Manner of articulation

We have described a number of classes of consonants according to their place of articulation, yet we are still unable to distinguish the sounds in each class from each other. What distinguishes [p] from [b], or from [m] ? all are bilabial sounds. What is the difference between [t], [d], and [n], which are all alveolar sounds ?

Speech sounds are also differentiated by the way the airstream is affected sa it travels from the lungs up and out of the mouth and nose. Such features or phonetic properties have traditionally been referred to as manners of articulation or simply manner features.

Voiced and voiceless sounds

If the vocal cords are apart when the airstream is pushed from the lungs, the air is not obstructed at the glottis, and it passes freely into the supraglottal cavities. The sounds produced in this way are voiceless sounds: [p], [t], [k] and [s] in the English words seep [sip], seat [sit] and seek [sik] are voiceless sounds.

If the vocal cords are together, the airstream forces its way through and causes them to vibrate. Such sounds are called voiced sounds are illustrated by the sounds [b], [d], [g], and [z] in the words bad [bæd], dog [dɒg], and buzz [bʌz]. If you put a finger in each ear and say ‘z-z-z-z-z’ you can feel the vibrations (although you hear a hissing sound in your mouth). When you whisper, you are making all the speech sounds voiceless.

The voiced/voiceless distinction is a very important one in English. It is this phonetic feature or property that distinguishes between word pair such as:

rope/robe   [roʊp] / [roʊb]

fate/fade    [feɪt] / [feɪd]

wreath/wreath [riθ] / [rið]

The first word of each pair ends with a voiceless sound and the second word with a voiced sound. All other aspects of the sounds of these words are identical; the position of the lips and tongue is the same in each of the paired words.

The voiced/voiceless distinction is also shown in the following pairs; the first word begins with a voiceless sound and the second with a voiced sounds :

fine/vine

[faɪn] / [vaɪn]

seal/zeal

[sil] / [sil]

choke/joke

[tʃoʊk] / [dʒoʊk]

The initial sounds of the first words of the following pairs are also voiceless, and for many speakers of English, the second words begin with voiced sounds.

peat/beat

[pit] / [bit]

tin / din

[tɪn] / [dɪn]

cane/gain

[keɪn] / [geɪn]

Aspirated and unaspirated sounds

In our discussion of the bilabial stop [p], we did not distinguish the initial sound in the word pit from the second sound in the word split. There is, however, a phonetic difference in these two stops. During the production of voiceless sounds the glottis is open and the air passes freely through the opening between the vocal cords must close in order to permit them to vibrate.

Voiceless sounds fall into two classes depending on the ‘timing’ of the vocal cord closure. In English when we pronounce the word pit, there is a brief period of voicelessness immediately after the p sound is released. That is, after the lips come apart the vocal cords remain open for a very short time. Such sounds are called aspirated because an extra puff or air is produced.

When we pronounce the p in spit, however, the vocal cords start vibrating as soon as the lips are opened. Such sounds are called unaspirated. The t in stick and the k in skin are unaspirated. If you hold a strip of paper in front of your lips and say pit, a puff of air (the aspiration0 will push the paper. The paper will not move when you say spit.

When a fully voiced [b] is produced, the vocal cords vibrate throughuot the articulation. In English, voiced stops may not be voiced.

Figure 6.2 shows in diagrammatic form the timing of the articulators (in this case the lips) in relation two the state of the vocal cords. Notice that in production of the voiced [b], the vocal cords are vibrating throughout the closure of the lips and continue to vibrate for the vowel production after the lips are opened. Most English speakers do not voice initial [b] to the full extent. Because we heavily aspirate an initial [p], there is no difficulty in distinguishing these two sounds. In the unaspirated p in spin, the vocal cords are open during the lip closure and come together and start vibrating as soon as the lips open. In the production of the aspirated p in pin the vocal cords remain

Apart for a brief period after the lip closure is released. These remarks apply to all English stops.

Aspirated sounds may be indicated by following the phonetic symbol with a raised h, as in the following examples :

pool

[phul]

tool

[thul]

cool

[khul]

spool

[spul]

stool

[stul]

school

[skul]

Nasal and oral sounds

The voiced/voiceless distinction differentiates the bilabials [b] and [p]. The sound [m] is also a bilabial and, in addition, it is voiced. What, then, distinguishes the [m] from the [b] ?

[m] is a nasal sound. When you produce [m], air escapes not only through the mouth (when you open your lips) but also through the nose.

In figure 6.1, the roof of the mouth is divided into the palate and the soft palate (or velum). The palate is hard bony structure at the front of the mouth. You can fell it with your thumb. As you slide your thumb back toward the throat you will feel the velum, which is where the flesh becomes soft and movable. Hanging down from the end of the velum is the uvula, which you can see in a mirror if you open your mouth wide and say ‘aaah’. When the velum is raised all the way to touch the back of the throat, the passage through the nose is cut off and air can escape only through the mouth.

Sound produced with the velum up, blocking the air from escaping through the nose, are called oral sounds, since the air can only escape through the oral cavity. When the velum is lowered, air escapes through the nose as well as the mouth : sounds produced this way are called nasal sounds. [m], [n], and [ŋ] are the nasal consonants of English. The diagrams in figure 6.3 show the position of the lips and the velum when [m], [p], and [b] are articulated.

The same nasal/oral difference occurs in beet [bit] and meat [mit], dear [dɪə] and near [nɪə]. The velum is raised in the production of [b] and [d], preventing the air from flowing through the nose, whereas in [m] and [n] the velum is down, letting the air go through both the nose and the mouth when the closure is released. [m], [n], and [ŋ] are therefore nasal sounds and [b], [d], and [g] are oral sounds.

These phonetic features or properties permit the classification of speech sounds into four classes; voiced,voiceless,nasal,and oral, in addition to th place-of-articulation classes discussed earlier. One sound may belong to more than one class, as shown in table 6.3.

Stops : [p] [b] [m] [t] [d] [n] [k] [g] [ŋ] [tʃ] [dʒ] [q] [G]

We are seeing finer and finer distinctions of speech sounds. [t] is a voiceless, alveolar, oral sound. But [s] is also voiceless and alveolar and oral. What is distinguishes [t] from [s] ?

In producing sounds, the airstream after entering the oral cavity be stopped (as in the production of [t] ), or partially obstructed (as in the articulation of [s]), or it may flow freely out of the mouth. Sounds that are stopped completely in the oral cavity for a brief period are, not surprising, called stops.

The final sounds in the words tap [tæp], tab [tæb], tam [tæm], tat [tæt], tad [tæd], tan [tæn], tack [tæk], tag [tæg], and tang [tæŋ] are stops that occur in English.

In the production of the nasal stops [n], [m], and [ŋ], although the air flows freely through the nose, the airflow is blocked completely in the mouth. Therefore, nasal consonants are stops.

Sound in which there is no stoppage in the oral tract are continuants. All the sounds of a language are either stops or continuants (non-stops).

Non-nasal or oral stops are also called plosives because the air that is blocked in the mouth ‘explodes’ when the closure is released. This explosion does not occur during the production of the nasal stops because the air has an ‘escape route’ through the nose.

[p], [b], [m] are bilabial stops, with the airstream stopped at the mouth by the complete closure of the lips.

[t], [d], and [n] are alveolar stops; the airstream is stopped by the tongue making a complete closure at the alveolar ridge.

[k], [g], and [ŋ] are velar stops with the complete at the velum.

[tʃ] and [dʒ] are alveopalatal or palatal affricates with complete stop closure.

Although there is no stoppage of air in the oral cavity, the air is completely stopped at the glottis in the production of the glottal stop [ʔ]; for this reason, some linguists classify it as a stop.

We have been discussing the sounds that occur in English. There are sounds, including stops, that occur in other languages but are not found in English. In Quechua, a major language spoken in Bolivia and Peru, uvular stops occur. These are produced when the back of the tongue is raised and moved backward to form a complete closure with the uvula. The letter q in words in this language, as in the language name, usually represents the uvular stop phonetically symbolised as [q]. The voiced uvular stop symbolised as [g] also occurs in Quechua. As noted earlier, glottal stops are also occur in a number of languages, such as Lebanese Arabic.

Fricatives: [s] [z] [f] [v] [θ] [ð] [ʃ] [ʒ] [x] [ɣ] [ʁ]

In the production of some continuants, the airstream is not completely stopped but obstructed from flowing freely. If you put your hand in front of your mouth and produce an [s], [z], [f], [v], [θ], [ð], [ʃ], or [ʒ] sound, you will feel the air coming out of your mouth. The passage in the mouth through which the air must pass, however is very narrow causing friction or turbulence. Such sound are called fricatives. (They are also sometimes referred to as spirants, from the Latin word spirare, ‘to blow’.)

In the production of labiodental fricatives [f] and [v], the friction is created at the lips, where a narrow passage permits the air to escape.

[s] and [z] are alveolar fricatives with the friction created at the alveolar ridge. The palatal or alveopalatal fricatives, [ʃ] and [ʒ], such as those in mesher [meʃə] and measure [meʒə], are produced with friction created as the air passes through the narrow opening behind the alveolar ridge. In English, the voiced palatal fricatives never begins words (except in words borrowed from French such as genre or gendarme which some English speakers produce with a French pronunciation). The voiceless palatal sound begins the words shoe [ʃu] nd sure [ʃʊə, ʃɔ] and ends the words rush [rʌʃ] and push [pʊʃ].

In the production of the interdental fricatives [θ] and [ð], represented by th in thin and then, the friction occurs at the opening between the tongue and teeth.

Most dialects of modern English do not include velar fricatives, although they occurred in an earlier stage of English in such words as right, knight, enough, and through, where the gh occur in the spelling. If you raise the back of the tongue as if you were to produce a [g] or [k], but stop just short of touching the velum, you will produce a velar fricative. The ch ending in German pronunciation of the composer’s name, Bach, is a velar fricative. Some speakers of modern English substitute a voiceless velar fricative in words such as bucket and a voiced velar fricative in such words as wagon for the velar stops that occur for other speakers in those words. [x] is the IPA symbol for the voiceless velar fricative and [ɣ] for the voiced velar fricative.

In some languages of the world, such as French, the uvular fricative [ʁ] occurs as the sound represented by r in French words such as rouge, ‘red’, or rose, ‘pink’. Voiced glottal fricatives, which do not occur in English, do occur in other languages. In Arabic pharyngeal fricatives are produced by pulling the tongue root towards the back wall of the pharynx. It is difficult to pull the tongue far enough to make a complete pharynx stop closure, but both voiced and voiceless pharyngeal fricatives can be produced and can be distinguished from velar fricatives.

All fricatives are continuants; although the airstream is obstructed as it passes through the oral cavity, it is not completely stopped.

Affricates

Some sounds are produced by a stop closure followed immediately by a slow release of the closure characteristic of a fricative. These sounds are called affricates. The alveopalatal sounds that begin and end the words church and judge are voiceless and voiced affricates, respectively. Phonetically, an affricate may be considered a sequence of a stop plus a fricative. Thus, the ch in church is the same as the sound combination [t] + [ʃ] as shown by observing that in fast speech white shoes and why choose may be pronounced identically. The voiceless and voiced affricates may be symbolised as [tʃ] and [dʒ], respectively. (In the American tradition, [č] and [ǰ] are the more commonly used symbols for these sounds.)

Because the air is stopped completely during the initial articulation of an affricate, these sounds are non-continuant and classified as stops.

Laterals : [l]

In the production of the sound [l], the front of the tongue makes contact with the alveolar ridge, but the sides of the tongue are down, permitting the air to escape laterally over the sides of the tongue. It is usually referred to as a lateral sound.

There is some obstruction of the airstream in the mouth but not enough to cause friction. Some linguists use this degree of obstruction to define a class of sounds called liquids which we discuss further in the following section. The sound [l] is the only clear example of this class in Australian English.

In some dialects of English (notably in British, but also in Australian to some degree), [l] is strongly affected by an adjacent vowel. When [l] follows a vowel (as in fell), or when it acts as a syllable on its own (as in bottle), the bulk of the tongue is raised towards the velum. This is often said to result in a ‘dark’ quality to the sound, and the variety is known as dark or velarised ɭ [ɫ]. When ɭ precedes a vowel (as in leaf), the bulk of the tongue drops sharply from the alveolar ridge, and the variety is known as clear ɭ.

In English, ɭ is regularly voiced. When it follows voiceless sounds, as in please, it may be automatically ‘devoiced’, at least partially. (This substitutes for the aspiration which we would normally expect on an initial p.) Many languages of the world have a voiceless ɭ. Welsh  is such a language; the name Lloyd in Welsh starts with a voiceless ɭ.

Glides : [j] [w] [r] [h] [ʔ]

The sound [j] and [w], the initial sound of you [ju] and woo [wu], are produced with little or no obstruction of the airstream in the mouth. When occurring in a word, they must always be followed directly by a vowel. In articulating [j] or [w] the tongue moves rapidly in gliding fashion either towards or away from a neighbouring vowel; hence the term glide. Glides are transition sounds that are sometimes called semivowels.

[j] is a palatal glide; the blade of the tongue is raised toward the hard palate in a position almost identical to that in producing the vowel sound [i] in the word beat [bit]. In pronouncing you [ju], the tongue moves rapidly from the [j] to the [u] vowel.

The glide [w] is produced by both raising the back of the tongue toward the velum and simultaneously rounding the lips. It is therefore a labio-velar glide, or a rounded velar glide. In dialects of English whose speakers have different pronunciations for the words which and witch, the velar glide in the first word is voiceless [M] (an ’upside-down’ w), and in the second word it is voiced [w]. The position of the tongue and lips for [w] is similar to that for producing the vowel sound in [u] lute, but the [w] is a glide because the tongue moves quickly to the vowel that follows.

As mentioned earlier, the r sounds that occur in various dialects of English and other languages differ somewhat from each other. We are using the symbol [r] for this whole class of sounds, most of which sare many of the characteristics of [w] and [j]. The most usual variant in Australian and New Zealand English is more like a vowel than anything else. The tongue tip is raised to just behind the alveolar ridge, but produces little obstruction to the airstream; [r] is therefore an alveolar glide. It is also like [j] and [w] in that it must always be followed directly by a vowel, as in red [red], bred [bred], and berry [beri]. The r that is written in for and fort is not pronounced (unless a combination of words brings a vowel into place, as in for a while). On the other hand, [r] is acoustically (that is, as a physical sound) very similar to [l]. For this reason r and l are often classed together as liquids, and they function as asingle class of sounds in certain circumstances. For example, the only two consonants permitted after an initial [k, g, p], or [b] in English are the liquids [l] and [r]. Thus we have plate [pleɪt], prate [preɪt], bland [blænd], and brand [brænd] but no word starting with ps, bt, pk, and so on. (Notice that in words as psychology or pterodactyl the p is not pronounced. Similarly in knight or knot the k is not pronounced, although at an earlier stage of English it was.)

Some language may lack liquids entirely, or have only a single one. The Cantonese dialect of Chinese has a single liquid [l]. Japanese, on the other hand, lacks an [l] but has an [r]. These differences make some English words difficult for Cantonese or Japanese speakers to pronounce. Other differences make these languages difficult for English speaker to pronounce.

Like, [l], [r] is very sensitive to both regional variation and varying phonetic context. In many dialects of English [r] is not subject to the restriction that it must immediately precede a vowel. In American English, for example, both for and fort would contain the sound [r]. It is also usual in American dialect to produce the [r] by curling the tip of the tongue back behind the alveolar ridge. Such sounds are called retroflex sounds. In some languages the r may be a trill, which is produced by the tip of the tongue vibrating against the roof of the mouth. It is possible that in an earlier stage of English the r was a trill. A trilled r occurs in many contemporary languages, such as Spanish, and is common is Scots English.

In addition to the alveolar trill, uvular trills also occur, as in French. A uvular trill is produced by vibrating the uvula. In other languages the r is produced by a single tap instead of a series of vibrating taps. In Spanish both the alveolar trill and the alveolar tap occur. If you substitute one for the other in certain contexts you will produce a different word. One may also produce an r by making the tongue flap against the alveolar ridge, as in a common pronunciation of the word three. Some speakers of British English pronounce the r in the word very with a flap. It sounds like a very fast d. Most American speakers, and some Australians, produce a flap instead of a [t] or [d] in words such as writer and rider, and latter and ladder. For many speakers these pairs are pronounced identically in normal conversational style. The IPA symbol for the alveolar tap or flap is [ɾ]. American linguists often use the symbol [D] to represent this sound.

After [t] and [d], a fricative version of r is normally produced. In the words train and drain, the r is represented by voiceless and voiced fricatives, respectively.

The [h] that starts words such as house [haʊs], who [hu], and hair [heə] is also a glide. The glottis is open as in the production of voiceless sounds. No other modification of the airstream mechanisms occurs in the mouth. In fact the tongue and lips are usually in the position for the production of the following vowel as the airstream passes through the open glottis. The air or noise produced at the glottis is heard as [h] and for this reason, it is sometimes classified as a voiceless glottal fricative. However, the [h] differs from ‘true’ consonants in that there is no obstruction in the oral cavity. It also differs from vowels, which are articulated by moving the tongue. When it is both preceded and followed by a vowel in English, as in ahead [əhed] and cohabit [koʊhæbət], it is often voiced.

The glottal stop [ʔ], like [h], differs from both consonants and vowels and is commonly classified as a glide. But because the air is completely blocked at the glottis, some linguists classify it as a stop.