Speech Production 5 Figure 1.2 The vocal tract. nasal tract or a/ t r a larynx the speech production mechanism as a whole.Figure 1.3 shows the four main components-the airstream process,the phonation process,the oro-nasal pro- cess,and the articulatory process.The airstream process includes all the ways of pushing air out(and,as we will see later,of sucking it in)that provide the power for speech.For the moment,we have considered just the respiratory system,the lungs pushing out air,as the prime mover in this process.The phonation process is the name given to the actions of the vocal folds.Only two possibilities have been mentioned:voiced sounds in which the vocal folds are vibrating and voice- less sounds in which they are apart.The possibility of the airstream going out through the mouth,as in v]or [z],or the nose,as in [m and [nl,is determined by the oro-nasal process.The movements of the tongue and lips interacting with the roof of the mouth and the pharynx are part of the articulatory process
Speech Production 5 the speech production mechanism as a whole. Figure 1.3 shows the four main components—the airstream process, the phonation process, the oro-nasal process, and the articulatory process. The airstream process includes all the ways of pushing air out (and, as we will see later, of sucking it in) that provide the power for speech. For the moment, we have considered just the respiratory system, the lungs pushing out air, as the prime mover in this process. The phonation process is the name given to the actions of the vocal folds. Only two possibilities have been mentioned: voiced sounds in which the vocal folds are vibrating and voiceless sounds in which they are apart. The possibility of the airstream going out through the mouth, as in [v] or [z], or the nose, as in [m] and [n], is determined by the oro-nasal process. The movements of the tongue and lips interacting with the roof of the mouth and the pharynx are part of the articulatory process. Figure 1.2 The vocal tract. 31269_01_Ch01_pp001-032 pp2.indd Sec2:5 10/29/09 4:57:14 PM Copyright 2010 Cengage Learning, Inc. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part
6 CHAPTER 1 Articulation and Acoustics Figure 1.3 The four main components of the speech mechanism. oro-nasal process articulatory process phonation process airstream process SOUND WAVES So far,we have been describing speech sounds by stating how they are made, but it is also possible to describe them in terms of what we can hear.The way in which we hear a sound depends on its acoustic structure.We want to be able to describe the acoustics of speech for many reasons(for more on acoustic phonet- ics,see Keith Johnson's book Acoustic and Auditory Phonetics).Linguists and speech pathologists need to understand how certain sounds become confused with one another.We can give better descriptions of some sounds (such as vow- els)by describing their acoustic structures rather than by describing the articu- latory movements involved.A knowledge of acoustic phonetics is also helpful for understanding how computers synthesize speech and how speech recognition works(topics that are addressed more fully in Peter Ladefoged's book Vowels and Consonants).Furthermore,often the only permanent data that we can get of a speech event is an audio recording,as it is often impossible to obtain movies or
6 CHAPTER 1 Articulation and Acoustics SOUND WAVES So far, we have been describing speech sounds by stating how they are made, but it is also possible to describe them in terms of what we can hear. The way in which we hear a sound depends on its acoustic structure. We want to be able to describe the acoustics of speech for many reasons (for more on acoustic phonetics, see Keith Johnson’s book Acoustic and Auditory Phonetics). Linguists and speech pathologists need to understand how certain sounds become confused with one another. We can give better descriptions of some sounds (such as vowels) by describing their acoustic structures rather than by describing the articulatory movements involved. A knowledge of acoustic phonetics is also helpful for understanding how computers synthesize speech and how speech recognition works (topics that are addressed more fully in Peter Ladefoged’s book Vowels and Consonants). Furthermore, often the only permanent data that we can get of a speech event is an audio recording, as it is often impossible to obtain movies or articulatory process phonation process oro-nasal process airstream process Figure 1.3 The four main components of the speech mechanism. 31269_01_Ch01_pp001-032 pp2.indd Sec2:6 10/29/09 4:57:14 PM Copyright 2010 Cengage Learning, Inc. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part
Sound Waves 7 x-rays showing what the speaker is doing.Accordingly,if we want permanent data that we can study,it will often have to come from analyzing an audio recording. Speech sounds,like other sounds,can differ from one another in three ways. They can be the same or different in(1)pitch,(2)loudness,and (3)quality.Thus, two vowel sounds may have exactly the same pitch in the sense that they are said on the same note on the musical scale,and they may have the same loudness,yet still may differ in that one might be the vowel in bad and the other the vowel in bud.On the other hand,they might have the same vowel quality but differ in that one was said on a higher pitch or that one of them was spoken more loudly. Sound consists of small variations in air pressure that occur very rapidly one after another.These variations are caused by actions of the speaker's vocal or- gans that are(for the most part)superimposed on the outgoing flow of lung air.Thus,in the case of voiced sounds,the vibrating vocal folds chop up the stream of lung air so that pulses of relatively high pressure alternate with mo- ments of lower pressure.Variations in air pressure in the form of sound waves move through the air somewhat like the ripples on a pond.When they reach the ear of a listener,they cause the eardrum to vibrate.A graph of a sound wave is very similar to a graph of the movements of the eardrum. The upper part of Figure 1.4 shows the variations in air pressure that occur during Peter Ladefoged's pronunciation of the word father.The ordinate (the vertical axis)represents air pressure(relative to the normal surrounding air pres- sure),and the abscissa (the horizontal axis)represents time(relative to an arbi- trary starting point).As you can see,this particular word took about 0.6 seconds to say.The lower part of the figure shows part of the first vowel in father.The major peaks in air pressure recur about every 0.01 seconds(that is,every one- hundredth of a second).This is because the vocal folds were vibrating approxi- mately one hundred times a second,producing a pulse of air every hundredth of a second.This part of the diagram shows the air pressure corresponding to four vibrations of the vocal folds.The smaller variations in air pressure that occur within each period of one-hundredth of a second are due to the way air vibrates when the vocal tract has the particular shape required for this vowel. In the upper part of Figure 1.4,which shows the waveform for the whole word father,the details of the variations in air pressure are not visible because the time scale is too compressed.All that can be seen are the near-vertical lines corresponding to the individual pulses of the vocal folds.The sound [f]at the beginning of the word father has a low amplitude (it is not very loud,so the pressure fluctuation is not much different from zero)in comparison with the fol- lowing vowel,and the variations in air pressure are smaller and more nearly ran- dom.There are no regular pulses because the vocal folds are not vibrating.We will be considering waveforms and their acoustic analysis in more detail later in this book.For the moment,we will simply notice the obvious difference be- tween sounds in which the vocal folds are vibrating(which have comparatively large regular pulses of air pressure)and sounds without vocal fold vibration (which have a smaller amplitude and irregular variations in air pressure)
Sound Waves 7 x-rays showing what the speaker is doing. Accordingly, if we want permanent data that we can study, it will often have to come from analyzing an audio recording. Speech sounds, like other sounds, can differ from one another in three ways. They can be the same or different in (1) pitch, (2) loudness, and (3) quality. Thus, two vowel sounds may have exactly the same pitch in the sense that they are said on the same note on the musical scale, and they may have the same loudness, yet still may differ in that one might be the vowel in bad and the other the vowel in bud. On the other hand, they might have the same vowel quality but differ in that one was said on a higher pitch or that one of them was spoken more loudly. Sound consists of small variations in air pressure that occur very rapidly one after another. These variations are caused by actions of the speaker’s vocal organs that are (for the most part) superimposed on the outgoing flow of lung air. Thus, in the case of voiced sounds, the vibrating vocal folds chop up the stream of lung air so that pulses of relatively high pressure alternate with moments of lower pressure. Variations in air pressure in the form of sound waves move through the air somewhat like the ripples on a pond. When they reach the ear of a listener, they cause the eardrum to vibrate. A graph of a sound wave is very similar to a graph of the movements of the eardrum. The upper part of Figure 1.4 shows the variations in air pressure that occur during Peter Ladefoged’s pronunciation of the word father. The ordinate (the vertical axis) represents air pressure (relative to the normal surrounding air pressure), and the abscissa (the horizontal axis) represents time (relative to an arbitrary starting point). As you can see, this particular word took about 0.6 seconds to say. The lower part of the figure shows part of the first vowel in father. The major peaks in air pressure recur about every 0.01 seconds (that is, every onehundredth of a second). This is because the vocal folds were vibrating approximately one hundred times a second, producing a pulse of air every hundredth of a second. This part of the diagram shows the air pressure corresponding to four vibrations of the vocal folds. The smaller variations in air pressure that occur within each period of one-hundredth of a second are due to the way air vibrates when the vocal tract has the particular shape required for this vowel. In the upper part of Figure 1.4, which shows the waveform for the whole word father, the details of the variations in air pressure are not visible because the time scale is too compressed. All that can be seen are the near-vertical lines corresponding to the individual pulses of the vocal folds. The sound [ f ] at the beginning of the word father has a low amplitude (it is not very loud, so the pressure fluctuation is not much different from zero) in comparison with the following vowel, and the variations in air pressure are smaller and more nearly random. There are no regular pulses because the vocal folds are not vibrating. We will be considering waveforms and their acoustic analysis in more detail later in this book. For the moment, we will simply notice the obvious difference between sounds in which the vocal folds are vibrating (which have comparatively large regular pulses of air pressure) and sounds without vocal fold vibration (which have a smaller amplitude and irregular variations in air pressure). 31269_01_Ch01_pp001-032 pp2.indd Sec2:7 10/29/09 4:57:15 PM Copyright 2010 Cengage Learning, Inc. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part
8 CHAPTER 1 Articulation and Acoustics Figure 1.4 The variations in air pressure that occur during Peter Ladefoged's pronunciation of the vowel in father. 0.0 0.2 0.4 0.6s T a th er expanded thisL」part expanded t 0.0 0.01 0.02 0.03 0.04s PLACES OF ARTICULATORY GESTURES The parts of the vocal tract that can be used to form sounds are called articulators.The articulators that form the lower surface of the vocal tract are highly mobile.They make the gestures required for speech by moving toward the articulators that form the upper surface.Try saying the word capital and note the major movements of your tongue and lips.You will find that the back of the tongue moves up to make contact with the roof of the mouth for the first sound and then comes down for the following vowel.The lips come together in the formation of p and then come apart again in the vowel.The tongue tip comes up for the t and again,for most people,for the final l. The names of the principal parts of the upper surface of the vocal tract are given in Figure 1.5.The upper lip and the upper teeth(notably the frontal inci- sors)are familiar-enough structures.Just behind the upper teeth is a small pro- tuberance that you can feel with the tip of the tongue.This is called the alveolar ridge.You can also feel that the front part of the roof of the mouth is formed
8 CHAPTER 1 Articulation and Acoustics PLACES OF ARTICULATORY GESTURES The parts of the vocal tract that can be used to form sounds are called articulators. The articulators that form the lower surface of the vocal tract are highly mobile. They make the gestures required for speech by moving toward the articulators that form the upper surface. Try saying the word capital and note the major movements of your tongue and lips. You will find that the back of the tongue moves up to make contact with the roof of the mouth for the first sound and then comes down for the following vowel. The lips come together in the formation of p and then come apart again in the vowel. The tongue tip comes up for the t and again, for most people, for the final l. The names of the principal parts of the upper surface of the vocal tract are given in Figure 1.5. The upper lip and the upper teeth (notably the frontal incisors) are familiar-enough structures. Just behind the upper teeth is a small protuberance that you can feel with the tip of the tongue. This is called the alveolar ridge. You can also feel that the front part of the roof of the mouth is formed 0.0 0.01 0.02 0.03 0.04 s expanded expanded 0.0 0.2 0.4 0.6 s f a th er this part Figure 1.4 The variations in air pressure that occur during Peter Ladefoged’s pronunciation of the vowel in father. 31269_01_Ch01_pp001-032 pp2.indd Sec2:8 10/29/09 4:57:15 PM Copyright 2010 Cengage Learning, Inc. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part
Places of Articulatory Gestures 9 Figure 1.5 The principal parts of the upper surface of the vocal tract. -soft palate hard palate (velum) teeth alveolar- uvula ridge Dharynx wall by a bony structure.This is the hard palate.You will probably have to use a fingertip to feel farther back.Most people cannot curl the tongue up far enough to touch the soft palate,or velum,at the back of the mouth.The soft palate is a muscular flap that can be raised to press against the back wall of the pharynx and shut off the nasal tract,preventing air from going out through the nose.In this case,there is said to be a velic closure.This action separates the nasal tract from the oral tract so that the air can go out only through the mouth.At the lower end of the soft palate is a small appendage hanging down that is known as the uvula. The part of the vocal tract between the uvula and the larynx is the pharynx. The back wall of the pharynx may be considered one of the articulators on the upper surface of the vocal tract. Figure 1.6 shows the lower lip and the specific names for the parts of the tongue that form the lower surface of the vocal tract.The tip and blade of the tongue are the most mobile parts.Behind the blade is what is technically called the front of the tongue;it is actually the forward part of the body of the tongue and lies underneath the hard palate when the tongue is at rest.The remainder of the body of the tongue may be divided into the center,which is partly beneath the hard palate and partly beneath the soft palate;the back,which is beneath the soft palate;and the root,which is opposite the back wall of the pharynx.The epiglottis is attached to the lower part of the root of the tongue. Bearing all these terms in mind,say the word peculiar and try to give a rough description of the gestures made by the vocal organs during the consonant sounds.You should find that the lips come together for the first sound.Then the back and center of the tongue are raised.But is the contact on the hard palate or on the velum?(For most people,it is centered between the two.)Then note the position in the formation of the l.Most people make this sound with the tip of the tongue on the alveolar ridge
Places of Articulatory Gestures 9 by a bony structure. This is the hard palate. You will probably have to use a fingertip to feel farther back. Most people cannot curl the tongue up far enough to touch the soft palate, or velum, at the back of the mouth. The soft palate is a muscular flap that can be raised to press against the back wall of the pharynx and shut off the nasal tract, preventing air from going out through the nose. In this case, there is said to be a velic closure. This action separates the nasal tract from the oral tract so that the air can go out only through the mouth. At the lower end of the soft palate is a small appendage hanging down that is known as the uvula. The part of the vocal tract between the uvula and the larynx is the pharynx. The back wall of the pharynx may be considered one of the articulators on the upper surface of the vocal tract. Figure 1.6 shows the lower lip and the specific names for the parts of the tongue that form the lower surface of the vocal tract. The tip and blade of the tongue are the most mobile parts. Behind the blade is what is technically called the front of the tongue; it is actually the forward part of the body of the tongue and lies underneath the hard palate when the tongue is at rest. The remainder of the body of the tongue may be divided into the center, which is partly beneath the hard palate and partly beneath the soft palate; the back, which is beneath the soft palate; and the root, which is opposite the back wall of the pharynx. The epiglottis is attached to the lower part of the root of the tongue. Bearing all these terms in mind, say the word peculiar and try to give a rough description of the gestures made by the vocal organs during the consonant sounds. You should find that the lips come together for the first sound. Then the back and center of the tongue are raised. But is the contact on the hard palate or on the velum? (For most people, it is centered between the two.) Then note the position in the formation of the l. Most people make this sound with the tip of the tongue on the alveolar ridge. Figure 1.5 The principal parts of the upper surface of the vocal tract. 31269_01_Ch01_pp001-032 pp2.indd Sec2:9 10/29/09 4:57:15 PM Copyright 2010 Cengage Learning, Inc. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part