07440

English
Phonetics
and Phonology
for Spanish
Speakers
Brian Mott
bi
Universitat, 49
2
a

e
d
i
c
i
n
(

)
( )
CONTIENE
CD
(1*/,6+3+21(7,&6$1'3+212/2*<
)2563$1,6+63($.(56
81,9(56,7$7
(1*/,6+3+21(7,&6$1'3+212/2*<
)2563$1,6+63($.(56
%ULDQ0RWW
Publicacions i Edicions
UNIVERSITAT DE BARCELONA
U
B
CONTENTS
LIST OF ILLUSTRATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
CD INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
SYMBOLS AND ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
FOREWORD to the second edition by 1ack Windsor Lewis . . . . . . . . . . . 23
PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1. PHONETICS AND PHONOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
1.2. Phonotactics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.3. The phonetics-phonology interIace . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
1.4. Structuralism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.5. Language universals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2. THE ORGANS OF SPEECH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.2. Initiation: the lungs and the act oI respiration . . . . . . . . . . . . . . . . . . . . 45
2.3. Phonation: the larynx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.4. Articulation: the supraglottal cavities . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.5. Coarticulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3. THE CLASSIFICATION OF SPEECH SOUNDS . . . . . . . . . . . . . . . . . 65
3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.2. The classifcation oI vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.3. Diphthongs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.4. Vowel systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.5. The Cardinal Vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
English Phonetics and Phonology Ior Spanish Speakers 8
3.6. The classifcation oI consonants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.7. Obstruents and sonorants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.8. Some statistics concerning vowels and consonants . . . . . . . . . . . . . . . . 85
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4. PHONETIC TRANSCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.2. Types oI phonetic transcription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.3. The symbols used Ior transcribing English . . . . . . . . . . . . . . . . . . . . . . 96
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5. THE ENGLISH PHONOLOGICAL SYSTEM . . . . . . . . . . . . . . . . . . . 107
5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.2. The English vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.2.1. Introauction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.2.2. The English vowels in aetail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
5.3. The English diphthongs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.3.1. Introauction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.3.2. The English aiphthongs in aetail . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.3.3. Levelling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
5.4. The English consonants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
5.4.1. Introauction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
5.4.2. The English consonants in aetail . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
5.4.2.1. The English plosives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
5.4.2.2. The English fricatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
5.4.2.3. The English affricates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
5.4.2.4. The English nasals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
5.4.2.5. The English approximants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
5.4.2.6. The aistribution of English /f/ ana /w/ . . . . . . . . . . . . . . . . . . . 141
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
6. CONNECTED SPEECH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
6.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
6.2. Assimilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.2.1. General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
6.2.2. Assimilation in the enaings -(e)s~ ana -(e)a~ in English . . . . . 151
6.3. Elision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
6.4. Liaison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
6.5. Gradation (use oI weak Iorms) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
Contents 9
6.5.1. Introauction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
6.5.2. The commonest weak forms in English . . . . . . . . . . . . . . . . . . . . . . 157
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
7. RHYTHM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
7.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
7.2. The rhythm oI English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
7.3. The rhythm unit or Ioot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
7.4. The rhythm oI English poetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
8. STRESS AND PRONUNCIATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
8.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
8.2. SuIfxation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
9. STRESS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
9.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
9.2. Word stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
9.3. English word stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
9.3.1. English wora stress. general tenaencies . . . . . . . . . . . . . . . . . . . . . 188
9.3.2. English wora stress. woras of one syllable . . . . . . . . . . . . . . . . . . . 189
9.3.3. English wora stress. woras of two syllables . . . . . . . . . . . . . . . . . . 189
9.3.4. English wora stress. woras of three syllables . . . . . . . . . . . . . . . . . 191
9.3.5. English wora stress. woras of four syllables . . . . . . . . . . . . . . . . . . 191
9.3.6. English wora stress. the effect of afhxes . . . . . . . . . . . . . . . . . . . . . 192
9.3.6.1. Prehxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
9.3.6.2. Sufhxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
9.3.7. English wora stress. compouna nouns ana syntactic units . . . . . . . 195
9.3.7.1. Other English compounas . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
9.3.8. English wora stress. woras with variable stress . . . . . . . . . . . . . . . 199
9.4. English sentence stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
9.4.1. Broaa ana narrow focus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
9.4.2. The nuclear stress. broaa focus sentences . . . . . . . . . . . . . . . . . . . 207
9.4.3. The nuclear stress. narrow focus sentences . . . . . . . . . . . . . . . . . . 210
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
10. INTONATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
10.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
10.2. Tone languages and intonation languages . . . . . . . . . . . . . . . . . . . . . . 218
10.3. The Iunctions oI intonation in English . . . . . . . . . . . . . . . . . . . . . . . . 221
10.3.1. The attituainal function of intonation . . . . . . . . . . . . . . . . . . . . . . 221
10.3.2. The grammatical function of intonation . . . . . . . . . . . . . . . . . . . . 222
10.3.2.1. Clause aivision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
10.3.2.2. Subfect ana preaicate aivision . . . . . . . . . . . . . . . . . . . . . . . . 223
10.3.2.3. Distinguishing between aehning ana non-aehning relative
clauses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
10.3.2.4. Questions versus exclamations . . . . . . . . . . . . . . . . . . . . . . . 224
10.3.2.5. Questions versus statements . . . . . . . . . . . . . . . . . . . . . . . . . . 224
10.3.2.6. Any absolutely any versus any chosen at ranaom . . . 225
10.3.2.7. Direct obfect, obfect of one verb or two? . . . . . . . . . . . . . . . . 225
10.3.2.8. Appositional phrases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
10.3.2.9. Distinguishing sentences not aistinguishable in writing . . . . 226
10.3.3. The accentual function of intonation . . . . . . . . . . . . . . . . . . . . . . . 226
10.3.4. The aiscourse function of intonation . . . . . . . . . . . . . . . . . . . . . . . 227
10.4. The meaning oI the tunes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
10.4.1. The fall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
10.4.2. The rise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
10.4.3. The fall-rise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
10.4.4. The rise-fall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
10.5. The intonation phrase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
10.5.1. Internal analysis of the intonation unit . . . . . . . . . . . . . . . . . . . . 240
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
11. LENGTH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
11.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
11.2. Length as represented in the English spelling system . . . . . . . . . . . . . 250
11.3. Further details oI length in English . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
11.3.1. Length in the history of English . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
11.3.2. Length in Moaern English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
11.3.2.1. General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
11.3.2.2. Pre-Fortis Clipping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
12. COMPARING SOUND SYSTEMS: ENGLISH, SPANISH AND
CATALAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
12.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
Contents 11
12.2. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
12.3. The sound systems oI English, Spanish and Catalan . . . . . . . . . . . . . . 261
12.3.1. The vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
12.3.1.1. Hiatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
12.3.2. The aiphthongs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
12.3.3. The consonants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
12.3.3.1. The plosives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
12.3.3.2. The fricatives ana affricates . . . . . . . . . . . . . . . . . . . . . . . . . . 269
12.3.3.3. The nasals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
12.3.3.4. The liquias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
12.3.4. Consonant clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
12.3.4.1. Initial clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
12.3.4.2. Final clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
12.3.4.3. Intrasyllabic clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
13. THE PHONEME AND DISTINCTIVE FEATURES . . . . . . . . . . . . . . 279
13.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
13.2. Other examples oI phonemes and allophones . . . . . . . . . . . . . . . . . . . 280
13.3. Neutralization oI phonemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
13.4. Diaphones and variphones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
13.5. Problems in phonemic analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
13.5.1. Phonetic similarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
13.5.2. The Biuniqueness Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
13.5.3. Problems of segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
13.6. Distinctive Ieature theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
13.6.1. Introauction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
13.6.2. Distinctive features. Jakobson ana Chomsky . . . . . . . . . . . . . . . . 290
13.6.3. Distinctive features for English . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
13.6.3.1. Mafor Class Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
13.6.3.2. Features for consonants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
13.6.3.3. Features for vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
13.6.4. Distinctive feature theory ana the phoneme . . . . . . . . . . . . . . . . . 294
13.7. Phonological rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
13.8. Trubetzkoy and the theory oI distinctive oppositions . . . . . . . . . . . . . 298
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
14. THE SYLLABLE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
14.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
14.2. The composition oI the syllable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
14.3. Theories oI the syllable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
14.3.1. The Sonority Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
14.3.2. John Wells theory of syllabicity . . . . . . . . . . . . . . . . . . . . . . . . . . 312
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
15. SOUND CHANGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
15.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
15.1.1. Observability of souna change . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
15.1.2. Graaualness versus abruptness . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
15.1.3. Regularity versus irregularity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
15.1.4. Factors which constrain regularity . . . . . . . . . . . . . . . . . . . . . . . . 320
15.1.5. Souna change. conaitionea or unconaitionea? . . . . . . . . . . . . . . . 321
15.1.6. Directionality of change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
15.2. Theories oI sound change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
15.2.1. The Ease Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322
15.2.2. The functional view of language change . . . . . . . . . . . . . . . . . . . . 324
15.2.3. The linguistic substratum, superstratum ana aastratum . . . . . . . . 325
15.2.4. Sociolinguistic variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
15.2.5. The Invisible Hana Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
15.2.6. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
15.3. Regular sound change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
15.3.1. The Great English Jowel Shift (GEJS) . . . . . . . . . . . . . . . . . . . . . 328
15.3.2. Grimms Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
15.3.3. The High German Consonant Shift . . . . . . . . . . . . . . . . . . . . . . . . 331
15.3.4. Other regular souna changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
15.4. Irregular sound change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
15.4.1. Omission ana aaaition of sounas . . . . . . . . . . . . . . . . . . . . . . . . . 334
15.4.1.1. Aphaeresis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
15.4.1.2. Syncope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
15.4.1.3. Apocope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
15.4.1.4. Prothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
15.4.1.5. Epenthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
15.4.1.6. Paragoge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
15.4.2. Assimilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
15.4.2.1. Assimilation affecting consonants . . . . . . . . . . . . . . . . . . . . . 338
15.4.2.2. Assimilation affecting vowels . . . . . . . . . . . . . . . . . . . . . . . . . 339
15.4.3. Dissimilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
15.4.4. Metathesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
15.4.5. Acoustic equivalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
15.4.6. Metanalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
15.4.7. Hypercorrection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
Contents 13
15.4.8. Folk etymology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
15.4.9. Analogy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
15.5. The eIIect oI sound change on phonological systems: splits
and mergers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
15.5.1. Splits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
15.5.2. Mergers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
15.6. On the dating oI sound change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
APPENDIX A: PASSAGES FOR PHONETIC TRANSCRIPTION . . . . 359
APPENDIX B: BRITISH AND AMERICAN ENGLISH . . . . . . . . . . . . . 363
Further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
GLOSSARY OF TERMINOLOGY (English - Spanish) . . . . . . . . . . . . . . 377
KEY TO EXERCISES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
ILLUSTRATIONS
This list oI captioned illustrations does not include the fgures that accompany the
descriptions oI each oI the vowels, diphthongs and consonants oI English in sections
5.2.2, 5.3.2 and 5.4.2.
PAGE
Figure 1. The levels oI language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Figure 2. Sounds (phones), allophones and phonemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Figure 3. Allophones and phonemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Figure 4. The organs oI speech . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Figure 5. The larynx: Iront and side views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Figure 6. A simple waveIorm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Figure 7. Complex waveIorm showing three peaks oI diIIerent Irequency . . . . . . . . . . . . . . 50
Figure 8. WaveIorm oI the English word lash as spoken by the author . . . . . . . . . . . . . . . . . 50
Figure 9. How the glottis opens and closes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Figure 10. The various states oI the glottis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Figure 11. The parts oI the tongue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Figure 12. Front view oI fat tongue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Figure 13. Front view oI sulcalized tongue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Figure 14. Tongue positions Ior English || and |k| . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Figure 15. Spectrogram oI the English word lash as spoken by the author . . . . . . . . . . . . . . . 55
Figure 16. Relationship between vowel height and F1 value, and vowel Irontness
and F2 value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Figure 17. Approximate F1 and F2 values Ior Iour English vowels . . . . . . . . . . . . . . . . . . . . 56
Figure 18. Velic closure and velar closure during the articulation oI |k| and || . . . . . . . . . . . 58
Figure 19. The area in the mouth in which vowel sounds are produced by the changing
shape oI the tongue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Figure 20. The twelve English vowel phonemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Figure 21. The English vowel /i/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Figure 22. Tongue shape Ior the English vowel /i/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Figure 23. Approximate tongue positions Ior English // and /u/ . . . . . . . . . . . . . . . . . . . . . . 70
Figure 24. Tongue shape Ior English // . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Figure 25. Tongue shape Ior English /u/ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Figure 26. The terminology used to indicate tongue height and the part oI the tongue which
is raised during the articulation oI vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Figure 27. The French Iront rounded oral vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Figure 28. The English diphthongs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Figure 29. Vowel systems oI the world`s languages: illustrative examples . . . . . . . . . . . . . . . 76
Figure 30. The central vowels oI English and Portuguese . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Figure 31. The Cardinal Vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Figure 32 The Spanish apico-alveolar |s| . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Figure 33 The English blade-alveolar |s| . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Figure 34 English 'dark |l| . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Figure 35 The 'clear |l| oI Spanish as in pala . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Figure 36 The Spanish palatal || . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Figure 37 The place oI articulation oI the English consonants . . . . . . . . . . . . . . . . . . . . . . . . 82
Figure 38. The English consonants according to place and manner oI articulation . . . . . . . . . 85
Figure 39. The International Phonetic Alphabet (revised to 2005) . . . . . . . . . . . . . . . . . . . . . 102
Figure 40. Yod in Standard Southern British and General American . . . . . . . . . . . . . . . . . . . . 140
Figure 41. Progressive and regressive assimilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Figure 42. Rhythm Reversal: automatic pilot ~ automatic pilot . . . . . . . . . . . . . . . . . . . . . . . 172
Figure 43 Stress preIerences oI a random sample oI native English teachers resident
in Barcelona and Saragossa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Figure 44. The words riaer and writer as said by the author on a Ialling tone
(CD track 20) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
Figure 45. The Mandarin Chinese word ma pronounced on Iour diIIerent tones by a Iemale
speaker (CD track 20) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Figure 46. The pitch patterns in Stockholm Swedish Ior pairs oI words like buren the cage`
and buren borne` . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
Figure 47. The English word yes said by the author on the Iollowing tunes:
low Iall, high Iall, low rise, high rise, Iall-rise, rise-Iall . . . . . . . . . . . . . . . . . . . 230
Figure 48. Typical English utterances showing the nucleus and one or more
oI the other parts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
Figure 49. Typical English utterances with a Ialling and rising nucleus . . . . . . . . . . . . . . . . . 241
Figure 50. English utterance with a stepping head beIore the nucleus . . . . . . . . . . . . . . . . . . . 242
Figure 51. Long and short values in English Ior each oI the fve vowel letters . . . . . . . . . . . . 250
Figure 52. The symmetrical fve-vowel system oI Spanish . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Figure 53. The vowels oI English and Catalan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Figure 54. The vowels oI Catalan (Barcelona): examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Figure 55. The vowels oI Valencian: examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Figure 56. The consonant phonemes oI English, Spanish and Catalan . . . . . . . . . . . . . . . . . . 267
Figure 57. WaveIorm oI the English words ton, stun and aone as pronounced
by the author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Figure 58. Jakobson`s and Chomsky`s Distinctive Features . . . . . . . . . . . . . . . . . . . . . . . . . . 291
Figure 59. Distinctive Ieatures Ior English consonants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Figure 60. Distinctive Ieatures Ior English vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
Figure 61. Distinctive Ieatures Ior a fve-vowel system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Figure 62 The composition oI the syllable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
Figure 63 The sonority peaks and troughs oI the English word pyfamas . . . . . . . . . . . . . . . . 309
Figure 64 Sonority scale oI the sounds oI English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
Figure 65 The Great English Vowel ShiIt (1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
Figure 66 The Great English Vowel ShiIt (2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
Figure 67 Grimm`s Law (1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
Figure 68. Grimm`s Law (2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
Figure 69 Grimm`s Law (3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
Figure 70 The High German Consonant ShiIt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
Figure 71 Omission and addition oI sounds in words . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
Figure 72 Types oI assimilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
Figure 73 SSB and American English vowels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
PREFACE
I taught English Phonetics and Phonology on my own at the University
oI Barcelona Irom 1972 to the early 1990`s, having previously been in charge oI
the subject at the University oI Saragossa Irom 1969 to 1972. Classes were oIten
overcrowded and the acoustic conditions usually poor. Nevertheless, despite these
setbacks and the intrinsic diIfculty such a technical subject presented to many
students, all oI them recognized it as a valuable part oI their linguistic training,
and throughout these years I was sometimes asked whether I intended to publish
the content oI my course. Subsequently, 1991 saw the frst edition oI my A Course
in Phonetics ana Phonology for Spanish Learners of English (EUB, University oI
Barcelona).
This frst edition leIt much to be desired as regards Iormatting and general
layout, but served its purpose Ior several years by providing students with back-up
material to my classes, which prior to 1991 were only supplemented by an anthology
oI notes, although, oI course, students were always reIerred to the standard works oI
Jones and Gimson. Thanks to useIul Ieedback Irom both colleagues and students, in
1996 I was able to produce a revised version, which updated the phonetic symbols
and included more phonetic transcription oI examples than the previous edition.
This second edition was still defcient in many ways, not least the typesetting and
general presentation, so in 2000 the old coursebook became English Phonetics ana
Phonology for Spanish Speakers, published by UB, no. 41 in its Manuals series.
This was a rewritten version oI the old text with a chapter on the syllable added, plus
many new exercises throughout the work, and numerous changes and new examples
aIter extensive revision. The same desire to improve the existing version and keep
up with progress in the feld has provided the impetus Ior this second edition, which
incorporates the modifcations to transcription in the Longman Pronunciation
Dictionary as presented in the third edition (LPD 2008), notably the extended use
oI the unstressed )/((&( vowel (see chapter 4), and takes account oI the recent shiIts
in the articulation oI the vowels oI RP (or SSB, as some preIer to label the model).
Although, on the whole, I Iollow LPD3 as regards phonetic notation Ior English,
there are a Iew minor cases in which I disagree with Wells` transcription. It is, aIter
all, extremely diIfcult to decide on some occasions which weak vowel is actually
used in an unstressed syllable. Such is the case oI the word event, Ior which LPD3
gives /vent/, but which could just as easily, and more consistently, be represented
as /ivent/. Use oI the .,7 vowel here seems unnecessarily conIusing as it contradicts
Wells` rules Ior use oI the unstressed )/((&( vowel as set out in LPD3, but Iortunately
such cases are Iew and Iar between.
As Phonetics has become increasingly technical and experimental in recent
years, it also seemed essential to include some explanation oI data obtained Irom the
acoustic analysis oI speech (though there are still modern elementary coursebooks in
the subject that manage very well without it notably Roach 2009). Accordingly, the
text has been provided with a limited number oI waveIorms, spectrograms and F0
tracings, though anyone wishing to look Iurther into these aspects oI the physics oI
speech will need to consult the more specialized books on the market, such as Ashby
& Maidment 2005 or Clark, Yallop & Fletcher 2007.
The present text provides rather more detail in some areas, such as the supra-
segmentals, than the average introductory course on English phonetics and phonology,
but it is hoped that it will serve both beginners and more advanced students and
teachers alike. Phonetics and Phonology in the English Department oI the University
oI Barcelona is now taught in two parts lasting a semester each, so that some chapers
oI the book can be covered in Fontica i Fonologia Anglesa I, and others in Fontica
i Fonologia Anglesa II, and students and academics Irom other institutions will be
able to adapt the book to their own needs. There is also abundant material Ior students
oI History oI Language and Ior language enthusiasts in general to delve into.
Thanks to the artistry oI Joan Carles Mora, there are many illustrations oI the
organs oI speech and, in particular, the sagittal sections oI the speech organs that
show the articulation oI the individual sounds oI English in chapter 5. Advances in
audio technology have allowed me to produce better recorded material and add to
that already presented on the CD accompanying previous editions.
I decided to eliminate most oI the recordings oI sounds Irom languages other
than English which I used in my 1991 and 1996 publications in order to constrain
the scope oI the book. For the same reason, the original chapter 15 oI the 1996
publication and various appendices have also been removed, except Ior the one on
British and American English, which has been revised and expanded in view oI the
equal importance oI the two varieties around the world.
Bibliographies are notoriously Irustrating to update in this age oI inIormation,
when new editions oI publications appear with alarming Irequency, but it is hoped
that, at least in the most important cases, the latest edition oI works has been cited.
Brian Mott
October 2010
1. PHONETICS AND PHONOLOGY
1.1. Introduction

Phonetics is an empirical science (i.e. one based on the observation oI Iacts) which
studies human speech sounds. It tells us how sounds are produced, thus describing the
articulatory and acoustic properties oI sounds, and Iurnishes us with methods Ior their
classifcation. It is concerned with the human sound-producing capacity in general and
examines the whole range oI possible speech sounds. ThereIore, the inIormation which
is aIIorded by phonetics need not apply necessarily to any language in particular. The
subject is a pure science and, strictly speaking, it does not Iorm part oI linguistics,
although, naturally, it plays an important role in the teaching oI Ioreign languages.
It is also useIul in the acquisition oI good diction, in speech therapy Ior people with
speech impediments, in helping the deaI and deaI-mutes, and in sound transmission.
As is known, vowels are made up oI Iormants, i.e. a number oI diIIerent Irequencies,
the most dominant oI which combine to produce their distinctive qualities. Only the
frst two Iormants are essential Ior the identifcation oI a vowel, and this Iact is oI
special interest to researchers in such felds as telecommunications, speech synthesis
and Automatic Speech Recognition (ASR). One oI the applications oI ASR is to be
Iound in the feld oI aviation. II machines can be trained to respond to messages, then
many oI the tasks normally perIormed by pilots can be taken over by them, which
means that the pilot will have his hands Iree to carry out more important jobs.
Phonetics is divided into three main branches:
(i) ARTICULATORY PHONETICS, which studies the nature and limits oI
the human ability to produce speech sounds and describes the way these
sounds are delivered;
(ii) ACOUSTIC PHONETICS, which studies the physical properties oI
speech sounds (e.g. pitch, Irequency and amplitude) during transmission
Irom speaker to hearer (Irom mouth to ear);
(iii) AUDITORY PHONETICS, which is concerned with hearing and the
perception oI speech, or our response to speech sounds as received
through the ear and brain.

Unlike phonetics, phonology is a branch oI linguistics, the other major areas
being grammar (including syntax) and semantics. II phonetics provides descriptions
oI sounds and ways oI classiIying them, phonology is a kind oI Iunctional phonetics
which employs this data to study the sound systems oI languages. It applies linguistic
criteria to the material provided by phonetics, so its concern is scientifc theory,
studying the linguistic Iunctions oI sounds.
Out oI all the speech sounds which it is possible to produce, individual
languages make use oI only a small number (see fgure 2). Thus, they act rather
like a sieve. The sounds which are used vary Irom language to language, and within
each language these sounds resolve themselves into 'Iamilies and Iorm a system
oI contrasts. It is these contrasts which are oI interest to the phonologist, who uses
the terms DISTINCTIVE, CONTRASTIVE, FUNCTIONAL or INFORMATION-
BEARING to describe such oppositions as that oI /k/ and /b/ in the words cat and
bat in English. The sounds /k/ and /b/ have a semantic value in that they serve to
distinguish words in English, and are called PHONEMES, which are the basic units
oI phonology.
It is important to distinguish these contrastive units, phonemes, which have
a communicative value within a given language system Irom other sounds that are
non-contrastive. For example, English has two principal types oI |l|, which are
impressionistically labelled 'clear and 'dark, respectively. The so-called 'clear
|l| occurs beIore vowels, as in the word lake; the other |l| (symbol ||) appears aIter
vowels, as in the words tall and chila. Now, iI I substitute 'dark |l| Ior 'clear |l| in
lake, I do not change the meaning oI the word. My pronunciation will sound a little
odd because oI the diIIerent distribution oI the two varieties oI |l| in English ('dark
|l| is not used beIore vowels), but as there is no phonemic opposition between these
two sounds, no semantic change occurs. These similar but non-contrastive sounds are
called ALLOPHONES.
In addition to the Iact that not all the diIIerent sounds in a language are
contrastive, it is equally important to note that diIIerent languages organize sounds
diIIerently and have diIIerent systems oI contrast (see fgure 3), a Iact which is oI
supreme importance Ior the language learner (see 12). In English, the two kinds oI |l|
we have described belong to the same phoneme (note that the phoneme is not a single
sound!), but in Russian, 'clear |l| and 'dark |l| are distinctive. This means that,
iI we substitute 'dark |l| Ior 'clear |l| in certain Russian words, we may produce
other words with diIIerent meanings, just as, iI I substitute /b/ Ior /p/ in the English
word pat, I produce a recognizably diIIerent sequence oI sounds with a diIIerent
meaning or, iI you like, another English word, bat. An example is provided by the
Russian Iorm aal, which means distance` iI pronounced with 'clear |l|, but he
gave` iI pronounced with 'dark |l|.
Let us take another example. In addition to the two |l|-sounds described,
there is another in English which is called devoiced. To understand in what way this
|l| is diIIerent Irom the others we have mentioned, say the words blaae and playea
31 Phonetics and Phonology
to yourselI. II you then try and isolate the segments |bl| and |pl| and say them by
themselves, it should be possible Ior you to notice that the |l| in |pl| is not quite
the same as the one in |bl|. At least the beginning oI the |l| in the sequence |pl|
sounds as iI it is preceded by aspiration (an |h|-sound). Once again, however, the
two types oI |l| do not serve to distinguish words, they do not make any diIIerence
to meaning. On the other hand, they may do in other languages and, in Iact, in Welsh
these sounds are in opposition (just as /p/ and /b/ or /p/ and /t/, Ior example, are in
opposition in English).
Other examples oI allophones are provided by the |k|-sounds in the English
words cool and keep, the |p|-sounds oI English spot and pot, and the |s| and |z| in
Spanish aesear and aesae, respectively. In all these pairs oI words, variants oI sounds
are used depending on the position in which they occur. But these positional variants
are not perceived as diIIerent by the native speaker and, as Iar as s/he is concerned,
they are the same sound, just as slightly diIIerent shades oI red are still reds, and a
jacket with two buttons and a jacket with three buttons are still jackets.
II you pronounce the words keep and cool slowly, you should be able to
Ieel your tongue making contact in each case with a diIIerent part oI the rooI oI
the mouth. In keep the contact is made Iurther Iorward than in cool. However, this
degree oI Irontness does not bring about a SYSTEMIC diIIerence, i.e. a change in
the system Irom one sound to another with a consequent change oI word meaning. It
is a NON-SYSTEMIC (NON-DISTINCTIVE / REDUNDANT / PREDICTABLE)
Ieature in English. But there is a language, Macedonian, a Slavonic language related
to Bulgarian, which opposes these two kinds oI |k|. Thus, in this language, kuka
with the |k| oI cool means hook`, whereas kukfa with the |k| oI keep means house`.
In English, certain consonants are aspirated (pronounced with a puII oI air
aIter them, rather like an |h|) beIore stressed vowels. This is the case oI the |p| in pot.
On the other hand, iI |s| precedes, as in spot, no aspiration is heard. The importance
oI this aspiration in English and its actual occurrence will be dealt with later in the
book (5.4.1, 12.1, 12.3.3.1), but Ior the moment suIfce it to say that it is a redundant
Ieature. II we pronounce spot with an aspirated |p|, we will not change the word;
it will just sound strange. However, in other languages aspiration may be used as a
distinctive Ieature. Thai opposes aspirated and unaspirated |p|, and in Hindi there are
not only aspirated and unaspirated |p|-sounds but also |b|-sounds distinguished by
the presence or absence oI aspiration, so that this language has the phonemes /p, p
,
b, b
/.
English has the phonemes /s/ and /z/ in words like Sue and :oo, respectively,
and Romanian also uses this contrast: virtuos virtuous` v. virtuo: virtuoso`. In
Spanish, these sounds also exist, but they are allophones oI the /s/ phoneme: |z|
occurs beIore certain consonants like |d| and ||, as in aesae and aesgarrar, while |s|
occurs in other positions (saber, aesear, mas).
The examples oI phonemic opposition which have been given so Iar are all
consonantal, but languages also have diIIerent vowel contrasts. English, Ior example,
has the phonemes /i/ and //, long and short varieties oI an |i|-type vowel, which
serve to distinguish words like sheep and ship, beat and bit, heat and hit, etc. Many
languages do not have such a contrast, and speakers oI these languages fnd it diIfcult
to hear and make the diIIerence when learning English. Similarly, Catalan has two
types oI |e|, as exemplifed in the words aeu god` (close |e|), and aeu ten` (open |e|).
More will be said about the meaning oI the terms close and open as applied to vowels
later in this book (3.2); Ior the time being, we can simply say that the open variety oI
|e| has a lower tongue position during articulation than the close variety. Spanish has
only one phoneme in this area oI articulation: /e/; thereIore, Spanish speakers have
diIfculty in distinguishing the two aIorementioned phonemes oI Catalan, although
Spanish does in Iact have a closer |e| in pera pear` than in perra bitch` with no
contrastive value.
As we have seen, the non-distinctive realizational variants oI phonemes,
(called allophones), tend to occur in specifc phonetic contexts, so that we can say
that the English /p/ phoneme has two principal allophones, one oI which is aspirated
and occurs in particular beIore stressed vowels, and the other unaspirated occurring
aIter |s|. As these allophones do not occupy the same positions in words, we say
they are in COMPLEMENTARY DISTRIBUTION. The opposite oI complementary
distribution is FREE VARIATION (see 13.1).
Words like sheep and ship, chip and ship which are distinguished by one
phoneme are called MINIMAL PAIRS (see also 13.2).
As phonology is concerned with the semiotic value oI sounds, it is related to
semantics. In Iact, Saussure, the Iather oI modern linguistics, attempted to inaugurate
a discipline labelled 'semiologic phonetics, which was later tentatively renamed
'phonology by his Iollower Albert Sechehaye. The term 'phonology was adopted by
the Prague School oI Linguistics in the early 1920`s and has remained in use since then.

1.2. Phonotactics

Apart Irom describing the sound system oI a language and determining its
phoneme inventory, phonology is also concerned with phonotactics, that is, statements
oI permissible strings oI phonemes. Two given languages may have certain sounds in
common, but these sounds may not be combined in the same way. For example, both
Spanish and English have the consonant sound which we call theta (symbol || the
initial sound oI the English word thin) but, whereas in English theta can be Iollowed
by |r| at the beginning oI words (as in three, threaa, thrill, etc.), in Spanish this is
not a possible consonant sequence. Similarly, Russian permits initial |d| as in gae
where`, Italian has initial |zb| as in sbaglio mistake`, and Czech has initial |hl| as
in hlava head`, while Modern English possesses none oI these consonant clusters
(although Old English did, in Iact, have |hl| in many words like hl
-
ua loud`).
Phonotactics deals not only with the way consonants combine but also
with the position consonants and vowels may occupy in the syllable or word.
1
For
example, |h| is possible at the beginning oI syllables in English, as in the words have
and behina, but not at the end. On the other hand, in Romanian |h| is both syllable-
initial and syllable-fnal (e.g. hartie paper`, pahar glass`, a oaihni to rest`, auh
spirit`).
Sometimes, languages have no initial clusters oI consonants at all. This is the
case oI Classical Arabic and some varieties oI Modern Arabic, which do, however,
have fnal clusters oI two or three consonants. The opposite is also Iound: Spanish has
no fnal clusters but does permit initial grouping.
The study oI sequential constraints is important because, when we learn a
Ioreign language, the individual sounds oI that language may not be entirely new
to us and thereIore not present a problem, but there may be groups oI sounds which
are problematic. English and Spanish, Ior example, both have |s| and |t| but, unlike
English, Spanish does not allow the initial group |st|, which exists in English words
like stuay and starter. In Spanish we always fnd a vowel beIore the |s| so that the
|s| and |t| belong to diIIerent syllables. ThereIore, the Spanish cognate oI stuay is
estuaiar, and the Spanish pronunciation oI starter will be |estate|.
The Iact that sounds may be restricted as to their positions in syllables and
words will be a problem Ior Ioreign language learners. The Iact that English has fnal
|m| and Spanish has not is a problem Ior Spanish learners oI English, which is why
rum appears in Spanish as ron. The fnal consonant oI English song, which we call
'eng, will also present diIfculty Ior some Spanish learners because, although this
sound occurs in Spanish, it is always Iollowed by |k| or ||, as in cinco and tengo.
Incidentally, Galician speakers will have no problem here because oI the existence
in their language oI words like unha (Ieminine indefnite article), in which 'eng,
represented by nh occurs beIore a vowel. Neither oI the two preceding areas oI
diIfculty will apply Ior Catalans, who are Iamiliar with fnal |m|, as in am custard
caramel`, and ||, as in tinc I have`. For the English learner oI Spanish, fnal |e| will
be diIfcult because |e| only appears in English in syllables closed by a consonant
(bea, breaa, met, etc.). For the Catalan learner oI English, fnal |b|, |d| and || will
be a problem because these consonants are not used fnally in Catalan, always being
substituted by |p|, |t| and |k|, respectively. But there exist pairs oI English words
like cab and cap, baa and bat, aog and aock in English, in which these two sets oI
consonants are in phonological opposition.
1. For notes on the appropriateness oI the syllable as the basis oI consonant cluster description, see
12.3.4. An additional point in Iavour oI taking the syllable as the basis Ior studying permissible strings oI
consonants is that it seems to be the basic neurologically planned unit in the brain. Evidence Ior this is to
be Iound in the type oI speech errors we call spoonerisms: a gay of aales a aay of gales. The consonants
switched around in this kind oI slip oI the tongue (or slip oI the brain!) occupy the same position in the
syllable. For Iurther details oI speech errors, see Fry 1977: 84-88.
Sometimes, there are constraints on consonant-vowel combinations. For
example, the English consonant 'eng is only preceded by certain short vowels, as
exemplifed in the Iorms sing, sang, sung, song, but never by long vowels (except in
onomatopoeic or invented words like boing), and not oIten by the short vowel |e|,
except in one or two words, like 'eng itselI, the common words length and strength,
and the Iorm ginseng a species oI plant`.
It is apparent, then, that a knowledge oI syllable structure is so intuitive to
the native speaker oI a language that s/he fnds it hard to break away Irom his or her
customary patterns on learning a Ioreign language. It is also interesting to note that,
since languages have their own personal rules oI combination, even people oI little
linguistic knowledge can make a guess at what language a text is written in iI they
are presented with an unseen piece oI writing. II I handed an Italian text to an English
person uninitiated in languages and asked them iI it was written in French or Italian,
they might well recognize the language by the number oI words that end in a vowel.
And it is this Ieel Ior syllabic structure that leads to the creation oI onomatopeic
jingles which are supposed to mimic the words oI other languages. Catalan has a
whole collection oI these, like Elastics blaus mullats fan fastic, which is supposed
to represent the sounds oI German, or Teta tonta tanta tinta tota tunta, intended to
parody the structure oI Thai. In Spanish there are lots oI jokes based on language
structure. Among these are the Japanese ones like:
Como se llama el Ministro de Educacion japons? Michiko Suda.
Como se llama el Ministro de Justicia japons? Nikito Nipongo.
Intuitive knowledge oI the syllabic structure oI AIrican languages, which have
many cases oI consonant combinations like |mb|, |nd| and 'eng ||, even in word-
initial position, is revealed in Spanish jokes like:
Como dicen pan en AIrica? Bimbo.
Como dicen :apato en AIrica? Bamba.
Knowledge oI the syllabic structure oI our native language is so internalized
that it makes it easy Ior us to guess what another person is saying to us. Guessing,
in Iact, is what we are doing most oI the time; understanding a message that is being
communicated to us is largely guesswork. Language is rarely spoken in optimum
acoustic conditions and, iI we actually only hear about one halI oI what is being said,
that is enough Ior us to interpret the message and provide what we have missed. This
works to the extent oI our correcting errors. Try asking an English speaker Ior a tlean
alass. II you say it smoothly enough without too much hesitation or getting tongue-
tied, s/he will not bat an eyelid and simply assume that you said a clean glass since
the initial consonant sequences |tl| and |dl| do not exist in Standard English (see Fry
1977: 83-84). The same would apply iI we said stliae instead oI striae in view oI the
Iact that initial stl is inexistent in English.
On the semantic level, there is also the context to help us so that, iI I hear Paint
the fence ana the ?ate or Check the calenaar ana the ?ate, I can Iairly easily provide
the missing consonants || and |d|, respectively.
The number oI consonants which can be grouped together varies considerably
Irom one language to another. English can have three or Iour consonants together,
as can be seen in the words glimpsea /lmpst/ and sixths /sk(s)s/, which will be a
problem Ior Spanish learners, but not necessarily Ior speakers oI Slavonic languages,
in which consonant clusters are common. Georgian, Ior example, has initial clusters
oI consonants ranging Irom two to six segments.
Japanese has no consonant clusters; the syllabic structure is the primitive
consonant-vowel (CV) combination. This means that, when borrowing takes place
into Japanese, words are made to conIorm to this pattern. The English word club, Ior
example, appears as kurabu; similarly, milk is pronounced something like miruku,
and baseball comes out as besuboru. The Japanese can now pay Ior articles in shops
with a puripeiao kaaao pre-paid card`, a thin polyester voucher whose value is
encoded on a magnetic strip, used to make purchases until the pre-paid amount runs
out.
The same syllable pattern is Iound in Swahili, in which the English word
hospital takes on the Iorm hosipitali, while English one has the Luganda Iorm wanu
and the Walpiri Iorm wani. Brazilian Portuguese also has a tendency towards this CV
structure, as is maniIest in the pronunciation oI aavogaao lawyer` as /adivoadu/
(Parkinson, in Harris & Vincent, 1988: 141), and in words like ritmo /ritimu/, apto
/apitu/ and chic /iki/ (Cunha & Cintra 1990: 47).
Finnish has no initial clusters, so professor was borrowed as rohvessori, |pr|
reducing to |r|, and |I|, non-existent in the Finnish sound system, being represented
(rather redundantly, by a combination oI two sounds) as |hv|. Compare the opposite
case oI Greek |p|, an aspirated |p|, becoming |I| in Latin, e.g. Greek pharos ~ Latin
3+$586 lighthouse`.

1.3. The phonetics-phonology interface

Although phonology is not concerned with all the articulatory minutiae oI
sounds which Iall within the ambit oI phonetics, it does deal with the rules which
govern the use oI allophones. For the phonetician, sounds are phenomena in the
physical world; Ior the phonologist, sounds are linguistic items whose intrinsic interest
is their Iunction, behaviour and organization. Thus, it may be said that the basic
notions in phonology are UNIT, REALIZATION and DISTRIBUTION, concepts
which Iorm the backbone oI practically all pre-1960`s structural linguistics regardless
oI linguistic feld. In phonology the UNIT is the phoneme, the REALIZATIONS are
the allophones, which are the actual exponents oI this abstract class labelled phoneme,
and the allophones have a particular DISTRIBUTION, i.e. they occupy certain
customary positions in the speech chain.
II phonetics provides inIormation on the physiological and acoustic properties
oI sounds, phonology investigates what properties have a Iunctional, communicative
value, so the business oI phonology is observation and analysis, and the subject is
marked by abstraction and generality. Phonology draws on phonetic substance, and
linguists have resorted to many an analogy with the physical world in an attempt
to show the relationship between the two areas. For example, Kenneth Pike said
(1947: 57): 'Phonetics gathers raw material. Phonemics ( phonology) cooks it.
More recently, a Iriend and colleague oI OxIord University, John Charles Smith,
used a meteorological metaphor comparing phonology to climate and phonetics to
weather, an analogy that he actually extended to language system as a whole versus
language use.
II phonetics studies phonic substance, variables, phonology has regard to the
choice oI constants, the systematization oI Iunctional linguistic units. Phonetics and
phonology complement one another and, although Trubetzkoy spoke oI them as
two separate felds, there is considerable overlap. Phonetics tells us that the English
phoneme /d/ is a voiced alveolar plosive and that English /n/ is a voiced alveolar nasal,
but Ior Trubetzkoy and Jakobson, the Iounders oI the Prague School oI linguistics,
who are oIten reIerred to as Iunctionalists because they described phonology as the
study oI the function oI speech sounds, /d/ and /n/ served to distinguish English words
like aeea and neea, coae and cone. They would say that /d/ and /n/ were distinguished
by the nasalization correlation, or that /d/ is | nasal| and /n/ is | nasal|. These
linguists were more concerned with what Ieature(s) distinguished phoneme A Irom
phoneme B than with phonemes as separate units. For Iurther details, see 13.6.

1.4. Structuralism

Twentieth-century linguistics or structuralism, oI which phonology Iorms part,
is said to begin with the Genevan linguist, Ferdinand de Saussure (1857-1913), oIten
reIerred to as 'the Iather oI modern linguistics, and to have been established by
Leonard Bloomfeld (1887-1949), mainly through his book Language, because oI the
technique he proposed Ior identiIying and classiIying Ieatures oI sentence structure,
in particular, the analysis oI sentences into their constituent parts. However, the
idea oI language as a network oI interdependent relationships, i.e. as consisting oI
a number oI units (sounds, words, meanings, etc.) which derive both their essence
and their existence Irom their relationship with the other units and have no prior or
independent existence, can be traced back much Iurther than the last century. In the
words oI Jakobson and Waugh (1979: 165):

The idea oI language as a structured, coherent system oI devices Irom the smallest to the
highest units has Ior ages been enrooted in sciences, striving against the superstitious and
liIeless image oI a Iortuitous aggregate oI scattered particulars.

In the nineteenth century, W. von Humboldt, a precursor oI present-day
linguistic views and a thinker who greatly infuenced Chomsky, said that 'Nothing
in language stands by itselI but each oI its elements acts as a part oI a whole, a
statement which Iorms a link with the well known reIerence made to language
by Meillet, a pupil oI Saussure, as a system 'ou tout se tient (where everything
depends on everything else`). Moreover, Saussure himselI oIten compares a language
system to a game oI chess: the pieces on the board only have a value in relationship
to the other pieces. In chess, a queen may be a very valuable piece to have, the most
valuable, in Iact, but not iI it is threatened and its mobility is thus blocked by one or
more oI the opponent`s pieces, in which case its value may be reduced or nullifed.
And iI the queen is lost, the value oI the other pieces will increase as they now have to
Iulfl her role. Again, moving a piece does not just change the potential oI that piece;
it alters the whole relational set-up between all the pieces. The situation in language
is similar. There exists a sort oI state oI 'otherness, so that a unit is what the others
are not. Any change in the other units as regards number or value will aIIect this
particular unit, too.
On the lexical level, compare the French word mouton with the English word
sheep. These are rough equivalents, but the situation oI each oI these units in their
respective systems is not the same because in English, the word sheep, the animal,
is opposed to mutton, the Iood, while French mouton is involved in no such contrast
because the French word has both the zoological and the culinary use. In other words,
it occupies more semantic space or a wider area oI meaning than English sheep.
As a phonological example, take the case oI the English phoneme /s/ and the
Spanish /s/. Apart Irom diIIerences in articulation, they can never be exactly the
same because they belong to diIIerent language systems and are involved in diIIerent
relationships in these systems. For one thing, English /s/ is opposed to /z/ (cI. Sue v.
:oo, ice v. eyes, etc.), while in Spanish there is no such phonological contrast (see
1.1). As another example, consider the Iact that Japanese has only one phoneme in the
area oI English /l/ and /r/ (see 13.4).
Saussure`s conception oI language as a system oI mutually defning entities was
to be a major infuence on several schools oI linguistics, like the London School and
the Prague School, and the theoretical distinctions or dichotomies that he proposed
are now the basis oI linguistic study.
Saussure insisted on separating DIACHRONIC Irom SYNCHRONIC
linguistics, his justifcation being that the history oI a language has no relevance Ior
the speaker oI the modern language. It is oI no signifcance to a speaker oI Modern
French that the frst-person singular pronoun fe I` derives Irom Latin (*2 because
these two pronouns Iulfl diIIerent roles in their respective language systems. In
Latin, personal pronouns were not usually expressed while in French the reverse is
true: these pronouns generally accompany the verb, as in English.
A distinction was also drawn by Saussure between LANGUE and PAROLE,
langue being the underlying linguistic system on the basis oI which all speakers are
able to understand and produce speech, and parole being speaker perIormance, the
actual utterances a speaker produces.
The language system was seen by Saussure as consisting oI a conglomeration
oI SIGNS, each invested with an ARBITRARY meaning. By arbitrariness what
Saussure meant was that there was no reason why a particular concept or SIGNIFIED
should be reIerred to by any particular SIGNIFIER. Thus, the Iact that the concept
tree` is reIerred to by tree in English, arbol in Spanish, arbre in Catalan, Baum
in German, and so on, is pure convention, a sort oI contract between speakers in
a community, so that they do not all go around calling things what they like. For
Saussure, each sign had a VALUE determined by the value or meaning oI all the other
units in the system, as was illustrated above.
The signs in a particular system are related to each other in two ways: either by
combination or by similarity/contrast. The frst oI these relationships Saussure called
SYNTAGMATIC, and the second PARADIGMATIC. Thus the words in any given
sentence are related to each other syntagmatically, but they are related paradigmatically
to all the other words which might have replaced them. In the utterance I am a man,
I is syntagmatically related to am but, at the same time, paradigmatically related to
he, we, etc. In the word man, the phonemes /m--n/ are syntagmatically related and,
iI we compare man with mat, we can see that /n/ and /t/ are paradigmatically related
in the English phonological system.
Twentieth-century linguistic structuralism, with its preoccupation with
language as a system, is opposed to nineteenth-century historicism and atomism.
The problem with nineteenth-century linguistics, or philology, as it was called, was
that it was too atomistic: in dealing with sound change, it said that sound A became
sound B and leIt it there. In other words, it said what happened, but not why, and it
did not consider the consequences that a single change might have Ior the system
as a whole. Saussure wanted to know what the eIIect on the entire system was. For
example, does the Iact oI sound A changing to sound B produce homophones? And
are there then subsequent changes in the language system to resolve this homonymic
clash? Any Romance linguist will understand the meaning oI the statement 'The cat
killed the cock in Gascony. Latin &$778 cat` and *$//8 cock` both produced gat in
Gascon and, as we are dealing with two common domestic animals, conIusion would
arise iI gat were used to reIer to both. ThereIore, the meaning cock` was covered
by other words like faisan pheasant` and metaphorical bigey ( 9,&5,8, Fr. vicaire
curate, vicar, local governor`). There is a similar example Irom Chilean Spanish,
where cantora was used Ior both singer` and chamber-pot`. As the association oI a
singer with something one urinates into was a rather inIelicitous one, the Peninsular
Spanish word cantante became used to designate singer`, a Iunction which it could
now perIorm without the embarrassing interIerence oI cantora, which was leIt Iree to
cover the meaning chamber-pot`. In Spanish, Latin )1,&8/8 Iennel` and *(1,&8/8
knee` Iell together as hinofo, so the word roailla (a diminutive oI rueaa 527$
wheel`) was coined to cover the meaning oI the second Latin term. Hinofo with the
meaning knee` now only survives in the expression postrarse ae hinofos.
Sometimes no therapeutic change is necessary in cases oI homonymy. For
example, meat and meet, Irom diIIerent Middle English words, can co-exist perIectly
well in English owing to the Iact that one is a noun and the other is a verb, and
interIerence or misinterpretation is unlikely. In the utterance Im going to meet my
friena, in which meet is a verb, it will never be interpreted as meat. Likewise, meat in
Have some meat will never be conIused with meet.
In a sense, every refection on language has always been structural. The human
race has always tried to see some kind oI order in language. The invention oI writing
itselI was based on structural intuition; writing systems are, aIter all, phonemic.
Again, the English child who says givea instead oI gave, Dont arg me on the basis oI
Dont argue, or Youa never ao that, nevera you? or the tot in the Canary Islands who
says tuscrobios on the basis oI microbios, are applying an internalized knowledge
oI the structure oI their native languages, albeit inaccurately. But it is not until the
twentieth century that we hear oI structural linguistics as a systematically studied and
established science.
There are, however, two objections that can be raised to structuralist accounts
oI linguistic change (McMahon 1994: 32). First, iI every element in a language
system is dependent upon every other element, how could change ever take place
without the whole language construct collapsing? Second, iI units have no meaning
in isolation but only derive their value Irom the system oI which they are a part, how
can we ever compare diIIerent languages or diIIerent stages oI the same language?
For example, blue in English cannot be compared with the word Ior blue` in Russian
or Welsh, in strict structuralist terms, because these words do not cover exactly the
same area oI the colour spectrum and are involved in diIIerent relational contrasts
with other terms in their respective systems.

1.5. Language universals

BeIore closing this chapter on phonetics and phonology, there is one other
important concern oI phonology to be mentioned: the involvement oI the subject in
the study oI language universals.
The study oI language universals is based on the premise that 'underlying the
endless and Iascinating idiosyncrasies oI the world`s languages there are uniIormities
oI universal scope. Amid infnite diversity, all languages are, as it were, cut Irom the
same pattern (Greenberg et alia 1966: xv), a statement reminiscent oI Chomsky`s
notion oI universal grammar. Many universals are implicational (iI A, thereIore B).
The Iollowing are a sample:
(i) Word-order universals:

Languages with dominant VSO (Verb-Subject-Object) order almost always have
prepositions.
Languages with dominant SOV (Subject-Object-Verb) order oIten have postpositions
('prepositions Iollowing the noun instead oI preceding).
(ii) Syntactic universals:
Languages with dominant VSO (Verb-Subject-Object) order usually have the adjective
aIter the noun.
II the pronominal object Iollows the verb, so does the nominal object.
(iii) Morphological universals:
II a language has the category oI gender (masculine and Ieminine), it also has the category
oI number (singular and plural).
II a language has infection, e.g. English fright-s, it always has derivation, e.g. fright-ful.
(iv) Phonological universals:
Short vowels, like the one in English ship, imply long ones, like that oI sheep.
II a language has nasalized vowel phonemes, it also has oral vowel phonemes.
/n/ is the commonest nasal phoneme, and the existence oI /m/ implies the existence oI /n/.
II a language has // ('dark |l|) as a phoneme, it also has /l/ ('clear |l|).
When a linguistic unit or process is more natural than another, we say that it
is UNMARKED, the term MARKED being used Ior units or processes which are
less expected. Thus, 'clear |l| is unmarked, while 'dark |l| is marked and, with
reIerence to the |k|-sounds oI cool and keep, the frst is unmarked, while the second
is marked.
It should be apparent Irom what has been said in this chapter that, despite
the diIIerent concerns oI phonetics and phonology, they are not separated Irom each
other in diIIerent watertight compartments. It is not natural to study speech sounds
in a vacuum with no reIerence to their linguistic Iunction, and we cannot study the
linguistic Iunction oI sounds without reIerence to their articulatory and/or acoustic
properties iI only to be able to identiIy and label the sounds we are reIerring to.
Phonology makes use oI phonetics, especially 'taxonomic phonetics, through use
oI terms like voicea, aental and obstruent, etc., so phonetics is an indispensable
Ioundation to phonology. But phonology is primarily concerned with the 'higher
levels oI organization. It deals with systems and patterns oI sounds in languages,
and there is an emphasis on the relationships between those sounds rather than their
substance.
The separation oI phonetics and phonology is an unIortunate and artifcial
one which has only come about during the twentieth century. BeIore the onset oI
the twentieth century people studying speech did not classiIy themselves as either
phoneticians or phonologists; such unnatural, watertight categories did not exist,
and scholars oI language were just scientists, interested in discovering Iacts. The
riIt started to appear with the arrival oI structuralism, and phonology developed as
an independent science mainly through the Prague School. Trubetzkoy said that
phonetics is 'the study oI sound pertaining to the act oI speech, while phonology is
'the study oI sound pertaining to the system oI language. Since structural linguistics
was supposed to be concerned with the system oI language, it Iollowed that phonetics
was not part oI this; thus, phonetics and phonology driIted apart.
More will be said about some oI the points mentioned in this preliminary
chapter in subsequent chapters oI this book.
Further reading
For very brieI, elementary, Iocused summaries oI the business oI Phonetics, on the one
hand, and Phonology, on the other, see Yule 2010 (Bib. 4), chs. 3 and 4, respectively.
Good general introductions to Phonetics are Malmberg 1963 and O`Connor 1973. For
an up-to-date, simple explanaton oI the concerns oI Phonology, see Nathan 2008.
See also Duchet 1982, Fudge 1970 (includes diIIerent defnitions oI the phoneme),
Gussenhoven & Jacobs 2005 and Hayes 2009. For the sounds oI the world`s languages,
see LadeIoged & Maddieson 1996. For more on Saussure, see Saussure 1974 (1915),
Culler 1986, and Gordon 1996. For good coverage oI the subject oI language universals,
see Comrie 1989. For a list oI universals oI grammar, see Greenberg 1966. For discussion
oI the interrelationship between phonetics and phonology, see Ohala 1991 and 2010.
Exercises
Exercise 1. Fill in the gaps in the Iollowing sentences:
1. Sounds which distinguish meaning are called ... .
2. Pairs oI words distinguished by a single sound like English ship and chip are
called ... .
3. The opposite oI complementary distribution is ... .
4. The study oI possible combinations oI sounds in a given language is called ... .
5. |d| and |n| are distinguished by the nasalization ... .
6. Ferdinand de Saussure is oIten reIerred to as 'the Iather ....
7. Trubetzkoy and Jakobson were the Iounders oI the ... School oI Linguistics.
8. ... described language as a system 'ou tout se tient where everything depends
on everything else`.
9. Saussure insisted on separating ... Irom ... linguistics.
10. The relationship oI the English pronoun I to he is ... .
11. The letters VSO, reIerring to syntactic order, stand Ior ... .
12. Phonology is very closely related to another branch oI linguistics called ... .
Figure 1. The levels oI language
Figure 2. Sounds (phones), allophones and phonemes
The range oI possible speech sounds is very large, but individual languages use only some oI these,
which are organized into 'Iamilies or phonemes, in contrast with one another
Figure 3. Allophones and phonemes
Sounds which are allophones in one language may belong to diIIerent phonemes
in another language
Figure 4. The organs oI speech
ENGLISH SPANISH
1 nasal cavity cavidad nasal
2 oral cavity cavidad bucal
3 lips labios
4 teeth dientes
5 alveolar ridge alvolos
6 hard palate paladar duro
7 soIt palate paladar blando (velo)
8 uvula uvula
9 tongue lengua
10 tip (apex) punta (apice)
11 blade corona (lamina)
12 Iront predorso
13 back postdorso
14 root raiz
15 jaw mandibula
16 pharynx Iaringe
17 epiglottis epiglotis
18 oesophagus (Iood passage, gullet) esoIago
19 larynx laringe
20 vocal Iolds (vocal cords) cuerdas vocales
21 trachea (windpipe) traquea
47 The Organs oI Speech
more quickly (sometimes taking only a quarter oI a second) and extend the time oI
exhalation to suit our needs. ThereIore, iI we are going to utter a very long sentence,
we prolong the period oI pulmonic defation in order to allow time Ior this syntactic
unit. This means that we may be using as little as fIteen percent oI the total respiration
cycle Ior breathing in.
Although most speech sounds are produced on a (PULMONIC) EGRESSIVE
AIR STREAM, that is, while we are breathing out, it is also possible to speak on
an INGRESSIVE AIR STREAM, which we sometimes do when we are sobbing,
gasping Ior breath, expressing sympathy or showing pain. As an example oI the
latter case, imagine that you are mending something with a hammer and hit yourselI
on the fnger. This will probably cause you to suck air in as a reaction to the pain,
and you may even utter an accompanying word to verbalize your discomIort as you
inhale.
The types oI non-pulmonic sounds known as CLICKS and IMPLOSIVES are
also made by drawing air in, rather than exhaling. Clicks are made with the VELARIC
or MOUTH-AIR MECHANISM and the process involved is similar to kissing or
sucking. The tongue makes contact with the rooI oI the mouth at two places, one
being the velum, and a vacuum is created. When air is allowed in to fll the vacuum, a
sound which we call a 'click is heard. A similar eIIect can be achieved by placing
a glass upside down on a smooth, damp surIace, sliding it in this upside down position
to create a vacuum and then raising it Irom the surIace.
Clicks are usually marginal to language, so they can generally be described as
paralinguistic, although they do have phonemic status in some languages spoken in
South and East AIrica. For example, Xhosa, spoken in the south-eastern part oI South
AIrica, has three phonemic clicks, and the initial consonant oI the word Xhosa itselI
is a click. In English, and other languages, the Iollowing clicks are Iamiliar:
i. a 'tut-tut click, used to express disapproval.
ii. a 'gee-up click used to make a horse or a donkey move.
iii. a clop-clop sound made in imitation oI the hooves oI horses.
iv. a noise like pst or tsk used in inIormal situations to attract attention.
These sounds all have velar closure produced by raising the back oI the tongue
to the velum together with closure at a place oI articulation Iurther Iorward in the
mouth but, as little reIerence has been made to place oI articulation so Iar in this
textbook, no details will be given at this juncture. Some symbols used by the IPA
(International Phonetic Association) Ior clicks are shown in the chart at the end oI
chapter 4.
Spanish people also use the tsk click in inIormal situations similar to English
people. Apart Irom this, they oIten use a hesitation click when they are speaking
which expresses Irustration and hesitation in the verbalization oI an idea, and tends to
precede an appeal such as A ver si me entienaes. Another situation in which Spanish
speakers use a click is when denying something: they utter a sound rather like the
English disapproval click while wagging their index fnger, a gesture that adds a
greater impression oI rejection to the noise.
Implosives are made on a GLOTTALIC INGRESSIVE AIRSTREAM. In the
production oI implosive consonants, air is contained in the mouth and pharynx by
a closure at some point while the larynx is lowered, thus rareIying the air. When
the closure is released, air rushes into the mouth, producing a hollow explosion. As the
glottal closure is incomplete, air still fows through the vocal Iolds, so implosives are
usually voiced. Implosives are Iound in many languages spoken in America, AIrica
and Asia, one notable example being Swahili.
There is still another kind oI non-pulmonic sound. This is the EJECTIVE.
Ejectives are the complete opposite oI implosives. They are unvoiced and made on
a GLOTTALIC EGRESSIVE AIRSTREAM: the larynx is raised to compress the air
trapped in the mouth and pharynx between the closed glottis and the oral place oI
articulation beIore it is released. Ejectives are commoner than the other two kinds
oI non-pulmonic consonant just described, and occur in languages spoken in America,
AIrica and the Caucasus (e.g. Georgian).
2.3. Phonation: the larynx
The larynx is the hard casing around the vocal Iolds/cords. It protects the Iolds
Irom damage and, as it is a kind oI valve in between the lungs and mouth, it plays an
essential role in speech production, eating and breathing.
When we are eating, the Iolds are tightly shut to prevent the entry oI Ioreign
bodies into the windpipe. Normally, pieces oI Iood do not reach the vocal Iolds
because they are defected away Irom the larynx by the epiglottis, and additional
protection is provided by the so-called Ialse vocal Iolds or ventricular Iolds, which
are situated just above the vocal cords proper and are also closed during swallowing.
By closing the cords and locking air in the lungs, we also assist in muscular
eIIort oI the arms and abdomen. II you are liIting something heavy and Iail to hold
your breath, you will notice how much less strength you have in your arms. However,
we normally fll the chest cavity with air automatically when we are doing heavy
manual work.
THE VOCAL FOLDS/CORDS are two bands oI elastic tissue lying horizontally
across the larynx Irom back to Iront. They are joined at the Iront, where we fnd the
thyroid cartilage, the hard, projecting lump which you can Ieel in your neck and
which is oIten reIerred to as the Adam`s apple. At the back, the Iolds are connected to
the arytenoid cartilages, which may be made to move apart, thus opening the Iolds in
a triangular shape. The opening between the Iolds is known as the glottis.
During normal breathing, the Iolds are wide apart so that the air Irom the lungs
passes through Ireely without producing vibration. In speech the Iolds may adopt
three principal positions:
i. wide open and not vibrating
ii. close together and vibrating
iii. tightly shut with the air Irom the lungs pent up behind them
Figure 5. The larynx: Iront and side views
In the frst instance, as the air Irom the lungs is unobstructed, we produce
what are called UNVOICED SOUNDS, such sounds as |p| and |I|. II the cords are
so close together that the air Irom the lungs makes them vibrate as it passes through
the glottis, then we produce what we call VOICED SOUNDS. Vowels are by nature
voiced sounds and there are also many voiced consonants, such as |b| and |v|. These
latter two are diIIerent Irom |p| and |I| respectively only as regards the action oI
the vocal cords. In Phonetics we can say that |p| is distinguished Irom |b|, and |I|
is distinguished Irom |v|, by THE VOICE CORRELATION. The unvoiced-voiced
opposition is a physiological dimension which is made great use oI in language since,
by adding voice to unvoiced consonants, we can double the number oI consonants,
and thereIore the number oI utilizable consonantal phonological oppositions, in a
given sound system.
When the vocal Iolds vibrate, the speed oI vibration varies. This speed oI
vibration, which we call FREQUENCY, is measured in Hertz (Hz), a unit which
corresponds to the number oI cycles per second (cps), a cycle being one complete
period oI vibration or oscillation oI the cords. Thus 5 Hz means fve cycles per second,
and 5 kHz means fve thousand cycles per second, and so on. A simple waveIorm can be
seen in fgure 6. However, bear in mind that most waveIorms are not simple or SINE/
SINUSOIDAL WAVES, as they are called, but complex, consisting oI two or more
overlapping component waves oI diIIerent Irequencies (fgure 7). In the case oI vowels,
the Irequencies will be harmonics oI the Iundamental Irequency produced by the vocal
Iolds (F0 F zero), so with an F0 oI 200 Hz, there will be harmonics at 400 Hz, 600
Hz, and so on (labelled F1, F2, etc.). II the Irequencies occur quite randomly and at
varying volumes, the result is a waveIorm with no regular pattern like that Ior the fnal
consonant oI the English word lash, shown in fgure 8. Compare the regularity visible
in the waveshape oI the frst part oI the word, where we have the initial consonant and
the vowel, with the aperiodicity displayed by the consonant at the end.
Figure 6. A simple waveIorm
Figure 7. Complex waveIorm showing three peaks oI diIIerent Irequency
Figure 8. WaveIorm oI the English word lash as spoken by the author
Frequency is controlled by a combination oI Iactors. First, there is the tension and
length oI the vocal cords which can be modifed, and then there is the air pressure Irom
the lungs which can be regulated. The greater the air pressure Irom the lungs, the Iaster
the vibration oI the Iolds. The Iaster the vibration oI the Iolds, the higher the PITCH oI
the note that is produced. In normal speech, the Irequency range is Irom about 60 Hz
to 350 Hz, more than two octaves. In any individual the range used is normally about
one and a halI octaves, though sometimes very high notes are used. A very high note
in a Iemale voice might be 1,000 Hz. The average male voice uses Irequencies around
130 Hz (one and a halI octaves below middle C on the piano) and the average Iemale
voice uses Irequencies around the 220 Hz mark (three semitones below middle C on
the piano), while 270 Hz is the average Ior children. The human ear`s useIul Irequency
range is very wide: 20-20,000 Hz, at least in young human beings. Other animals, like
dogs and bats, can capture sounds at even higher Irequencies.
The power transmitted along a sound wave is called the INTENSITY and is
measured in decibels (dB), named aIter Alexander Graham Bell, the inventor oI the
telephone. Normal conversation at one metre is around 60 dB, heavy traIfc is around
80 dB, while 120 dB is very loud noise indeed.
Apart Irom variation in the Irequency oI vibration oI the vocal Iolds, there
may also be variation in the MODE OF VIBRATION, giving rise to diIIerent types oI
voice. In NORMAL VOICE or MODAL VOICE, an average amount oI air is escaping
Irom the lungs; this is the mode we use most oI the time when we are speaking. But
there also exist other types oI voice called CREAKY VOICE, BREATHY VOICE,
and WHISPER or MURMUR.
Creaky voice is sometimes called 'glottal Iry because oI the sputtering eIIect
achieved. It is produced when we are speaking under our breath, or to avoid disturbing
people in the vicinity oI our conversation, or in order to keep a conversation private.
It is also used by ventriloquists and will result Irom trying to speak while liIting a
heavy weight because the vocal Iolds are almost completely closed: they are only
open slightly at a point near the Iront and thereIore vibrate along less than halI oI
their length. The deIective periodicity that we perceive arises Irom weak breath Iorce:
we are expelling less than the average amount oI air Irom the lungs. Creaky voice is
sometimes used in English when we are speaking on a very low tone. This Irequently
happens at the end oI utterances when the voice drops. Some people simply use creaky
voice idiosyncratically as an aIIectation because it is associated with sophistication
in English.
Breathy voice is the opposite oI creaky voice in the sense that, instead oI using
weak breath Iorce, we use strong breath Iorce and expel the air rapidly Irom the
lungs, in no more than two seconds. The vocal Iolds are open but, as the expulsion
oI air is so strong, they are made to vibrate, or fap in the breeze like a pair oI sheets,
as it were. Because breathy voice may be used to say a very passionate I love you, it
is sometimes reIerred to as 'bedroom voice. It is not used a great deal in every-day
speech, but a shocked No' might be uttered using this kind oI phonation.
In whisper or 'library voice, the vocal Iolds vibrate but there is a considerable
opening through which quite a lot oI air escapes at the end oI the Iolds beside the
arytenoid cartilages. Thus, voicing is accompanied by a continuous escape oI air
generating the sound we know as whisper. II Ior breathy voice the glottis is open
and the air takes no more than two seconds to escape Irom the lungs, Ior whisper the
glottis is narrowed and the lungs empty in about fve to seven seconds.
A third consideration regarding the vibration oI the vocal Iolds is the
AMPLITUDE OF VIBRATION, which is related to loudness. The wider the Iolds
move apart in the open phase, the louder the resultant sound will be. Compare this
action oI the vocal Iolds with that oI a guitar string when you pluck it. The harder
you pluck it, the Iurther you pull it away Irom its position oI rest, and the louder the
sound will be.
Figure 9. How the glottis opens and closes
open glottis whisper voice creak closed glottis
(voicelessness) (glottal stop)
Figure 10. The various states oI the glottis
When the vocal cords are tightly shut so that the air Irom the lungs builds up
behind them, a GLOTTAL STOP (phonetic symbol: ||) is produced iI they suddenly
open. Producing a glottal stop is rather like coughing and the acoustic eIIect is similar
to that oI a cork being pulled out oI a bottle oI champagne.
The glottal stop is used marginally in most languages and oIten has a co-
articulatory Iunction (see 2.5), i.e. it is articulated together with some other sound.
In English, Ior example, fnal |p|, |t| and |k| in the words rap, rat and rack are
characteristically co-articulated with the glottal stop, as is the fnal consonant oI the
word teach. But the glottal stop can also actually replace other sounds in English
rather than accompany them, which is oIten the case oI non-initial |t|. Glottalization
oI |t| is a regular Ieature oI Cockney (the lower class speech oI London) and other city
dialects in England and Scotland in such words as butter and what, but is stigmatised
in most positions. As an anecdote to underline how natural glottalizing is to some
varieties oI English, there is the story oI the Iather who told his son to pronounce
butter with a |t| and not with a glottal stop, which the boy did; whereupon the Iather
immediately said |ts be| (Thats better) with the reviled glottal stop, showing
just how spontaneous and subconscious speech habits are.
One particular context in which glottalization oI |t| is not condemned in
English is beIore a nasal consonant, e.g. mutton, bitten, oIten pronounced |mn|
and |bn|, respectively. Although the glottal stop usually substitutes Ior |t|, it may
also appear in Cockney in place oI |p| and |k| and even |I|, as in the words people
|pil|, poker |p| and ofhce |s|. ThereIore, its phonemic status is not entirely
clear, though it is generally treated as an allophone oI the /t/ phoneme.
The glottal stop is also used to reinIorce initial vowels in English (and probably
other languages). For example, when reading a list oI words beginning with a vowel,
we usually begin each with a glottal stop. Thus, eat, ape, ear will be read out as
|it|, |ep|, ||. This abrupt beginning to vowels is known as the HARD ATTACK
(see also 12.3.1). We can also separate contiguous vowels in hiatus in this way, as in
co-operate |kpret|.
The glottal stop is a regular Ieature oI Danish, in which it is called the stoa,
and distinguishes pairs oI words like hun /hun/ she` and huna /hun/ dog`. It also
occurs in German, which has diIIerentiated Iorms like the verbs vereisen /Iaizn /
to Ireeze` ( Eis ice`), with a glottal stop, and verreisen /Iraizn / to go away`
( Reise journey`), with |r|.
There is a very marginal use oI the glottal stop in English when || and
|mm| may both be used to mean no`, as opposed to |h| and |mhm|, which
stand Ior yes`.
2.4. Articulation: the supraglottal cavities
There are three supraglottal cavities:
i. THE PHARYNX, the area at the back oI the mouth
ii. THE ORAL CAVITY or MOUTH
iii. THE NASAL CAVITY
These cavities act as resonators Ior the sound generated by the vocal Iolds in
the larynx. II we compare the Iunction oI the vocal Iolds to the strings oI an instrument
like the guitar or violin, then we may draw a parallel between the supraglottal cavities
and the hollow body oI such instruments, which provides the necessary amplifcation
Ior them. The principle oI resonance is that something which is itselI vibrating, like
the vocal Iolds, can Iorce into vibration some other system, the resonator, in our case
the volume oI air in the supraglottal cavities.
Figure 11. The parts oI the tongue
Figure 12. Front view oI fat tongue Figure 13. Front view oI sulcalized tongue
Figure 14. Tongue positions Ior ||, the English vowel oI cart, and |k|,
the consonant sound in the English word cook
The resonances set up in the vocal tract, which, in the case oI vowels, are
harmonics oI the Iundamental Irequency (see 2.3), are called FORMANTS. As the
vocal tract changes shape to produce diIIerent sounds, diIIerent Iormant Irequencies
come into play. Sounds are made up oI many diIIerent Iormants, but only the frst
three are important in helping us to distinguish them Irom one another: F1, F2 and
F3, particularly the frst two. In Iact, transmission oI speech via the telephone relies
mainly on F1 and F2, which is why our voices sound rather impoverished when we
use this device.
Only vowels have a sharply defned Iormant structure, as can be seen in the
SPECTROGRAM oI the English word lash (fgure 15 below), whose waveIorm you
viewed in fgure 8. In the spectrogram, the Iour dark horizontal lines above the vowel
are its frst Iour Iormants. Compare the lack oI a clear Iormant structure Ior the fnal
consonant, while the initial consonant, being a liquid and thereIore more vowel-like,
has a clearer one, though less well defned than Ior the vowel. Note also that, whereas
the waveIorm in fgure 8 shows amplitude on the vertical axis, the spectrogram marks
Irequency in that position. In both kinds oI illustration, the horizontal axis shows time.
Figure 15. Spectrogram oI the English word lash as spoken by the author
II we measure the averages Ior the frst two Iormants oI the vowel in lash /l/ in
a programme like PRAAT (downloadable Irom the web), we will see that they are very
approximately F1: 850 Hz and F2: 1600 Hz. As the diIIerence between these two
Irequencies, F1 and F2, is about 750 Hz, and we know that the diIIerence between F0 and
F1 will be the same, this means that we can presuppose an F0 oI about 100 Hz (850 Hz minus
750 Hz), which is normal Ior a male voice. Likewise, we can calculate the F3 by simply
adding 750 Hz to the F2 value, which is 1600 Hz. This will give us an F3 oI 2350.
The shape oI the oral cavity is determined mainly by the movement oI the
tongue. This is the most elastic and mobile oI the articulators called an ACTIVE
ARTICULATOR as opposed to a PASSIVE ARTICULATOR like the palate. It can
be stretched Iorward in the mouth, retracted and bunched up, and also move up and
down while adopting these other positions so that either the Iront or the back or the
centre is raised or lowered.
II the Iront oI the tongue is raised halI way, then a vowel like the one in the
English word bea is produced. II the back oI the tongue is raised halI way, then we
hear a vowel like that oI the word ball or bought. II the Iront oI the tongue is raised
until it is Iairly close to the palate, we produce a vowel like that oI leave and, iI the
tongue is lowered and retracted slightly Irom this position, we have the vowel in
live. The vowel oI cart is produced by holding the tongue maximally low so that the
highest point is a little in advance oI the centre oI the back. As a matter oI Iact,
this is the vowel a doctor in England will ask you to reproduce in order Ior him to
examine your throat. It would be no use asking you to say the vowel in leave as the
Iront-tongue raising characteristic oI this vowel would obstruct the doctor`s vision.
Figure 16. Relationship between vowel height
and F1 value, and vowel Irontness and F2 value
There is an inverse relationship between F1 and vowel height, so that the vowel
in leave will have a lower F1 than the vowel in live, which is more open. Typical F1
values Ior these vowels in RP or Standard English (see 5.1) might be 280 Hz Ior the
frst and 380 Hz Ior the second; the F1 Ior the vowel in lash above, a much lower
vowel than the latter two, was around 850 Hz. F2 corresponds directly with Irontness,
the Ironter vowels having higher values than the backer ones. Thus, the vowel oI
leave, which has higher Iront oI tongue raising than the vowel oI live, might have
a reading oI about 2500 Hz, while the vowel in live, which is slightly backer, could
have an F2 oI 2000 Hz. You may recall that the F2 Ior the vowel in lash above, which
is again articulated with the tongue a little Iurther back, was around 1600 Hz. A vowel
that is both very low and very back, like the one in the English word cart /kt/ (see
fgure 14), might have values like F1: 700 Hz, F2: 1000 Hz. II you compare the F1
and F2 values Ior the English vowels discussed in this paragraph, you`ll notice that
there is a greater diIIerence between the Irequencies Ior the high Iront vowels than
Ior the lower and backer vowels (see fgure 17), thus it is possible to make a guess at
the degree oI Irontness and openness oI a vowel on the basis oI fgures oI this kind.
F1 F2
leave /liv/ 280 2500
live /lv/ 380 2000
lash /l/ 850 1600
cart /kt/ 700 1000
Figure 17. Approximate F1 and F2 values Ior Iour English vowels
For some sounds there is a groove down the middle oI the tongue. This is the
case oI |s| in English words like sit and case. When the tongue adopts this shape, we
say that it is SULCALIZED.
In the articulation oI some sounds the lips play an important role. When sounds
are produced with the lips pushed Iorward or protruded, we call them ROUNDED or
LABIALIZED. In the languages oI the world, this is usually the case oI back vowels,
that is, the ones produced with back-tongue raising, and there is a good acoustic
reason Ior this. Back vowels have a GRAVE resonance, whereas Iront vowels are
ACUTE. Protrusion oI the lips elongates the oral cavity and enhances graveness (in
spectrograms it is the F3 that is related to this degree oI rounding); so, by combining
backness with rounding/labialization, we are maximizing graveness and thus making
our back vowels as distinct as possible Irom the Iront ones, just as traIfc lights use
red and green Ior stop and go, respectively, rather than two shades oI red or two
shades oI green, which would lead to conIusion. In acoustic terms, we could say that
rounding the lips has the eIIect oI lowering all the resonant Irequencies oI the vocal
tract or, putting it another way, it fattens the acoustic signal.
It should, however, not be Iorgotten that rounding can also combine with
Irontness and, sometimes, we also fnd back vowels that are unrounded. The French
words pur pure` and pire worse` are distinguished by the presence oI labialization
on the vowel in the frst case and absence oI it in the second, both vowels being Iront
articulations. In phonological terms, we can say that pur and pire are diIIerentiated
by the LABIALIZATION CORRELATION.
II we bring our two lips or our lips and teeth into contact or into close proximity,
or iI we raise the tongue so high that it almost makes contact or eIIectively touches the
passive articulators at some point, then consonant sounds are produced. As we shall see
in the next chapter (3.1), this absence versus presence oI contact between the organs
is the basis oI our physiological distinction between a vowel and a consonant. As an
example oI the production oI a consonant, take the case oI English |t|. To produce this
consonant, the tip oI the tongue is brought into contact with the alveolar ridge, that
is, the hard gum ridge just behind the upper teeth. This contact momentarily obstructs
the fow oI air Irom the lungs. When the contact is released, we hear the characteristic
explosion oI the |t| as the air escapes. For |k|, the initial and fnal consonants in the
English word cook, the back oI the tongue touches the velum beIore drawing away
again; Ior |p| the lips are closed and then abruptly opened, but there is no voice, i.e. no
vocal Iold vibration, which there is, on the other hand, Ior |b|, and so on.
Admission oI air to the nasal cavity is controlled by the velum. II the velum is
raised and touches the back wall oI the pharynx, then air passes exclusively through
the mouth, thus producing oral sounds. On the other hand, iI the velum is lowered, so
that air is allowed to enter the nasal cavity, then we talk oI NASAL or NASALIZED
sounds. The acoustic eIIect oI using the nasal cavity as a resonator is to add, not a
resonance, but an anti-resonance, which means that the strength oI a particular range
oI harmonics is reduced. The nasal cavity is unable to vibrate at certain Irequencies
and thus removes those Irequencies Irom the source sound.
The term nasal is usually reserved Ior the consonants |m, n, |, while
nasali:ea is applied to vowels since the passage oI air through the nose in the case oI
vowels is concomitant with its passage through the mouth, there being no complete
obstruction in this lower chamber. Consonants can have Iull nasal quality because oI
the obstruction to the air in the mouth.
II our nasal passage is congested because we have a cold, then consonants
which are usually nasal will be articulated as oral. For example, the greeting Gooa
morning /d mn/ might sound rather like /d bd/ since |b|, |d| and ||
are the oral equivalents oI |m|, |n| and ||, respectively. To use our phonological
terminology once again, we can say that these two sets oI consonants are distinguished
by the NASALIZATION CORRELATION.
It is important to recognize the close relationship between the latter two sets oI
consonants because |m| and |b|, |n| and |d|, and || and || Irequently occur together
in languages. Take, Ior example, Spanish words like bamba, anaa and mango /
mao/, and English words like chamber, sana and hnger /I/. Moreover, in
the history oI languages, these consonant groups sometimes simpliIy, as happened
to fnal |mb| and || in English when they reduced to |m| and ||, respectively:
climb /klam/, aumb /dm/, thing //, hang /h/. Latin 0$1'
-
$5( became manaar
in Spanish, but manar through loss oI |d| in Catalan. The reverse may also happen:
Old English th
-
ymel became Modern English thimble; Spanish camion is sometimes
pronounced cambion in Aragon, and so on. For more examples, see 15.4.1.5.
When the velum or soIt palate is raised to shut oII the nasal cavity, we speak
oI VELIC CLOSURE. The expression VELAR CLOSURE reIers to the action oI
raising the back oI the tongue to the velum in the production oI the consonants |k, ,
|. Thus the speech organs will adopt the approximate positions shown in fgure 18
during the articulation oI the oral consonants |k| and || in English.
Figure 18. Velic closure and velar closure during the articulation oI |k| and ||
2.5. Coarticulation
So Iar in this chapter, we have spoken oI Ieatures like labial, velar and nasal
as pertaining to individual sound segments. Thus, Ior example, we have described |b|
as labial, |k| as velar and |n| as nasal, and so on, as iI the Ieatures were properties oI
particular sounds. However, Ieatures may be spread over more than one segment. Thus
the vowel in the English word man is always somewhat nasalized as it stands between
two nasal consonants: phonological Iorm /mn/, phonetic realization |m
n|. So the
word is nasalized throughout the articulation oI the three segments, not just nasalized
at the beginning and at the end. Such additional Ieatures acquired by a sound due to
its position in a sequence are reIerred to as SECONDARY ARTICULATIONS and
are one type oI COARTICULATION. Apart Irom nasalization, the principal types oI
secondary articulation are:
i. labialization represented by a raised |w|: English switch |swt|
ii. palatalization represented by a raised yod (|j|): English tube |tjub|
iii. velarization represented by a tilde (|`|) placed through a symbol:
English ball |b|
In the frst example, the initial consonant is strongly labialized through
infuence oI the Iollowing semi-vowel; in the second example, the |t| is attracted
towards the palatal region by the Iollowing yod and could become Iully palatalized
as |t| (|tub|); in the third example, the |l| is velarized (dark) through being in
postvocalic position.
The Ieatures we have described need not merely aIIect the pronunciation
oI particular sound sequences, but may colour a sound system as a whole, so that
diIIerent communities and even individuals use particular VOICE SETTINGS or
PHONETIC SETTINGS, which give them their typical voice quality or TIMBRE/
TAMBER. American speakers oI English, Ior example, are oIten considered to make
use oI nasalization.
As well as having a linguistic Iunction, the settings are also used
paralinguistically to signal mood or emotion. These long-term adjustments oI the
vocal apparatus can be divided into (i) supralaryngeal settings, like the ones we
have just described, and (ii) laryngeal settings, like creaky voice, breathy voice and
whisper, as mentioned under 2.3. To these we could add harsh voice, in which the
larynx is very tense and the ventricular Iolds or Ialse vocal Iolds, which are located
just above the vocal Iolds, are pushed down onto the true vocal Iolds with the result
that the speaker sounds angry and aggressive. This type oI voice is typical oI the jazz
singer Louis Armstrong.
The general supralaryngeal setting oI a language is determined by the most
Irequent sounds in that language. In English there are a number oI alveolar consonants,
so English tends to sound alveolarized. Moreover, the tongue-tip is used quite a lot,
while in French it is the blade or the Iront oI the tongue that comes into action.
English is also characterized by neutral lips held loosely apart, probably because ||,
the vowel oI hit, and schwa, the second vowel oI better, which have this Ieature, are
very common, though there is occasionally more opening oI the mouth Ior sounds
like |a|, |a| and ||. French, on the other hand, is characterized by a great deal oI
lip-rounding. In Russian, as in English, the lips are spread as there are a lot oI palatal
articulations. The tongue is hardly ever visible in the English setting, as opposed to
French with its Irequent lowering oI the jaw because open vowels are common in this
language. On account oI the rounded lips and concave cheeks, French words give the
impression oI being Iorced out oI the mouth.
In Turkish and Persian, dental consonants are common, the tongue is well
tapered and the pointed tip is the predominating articulator, so these languages may
be described as dentalized. The languages oI India and Pakistan have retrofexion
with open jaws. In Arabic, and oIten in German (especially among men), pharyngeal
contraction occurs; this type oI speech may be reIerred to as pharyngealized
(pharyngealized sounds have a raised || attached to them).
German, Italian and Polish (and many other languages) resemble English as Iar
as the hard contact oI the tongue with the passive articulators is concerned, while in
Spanish and Danish there is very noticeable lenition, i.e. weak contact oI the speech
organs. In the languages oI India and Pakistan, bilabial contact is weak, probably
owing to the jaw setting conditioned by the presence oI RETROFLEX sounds, i.e.
sounds produced with the tip oI the tongue curled back. American English and Irish
English have a retrofex |r|, while the speech oI Birmingham (England), Liverpool,
the Bronx and some types oI Houston accents is noticeably velarized, the back oI the
tongue being raised towards the velum.
Another type oI coarticulation is DOUBLE ARTICULATION. This involves
two simultaneous articulations oI equal rank, as in the Yoruba words |ak
pa| meaning
arm` and ag
ba meaning adult` (LadeIoged & Johnson 2006: 172), in which the

plosive consonants are pronounced with simultaneous closure at the velar and labial
places oI articulation. To underline the Iact that the articulation is co-terminous,
rather than sequential, we use a tie in transcription.
An example Irom English is provided by the semivowel or approximant |w|,
which is pronounced with concomitant rounding oI the lips and raising oI the back
oI the tongue towards the velum, and thereIore described as labial velar (labiovelar).
Remember also that the English unvoiced plosives in syllable codas are oIten
co-articulated with glottal closure (see 2.3): rap |rp|, rat |rt|, rack |rk|.
Apart Irom |w|, there are other sounds in English that are characterized by
a particular confguration oI the tongue with attendant lip-rounding which does not
derive Irom the infuence oI contiguous labialized sounds: || and |r|, Ior example,
have a post-alveolar place oI articulation together with lip-rounding Ior many
speakers (London speakers oIten substitute a kind oI |w|-sound Ior |r|, in Iact). This
should be distinguished Irom the case oI sounds with secondary labialization which
derives Irom context, a kind oI assimilation (see above and also 6.2 and 15.4.2), just
like the devoicing that |l| and |r| undergo in English aIter |p|, |t| and |k| in words like
place |pl es|, try |tr a| and clock |kl k|.
The present chapter is oI Iundamental importance to the student oI phonetics
and phonology as it Iorms the basis oI all Iurther study in these felds. Familiarity
with the speech organs and how they Iunction is the key to an understanding oI many
problems oI speech production. Historical changes by which one sound develops
into another or is replaced by another can only be Iully appreciated in the light oI
our knowledge oI physiological and acoustic phonetics. The infuence oI one sound
over another in the speech chain (Ior instance, change oI |n| to |m| beIore |p| or
|b|) cannot be understood without recourse to inIormation about the workings oI
the speech organs and the essential relationships between sounds. ThereIore, you
are strongly recommended to make sure that you have understood the essentials oI
INITIATION, PHONATION and ARTICULATION as set Iorth at the beginning
oI this chapter beIore proceeding with this course.
Further reading
All phonetics books provide a chapter on how the speech organs work. Fry 1977,
Denes & Pinson 1993, and Laver 1994 provide good coverage, while Ashby &
Maidment 2005 oIIers a very clear, up-to-date account. CatIord 2001 teaches readers
how sounds are produced by careIully explaining to them how to experiment with their
own vocal tract. For detailed inIormation on the anatomy and physiology oI speech
and the acoustics oI speech production and perception, see Ball & Rahilly 1999;
Clark, Yallop & Fletcher 2006; Reetz & Jongman 2009. Fry 1979 deals thoroughly
with the acoustic aspects. Ogden 2009 also provides a very complete explanation oI
the diIIerent types oI articulation, accompanied by the relevant acoustic inIormation.
For a book that integrates a comprehensive introduction to speech articulation,
acoustics and perception (chapters 2 and 5) into the study oI language and mind, see
Byrd & Mintz 2010.
For masterIul treatment oI the physics and acoustics oI speech sounds, see
LadeIoged 1996, and LadeIoged & Johnson 2006. Rogers 2000 provides considerable
articulatory and acoustic detail oI the English spoken not only in Britain and America
but in many other places around the world, together with numerous illustrative
spectrograms. Martinez 2007 provides a large number oI spectrograms oI illustrations
Irom Spanish.
Exercises
Exercise 1. Give an alternative name in English Ior:
1. trachea 2. oesophagus
3. tip oI the tongue 4. velum
5. vocal cords 6. thyroid cartilage
7. creaky voice 8. breathy voice
9. whisper 10. glottal stop beIore initial vowels.
11. labialization 12. vocal Iolds
Exercise 2. Complete the Iollowing:
1. The three major parts oI the speech apparatus are ..... .
2. The most important organ in the mouth is the ..... because oI its mobility.
3. Velic closure is raising oI the ..... to touch the ..... .
4. Velar closure is raising oI the ..... to touch the ..... .
5. The tongue is an active articulator, but the palate is a ..... articulator.
6. The resonances set up in the vocal tract are called ..... .
7. The opening between the vocal cords is called the ..... .
8. Clicks are produced by the ..... mechanism.
9. The appendage attached to the soIt palate is called the ..... .
10. The supraglottal cavities are ..... .
11. The rooI oI the mouth is divided into ..... .
12. Most speech sounds are produced by using air expelled Irom the lungs and
thereIore they are ..... .
13. The frst stage oI speech production, i.e. expelling air Irom the lungs, is reIerred
to as ..... .
14. An F1 value shows degree oI vowel ....., and an F2 value shows degree oI vowel
..... .
15. A simple waveIorm is reIerred to as a ..... wave.
16. Frequency oI vocal Iold vibration is measured in ..... .
17. A visual display oI the component parts oI a sound wave is called a ..... .
18. F1 and F2 are ..... oI the Iundamental Irequency (F0).
19. The Iaster the vocal Iolds vibrate, the higher the ..... oI the sound produced.
20. The human ear`s useIul Irequency range is ..... Hertz.
Exercise 3. Label the organs oI speech on the Iollowing diagram, and then look back
at fgure 4 to check your answers.

07440

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

07440

Uploaded by

Copyright:

Available Formats

English

ba meaning adult` (LadeIoged & Johnson 2006: 172), in which the

You might also like