sources are glottalic and velaric - we probably won’t discuss these.

1. Glottis

2. Vocal Folds/Chords

3. Lower Teeth

4. Upper Teeth

5. Lips

6. Nasal Cavity

7. Alveolar Ridge

8. Hard Palate

9. Soft Palate/Velum

10. Uvula

11. Epiglottis

12. Larynx

2

3.1 Place of Articulation

• Speech sounds can be categorized according to Place of Articulation. - i.e., according to the ‘active articulator’ (the articulator that moves)

(1) LABIAL: using the lips [p, b, m, f, v]

(2) CORONAL: using the tip or blade of the tongue [t, d, s, S, Z]

(3) DORSAL: using the back of the tongue [k, g, q, Q]

• place of articulation can further be categorized according to the. ’passive articulator’ (where the active articulator moves to)

LABIAL: CORONAL: DORSAL:Bilabial [p, b, m] Dental [T, D] Palatal [j]Labiodental [f, v] Alveolar [t, d, n, s] Velar [k, g, N]

Retroflex [ú, ã ] Uvular [q, ö, X, K]Palato-Alveolar [S, Z]

PoA Examples from English

Bilabial: pit [phIt], tap [thæp], moon [mun], same [seIm]

Labiodental: fan [fæn], phone [foUn], very [vE.ôi], save [seIv]

Dental: this [DIs], weather [wE.D@ô], think [TINk], path [phæT]

Alveolar: rat [ôæt], sit [sIt], do [du], no [noU], lie [laI]

Palatal-Alveolar:

splash [splæS], shop [SAp], rouge [ôuZ], genre [ZAn.ô@]

Palatal: yawn [jAn], yell [jEl]

Velar: pick [phIk], pig [phIg], sing [sIN]

Glottal: uh-oh [2.PoU], happy [hæ.phi]

Exercises (Use your IPA chart!)

1. What place of articulation do the first segments in these words have? pie,buy, my, finish, value, tie, die, night, sigh, zip, laugh, write, wrong, rain, shy, she,show, you, yellow, cat, gain, height

2. What place of articulation do the final segments in these words have? pack,pick, hang, sing, wrong, rough, fish, rat, car, tall, give, drive, meat, pass

3.2 Manner of Articulation

• Speech sounds can be categorized according to Manner of Articulation. - i.e., the degree of closure between the articulators

1. STOP: complete closure of articulators– oral (nasal cavity blocked by velum)2 [p, t, k]– nasal (nasal cavity open) [m, n, N]

Diagram from Ladefoged 2001

2. FRICATIVE: close approximation of articulators. (turbulent airstream) [s, z, T, D, f, v, S, h]

Diagram from Ladefoged 2001

3. APPROXIMANT: close approximation of articulators. (non-turbulent airstream) [w, j, l, r]

4. AFFRICATE: stop with fricative release [>tS,

>dZ]

Exercises (Use your IPA chart!)

1. What manner of articulation do the first segments in these words have?pie, buy, my, finish, value, tie, die, night, sigh, zip, laugh, write, wrong, rain, shy,she, show, you, yellow, cat, gain, height

2. What manner of articulation do the final segments in these words have?pack, pay, hang, fare, wrong, rough, fish, rat, car, tall, give, drive, meat, pass

2Another name for an oral stop is a plosive.

3

3.3 Laryngeal Mechanism

• Egressive pulmonic consonants can be categorized according to

(i) VOICING - i.e., whether the vocal folds are vibrating

– VOICED (closed, vibrating vocal folds){b, d, g, z, v, m, n, a, i , u, e, o, dZ, D, , ...}

– VOICELESS (open vocal folds){p, t, k, s, S, f, tS, T, ...}

(ii) ASPIRATION- i.e., whether the sound is followed by a puff of air

– ASPIRATED (puff of air){ph, th, kh

}

– UNASPIRATED (no puff of air){p, t, k}

CONSONANT ARTICULATORY FEATURES: Summary

Consonants can be categorized according to the following features:

1. Place of Articulation (bilabial, alveolar, velar, etc.,)

2. Manner of Articulation (stop, nasal, fricative, etc.,)

3. Laryngeal Properties (voiceless, aspirated, etc.,)

4 Articulatory Phonetics: Vowels

• Recall, speech sounds are made by pushing air through your vocal tract. - and vowels involve a lack of construction and free-flowing air

• ...but your tongue and lips are placed in different locations:

Diagram from Ladefoged 2001

1. Vowel [i] as in heed2. Vowel [I] as in hid3. Vowel [E] as in head4. Vowel [æ] as in had

5. Vowel [A] as in father

6. Vowel [U] as in good

7. Vowel [u] as in food

• The different tongue/lip configurations define different cavities of air withinyour vocal tract

• These cavities have different resonant/harmonic frequencies, which is whydifferent vowels sound distinct

4

4.1 Vowel Features

VOWELS are characterized in terms of the following articulatory features:. height, front/backness, rounding, tense/laxness3

We’ll use this minimal vowel chart to go through the features. (see IPA chart for full list of symbols)

(i) HEIGHT (i.e., high [i] vs low [a] ). relates to the vertical position of the tongue

3Some languages make use of a feature Advanced Tongue Root (±ATR)

(ii) FRONT/BACKNESS (i.e., front [y] vs back [u]). relates to the front/back position of the tongue

(iii) ROUNDNESS (i.e., round [y] vs unround [i]). relates to roundness of lips

(iv) LAXNESS (i.e., tense [i] vs lax [iI). relates to the degree of height/backness

NOTE 1: Some dialects of English do not have the vowel [O]. These dialects,a

which have the caught-cot merger pronounce the word pairs odd/awed, col-lar/caller, wok/walk with [A]. Dialectsb which lack the caught-cot merger pro-nounce the first member of the pair with [A] and the second member of thepair with [O].

aeg., my dialectbeg., Lady Mary’s RP dialect from Downton Abbey

NOTE 2: Some dialects of English do not have the vowel [O]. These dialects,a

which have the caught-cot merger pronounce the word pairs odd/awed, col-lar/caller, wok/walk with [A]. Dialectsb which lack the caught-cot merger pro-nounce the first member of the pair with [A] and the second member of thepair with [O].

aeg., my dialectbeg., Lady Mary from Downton Abbey

5

IPA Symbol Example

[i] English: sheep [Sip], eat [it], free [fôi]

[y]French: tu [ty], lu [ly], bu [by], German: mude[mud@], uber [yb5], Cantonese: ”book” [Sy]1,”pig” [

>tsy]1, ”fish” [jy]6

[I] English: lip [lIp], hit [hIt], sick [sIk]

[W]

[u] English: root [ôut], hoop [hup], soon [sun]

[U] English: could [khUd], look [lUk], full [fUl]

[e]French: aimer [e.me], English: hey [hej]/[heI], lay[lej]/[leI], page [phej

>dZ]/[pheI

>dZ]

[o] French: peau [pho], English: low [low]/[loU], boat[bowt]/[boUt], soak [sowk]/[soUk]

[@] English: about [@.bawt], sofa [sow.f@], amazing[@.mej.zIN]

[E] English: let [lEt], wreck [ôEk], hen [hEn]

[2] English: bus [b2s], cut [k2t], some [s2m]

[O] English: awed [O:d], hawk [hO:k], bought [bO:t]

[æ] English: hat [hæt], can [khæn], map [mæp]

[a] Japanese: ”fish” [sa.kha.na]

[A] English: hot [hAt], hop [hAp], sock [sAk]

[aj]/[aI] English: pie [phaj], ride [ôajd], time [thajm]

aw]/[aU] English: cow [khAw], loud [lawd], clown [khlawn]

[Oj]/[OI] English: toy [thOj], droid [>dZôOjd], coin [khOjn]

VOWEL ARTICULATORY FEATURES: Summary

Vowels can be categorized according to the following features:

1. Height (high, mid, low)

2. Front/Backness (front, central, back)

3. Roundness (round, unround)

4. Laxness (tense, lax)

Study Guide

You should know these terms, and what they refer to:

1. Place of Articulation, Manner of Articulation, Laryngeal Properties

2. Labial, Coronal, Dorsal (PoA acc. to active articulator)

3. Bilabial, Labiodental, Dental, Alveolar, Palatal-Alveolar, Palatal, Velar,Glottal

4. oral stop, nasal stop, fricative, approximant, affricate

5. voicing, aspiration

6. height, front/backness, roundness, laxness

You should be able to:

1. Label a diagram of the articulators in the vocal tract

2. Categorize an IPA symbol according to. (i) PoA, MoA, and laryngeal mechanism, or. (ii) height, front/backness, roundness, tense/laxness

6

5 Acoustic Phonetics

• Phonetics: The study of speech sounds

– Articulatory - i.e., how speech sound are made

– Acoustic - i.e., their physical properties (waveform analysis, frequency,wavelength, etc. )

– Perceptual - i.e., how speech sounds are perceived (eg., pitch vs fre-quency)

• Speech sounds are sounds; this means they have all the characteristic mea-surable properties of sounds, eg.

(1) Wavelength

(2) Frequency/Period

(3) Amplitude

→ Acoustic phonetics is the study of these properties.

• A Waveform shows the changing amplitude of a sound over time

Time (s)5.687 6.439-1

0.932

0

6.0633676

• This allows you to easily see changes in loudness over time,but it’s difficult to see changes in frequency over time.

• And frequency is much more useful for describing speech sounds.

• Some ways frequency is useful for describing speech sounds:

1. Vowel quality can be characterized by resonant frequencies (Vowel qual-ity ≈ the difference between i, e, o, u, etc.)

2. Pitch/tone correlates with fundamental frequency

3. A consonant’s place of articulation can be characterized by their effecton the resonant frequencies of neighbouring vowels

→ These (vowel quality, tone, place of articulation) are all things that lan-guages use to distinguish/contrast meaning

• spectrograms are much more useful for investigating speech sounds:

• A Spectrogram of the word [fud] ‘‘food’’

Time (s)5.613 6.5480

4000

Frequency(H

z)

5.73319396 6.35870804

• What is a Spectrogram?

° A spectrogram is a compilation of spectrums or spectral slices

° Where a spectrum/spectral slice shows, for a particular point in time

(1) Frequency (Hz):Shown on the x-axis (→)

(2) Amplitude (dB):Shown on the y-axis (↑)

7

• How to Read a Spectrogram

(1) Vowel Pitch and Consonant VoicingThe dark band at the bottom is the fundamental frequency (f0)

This correlates with the resonant frequency of vibrating vocal folds

(i) A high f0 correlates with high pitch4

(ii) A low f0 correlates with low pitch(iii) Voiceless sounds lack a f0

(2) Vowel QualityThe frequencies where there is lots of energy (dark bands) are resonant,or formant frequencies, (F1, F2, F3...)

These are the resonant frequencies associated with the vocal tract con-figuration (which acts like a Helmholtz Resonator) you have when youpronounce vowels

(i) High/Closed vowels are associated with a low F1(ii) Low/Open are associated with a high F1

i.e., Vowel height inversely correlates with the height of F1(iii) Front vowels are associated with a high F2(iv) Back vowels are associated with a low F2

i.e., Vowel frontness correlates with the height of F25

(3) Consonant Manner of Articulation

(i) Stops: Characterized by white spaces! (absence of energy/silence)(ii) Fricatives: Characterized by noisy energy in the high frequency

range (eg., 3000-5000 Hz range)(iii) Glides: Look like vowels!(iv) Nasals: Characterized by an anti-formant

(4) Consonant Place of Articulation

(i) Bilabials: Formant Lowering(ii) Alveolars: F1 lowers; F2 beds towards 1600 Hz

(iii) Velars: Characterized by ”velar pinch”

4Note that ’tone’ , pitch. ’Tone’ refers specifically to contrastive pitch.5Ladefoged (1993) suggests F2-F1, but Ladefoged (2001) just uses F2.

6 Perceptual Phonetics

Perceptual Phonetics - the study of how speech sounds are perceived (eg., pitchvs frequency)

• The physical properties of sound waves ,what we hear

• Our ears, eyes and brain filter/categorize the sound waves into what weultimately perceive

• Eg. Fundamental frequency (f0) and pitch (the way we perceive f0) arelogarithmically, not linearly, related.

° Perceptual phonetics is the study of how we perceive speech.

• The McGurk Effect

First listen and watch the video with your eyes open; then close your eyes -what do you hear?

References

Ladefoged, Peter. 1993. A course in phonetics. 3rd edition. New York, HarcourtBrace.

Ladefoged, Peter. 2001. A course in phonetics. 4th edition. Orlando: HarcourtCollege Publishers.

8

7 Practice with the IPA Chart

1. Provide the IPA symbol for the following sounds:

a) Voiceless, unaspirated, uvular stop .

b) Voiceless, palatal fricative .

c) Voiced, post-alveolar fricative .

d) Voiced, bilabial stop .

e) Voiceless, aspirated, alveolar stop .

f) (Voiced) alveolar nasal stop .

g) (Voiced) alveolar approximant .

2. Provide an articulatory description for the symbols in table 1 and 2.

3. Convert the following IPA transcription into English orthography:

(i) DEô w2z w2ns @phAn @ thajm @ phOô wIdow hu hæd @n ownli s2n nejmd>dZæk ænd @ khaw nejmd mIlkhi wajt

(ii) Al Dej hæd thu lIv An w2z D@ mIlk D@ khaw gejv EvrI mOônIN, wI>tS Dej

khEôid thu D@ mAôkIt ænd sowld.

(iii) b2t w2n mOônIN mIlkhi wajt gejv now mIlk ænd Dej dId nAt now w2t thudu.

(iv) “w2t Sæl wi du, w2t Sæl wi du?” sEd D@ wIdow, rININ hEô hændz

(v) “>tSIô 2p m2DEô, aj wIl fajnd w@ôk s2mwEô” sEd

>dZæk

(vi) “wi m2st sEl mIlkhi wajt ænd wIT D@ m2ni, owphIn @ SAp, Oô s2mTIN” sEdhIz m2DEô

IPA Voicing/Aspiration Place Manner

[t] voiceless, unaspirated alveolar stop

[kh]

[S]

[B]

[N]

[P]

[M]

[ñ]

[D]

[T]

[z]

[j]

[x]

[m˚

]

Table 1: Consonant Articulatory Features

(vii) sow>dZæk thUk D@ khaw ænd staôtId wælkhIN thuwOôdz D@ mAôkit.

9

IPA Tense/Lax Height Backness Round/Unround

[i] tense high front unrounded

[y]

[A]

[2]

[O]

[u]

[W]

[U]

[@]

[æ]

[e]

[E]

[ø]

[œ]

Table 2: Vowel Articulatory Features

10

CONSONANTS (PULMONIC) © 2015 IPA Bilabial Labiodental Dental Alveolar Postalveolar Retroflex Palatal Velar Uvular Pharyngeal Glottal

Plosive Nasal Trill Tap or Flap Fricative Lateral fricative Approximant Lateral approximant

Symbols to the right in a cell are voiced, to the left are voiceless. Shaded areas denote articulations judged impossible.

CONSONANTS (NON-PULMONIC) Clicks Voiced implosives Ejectives

Bilabial Bilabial Examples:

Dental Dental/alveolar Bilabial

(Post)alveolar Palatal Dental/alveolar

Palatoalveolar Velar Velar

Alveolar lateral Uvular Alveolar fricative

VOWELS Front Central Back

Close

Close-mid

Open-mid

Open Where symbols appear in pairs, the one to the right represents a rounded vowel.

OTHER SYMBOLS Voiceless labial-velar fricative Alveolo-palatal fricatives Voiced labial-velar approximant Voiced alveolar lateral flap Voiced labial-palatal approximant Simultaneous and

Voiceless epiglottal fricative Affricates and double articulations can be represented by two symbols joined by a tie bar if necessary.

Voiced epiglottal fricative Epiglottal plosive

SUPRASEGMENTALS Primary stress Secondary stress Long

Half-long

Extra-short

Minor (foot) group

Major (intonation) group

Syllable break

Linking (absence of a break)

DIACRITICS Some diacritics may be placed above a symbol with a descender, e.g. Voiceless Breathy voiced Dental

Voiced Creaky voiced Apical

Aspirated Linguolabial Laminal

More rounded Labialized Nasalized

Less rounded Palatalized Nasal release

Advanced Velarized Lateral release

Retracted Pharyngealized No audible release

Centralized Velarized or pharyngealized

Mid-centralized Raised ( = voiced alveolar fricative)

Syllabic Lowered ( = voiced bilabial approximant)

Non-syllabic Advanced Tongue Root

Rhoticity Retracted Tongue Root

TONES AND WORD ACCENTS LEVEL CONTOURor Extra or Risinghigh High Falling Mid High

rising Low Low

rising Extra Rising-

low fallingDownstep Global rise Upstep Global fall

THE INTERNATIONAL PHONETIC ALPHABET (revised to 2015)

Typefaces: Doulos SIL (metatext); Doulos SIL, IPA Kiel, IPA LS Uni (symbols)