Basics of Articulation (PDF)

2 Basics of Articulation Manner and Place in English Don’t speak of letters when you mean sounds. Languages are made up of sounds. Daniel Jones, The Sechuana Reader, 1916, p. xxxvi (emphases original) Chapter outline 2.1 Th...

2 Basics of Articulation Manner and Place in English Don’t speak of letters when you mean sounds. Languages are made up of sounds. Daniel Jones, The Sechuana Reader, 1916, p. xxxvi (emphases original) Chapter outline 2.1 The dance of the articulators 15 2.2 Phonetic transcription 16 2.3 The building blocks of speech 20 2.3.1 Airstream, larynx, and velum 20 2.3.2 Manner of articulation 21 2.3.3 Place of articulation for consonants 24 2.3.4 Vowels 26 Chapter summary 29 Further reading 29 Review exercises 30 Further analysis and discussion 32 Further research 32 Go online 32 The Sounds of Language: An Introduction to Phonetics and Phonology, First Edition. Elizabeth C. Zsiga. © 2013 Elizabeth C. Zsiga. Published 2013 by Blackwell Publishing Ltd.  Basics of Articulation   15 Daniel Jones is correct: (Spoken) languages are made up of sounds. In this chapter, we con- sider first how sounds combine to create linguistic messages, and then how articulator movements combine to create specific sounds. Section 2.2 considers the tricky problem of how to write down sounds in a systematic and unambiguous way. Section 2.3, the bulk of the chapter, describes the vocal tract movements and configurations that act as “building blocks” to create the sounds of English. 2.1 the dance of the articulators It is a basic tenet of linguistics that human language is discrete, combinatorial, and unbounded. That is, the stream of language can be broken down into smaller, individual pieces, and these pieces (which may in themselves mean nothing) can be combined in different ways to create an infinite number of messages. Thus, human language is like a set of Lego bricks that can be combined into a boat or a barn or a starship, depending on the plan of the builder. At the level of sentence structure, every language user knows a set of individual words that can be combined in new and different ways to create an infinite number of sentences. Or, through the property of recursion, which permits embedding one sentence inside another, a speaker can create sentences that are indefinitely long, with no limits except the practical ones of memory and patience: “Mary said that John said that Sally said that her mother said... that Robin likes Joey.” At the level of sound structure, a dictionary full of words is made up of different com- binations of discrete sounds, chosen from a limited inventory of vowels and consonants, which by themselves mean nothing. The set of words known by any individual is large but not infinite (usually estimated between about 40,000 and 300,000, depending on experience and education.) However, recursion can sometimes apply at the word level as well, as in great-great-great-great-grandmother or re-re-re-re-re-apply. In addition, the set of words is open-ended: new words are created every day. If I told you about a delicious new fruit that combined the tastes of a mango and banana, you’d have no problem adding “mananas” to your grocery list, and thus your vocabulary. Finally, we will see that every vowel and consonant is made up of a combination of vocal tract movements we will call articulatory gestures, which combine to create a particular sound. These properties (discrete, combinatorial, unbounded, recursive) set human language apart from every other animal communication system. While birdsongs or dolphin calls may be complex and meaningful, they do not, so far as we know, combine individual meaningless pieces in different ways to create an open-ended number of messages, and no clear example of recursion has been discovered in animal communication. Those of us familiar with English spelling probably accept the ideas of “speech sound” or “vowel” and “consonant” without thinking, because we are used to seeing words repre- sented as a sequence of letters. However, the idea that speech can be broken down into discrete elements is far from obvious. As a physical object in the world, speech exists as a continuous stream. As a person is speaking, the lips, tongue, larynx, lungs – all the vocal organs – are in constant, continuous motion. The lips open and close, the tongue body slides forward and back, the tongue tip flips up and down, all without stopping. Words, and the vowels and consonants that make them up, are run together and overlapped. We don’t pause between sounds as though each one was a letter in a spelling bee. This running together is especially obvious when you’re listening to a language you don’t know. When you’re not able to associate meanings with the different chunks of speech, you cannot tell where one word stops and the next begins. 16   Basics of Articulation You might think of speech movements as a dance of the articulators. As you watch a skilled dancer move across the floor, you see her feet, legs, arms, and head constantly and fluidly moving: dancing does not consist of jumping from one pose to the next. An unskilled onlooker would have trouble picking out the individual steps. Only if you saw the dance being choreographed would you know that you were watching a sequence of discrete steps that were learned separately and then put together. The sounds of speech are like the steps in a dance. Just as each dance step may combine movements of different parts of the body, so each sound of speech combines movements of different parts of the vocal tract. Just as dancing combines a set of steps in a sequence, but does not consist of holding one pose and then jumping to the next, so speech combines a set of sounds in sequence, but does so fluidly, with no obvious breaks between. And just as a dancer who knows his dance well can execute the movements from “muscle memory” without thinking “left foot here, right foot there” at every step, so humans who are proficient language users execute the movements of speech without any conscious thought at all about the vocal tract movements they are executing. The movements that create each step of the dance of the articulators are articulatory gestures. In order to define a dance step, you would specify which part of the body to move (left foot, right arm, etc.) and then specify the position to which it moves (left foot slides back, right arm raised over head). In the same way, to define a vocal tract gesture, you choose the articulator, and specify the position, or goal, toward which it moves. In most cases, for speech, this goal is a constriction in some part of the vocal tract, which will result in some particular sound. When we define the position to which the articulator moves (the combina- tion of active articulator and passive articulator), we are defining the place of articulation. When we define the type of constriction that is made (complete stoppage of air vs. narrow channel, for instance) we are defining the manner of articulation. In addition to the place and manner of articulation of a vocal tract constriction, the definition of a speech sound will also include the states of the velum and larynx, as well as the method of getting the air to move. Every speech sound, then, is defined by a particular combination of the following components: airstream mechanism state of the larynx state of the velar port combination of active and passive articulator (= place of articulation) manner of articulation This chapter discusses each of these in turn. Before we can begin to talk about specific sounds, however, we need to deal with an important problem: how can sounds be written down? 2.2 phonetic transcription As students of speech, we need to consider how we can most simply and effectively symbolize the sounds that are the object of our study. A system of symbols for sounds is a phonetic alphabet. Writing down sounds using a phonetic alphabet is called phonetic transcription. Since this is a textbook, I need to be able to communicate to you which speech sound I am describing, without having the opportunity to actually pronounce it for you. So far, I have been able to get away with using example sounds for which the sound–spelling correspond- ence is reasonably unambiguous: we all know what sound the letter “m” makes, or “b.” Or I  Basics of Articulation   17 can cite English words, and say “the sound at the beginning of the word ‘mall,’” and you can say it to yourself and know what I mean. But this won’t get us very far. 2.1 The relationship between language and writing would be a course in itself. Some writing systems do not represent sounds at all: Chinese logograms, for example, represent word meanings rather than pronunciations. This is a very useful property, given that many of the language varieties that we group under the label “Chinese” are so different as to be mutually unintelligible. But residents of Hong Kong in the south and Beijing in the north, for example, who might not be able to understand each other speak, can all read the same newspaper. Many other writing systems (such as Japanese and Korean) use symbols that correspond to syllables. For languages that use an alphabet (English, Spanish, Greek, Russian, Arabic, etc.), the correspondence between sounds and letters may be more or less straightforward (as in Spanish), or to a large extent arbitrary (English). When English dictionaries first began to appear about 500 years ago, the dictionary writers did their best to write words the way they sounded, and thus there was a much better cor- respondence between sound and spelling than there is now. For example, “made” really was pronounced “ma-deh” and there really was a “k”-sound and “gh”-sound in “knight.” But pronunciation has changed a lot in half a millennium, and spelling has not kept up. What with “silent” letters like “e” in “made” and “k” in “knight,” words like “read” that can be pronounced either “red” or “reed,” and incomprehensible sets like “through,” “though,” “thought,” and “tough,” English spelling is almost as much a matter of memorization as it is sound–letter correspondence. If English words were spelled the way they sound, elemen- tary school children would not have to study lists of spelling words every week, and spelling bees would be no challenge at all. Even if we pick “easy” spelling words, however, another problem is that different people pronounce words in different ways. In English, this is especially problematic for the pronun- ciation of vowels. If I wanted to talk about “the vowel in the word ‘bite,’” you would have to ask if the bite occurred in Atlanta, Baltimore or Seattle (not to mention London or Perth) before you could know which vowel was meant. (Variation between dialects can become a real political issue for languages that are being written down for the first time. Even if the sound–letter correspondence is transparent, which group’s pronunciation gets to be the “official” one?) Finally, a complete system of phonetic transcription cannot be limited to the sounds of any one language. We do not have unambiguous letters in the Roman alphabet for the clicks in some Southern African languages or the sounds that are made in the back of the throat in Arabic. For the purpose of spelling foreign names, we use the closest English equivalent, or letters that are not much used otherwise, like q and x. But sometimes it is not clear which letter is the closest: is it Peking or Beijing, Bombay or Mumbai? (Answer: neither, really: it’s a non-English sound for which we do not have an obvious spelling.) The sound spelled with an X in the name of the African language Xhosa is nothing like the sound spelled with an X in the Chinese family name Xu. (The former is a click, the latter, a sound similar to “sh.” If you are limited to English spelling, it is hard to be any more specific.) In 1887, a group of phoneticians, the International Phonetic Association, tackled the problem of how to describe precisely and unambiguously any sound they might encounter in their efforts to describe all the languages of the world. They decided to create a new alphabet, the International Phonetic Alphabet (IPA), based on a set of principles, two of the most central being: 18   Basics of Articulation 1. The alphabet would be universal. There would be enough symbols so that every sound in every human language could be represented. 2. The alphabet would be unambiguous. There would be a one-to-one correspondence between sounds and symbols: every sound would have one symbol, and every symbol would stand for only one sound. English spelling frequently falls short on both these principles. It was noted above that there are many sounds for which we do not have symbols. The IPA currently contains 83 unique symbols for consonants and 28 symbols for vowels. The Roman alphabet, of course, has 21 and 5. (Don’t worry too much about memorizing every symbol in the IPA right away. That is why there is a chart in this book to look them up (Figure 3.1). Plus, there are regularities that make learning the IPA somewhat easier than you think.) As new sounds have been discovered, the IPA has been revised to include new symbols for them. The latest revision, as of this writing, took place in 2005. The new sound that prompted the revision is described in Chapter 3. 2.2 In Focus In English spelling, the one sound/one symbol correspondence is constantly violated. There are, for example, at least four different ways to pronounce the letter “c,” as in “each vicious circle” (five if you count “silent c” as in “back”). The “s-sound” can be spelled with either “s” as in “sent” or “c” as in “cent,” a combination of the two, as in “scent,” or with a double “s” (“assent”). The letter “x” stands for a sequence of two sounds, [k] followed by [s] (“box” rhymes with “locks”), while the sequence of letters “ph” stands for one sound, the same as “f.” The English system is not totally chaotic. Various rules can be used to figure out whether a “c” should be “hard” or “soft,” for example. And it can be useful to have different spellings for words that sound the same: the sentence “I was looking for a big sail” is ambiguous in speech but not in writing. But the level of ambiguity, randomness, and mismatches is high enough to keep the system of English spelling from being useful as a clear way to refer to sounds, even for English alone, or even if only one dialect was being considered. Numerous different phonetic alphabets have been developed, with different purposes in mind. Some were developed for a single language, with the goal of writing the language down and teaching people to read. For this goal, simplicity is probably the most important crite- rion. At the time when many of these systems were created, the set of symbols was con- strained to those available on typewriter keys. Other systems are meant to be used in dictionaries, or for the purpose of teaching a foreign language, for which purposes keeping the phonetic alphabet close to the “spelling alphabet” might be most important. Tradition often plays an important role. If you grew up learning about “long e” and “short e,” or if scholars have written words a certain way for hundreds of years, a phonetic alphabet that represents sounds in the same ways may be preferred. There is no one correct system to use for every case. However, for the purposes of the general description of all the sounds of the many lan- guages of the world, the principles of the International Phonetic Alphabet best meet the need  Basics of Articulation   19 and are most widely accepted in the linguistic community, and thus this book will use IPA. Where symbols from other transcription systems are commonly in use, these will be noted. We begin, in this chapter, with the sounds of English. This is not because English has some special linguistic status among the languages of the world – like every other language, English has some properties that are pretty common and some others that are more unusual – but because it is the one language that every reader of this book is familiar with. You will find that most, if not quite all, of the IPA symbols for English sounds are already familiar, being based on their most common values in the Roman alphabet. Table 2.1 gives the IPA symbols for the consonants of English, along with example words. Note that only the basic symbols, enough to distinguish different English words, are given in Table 2.1. This basic style of transcription is known as broad transcription. Various addi- tional articulatory details or modifications, which are often indicated by diacritical marks added to the basic symbol, are discussed in the text. A style of transcription that includes Table 2.1 IPA symbols for the consonants of English. Initial Final Medial Alternate symbol p pat pie pen pin whip upper b bat buy Ben bin bib rubber m mat my men minion whim summer f fat fight fen fin whiff suffer v vat vie vendor vintage wave ever θ thigh thin with Ethel ð that thy then bathe weather t tat tie ten tin wit retool d data dye den din mid redo n Nat night ninja win renew s sat sigh sensor sin miss presser z zap zen zip wiz buzzer l lateral lie lentil lip will filler r rat rye rent rip where terror ∫ shack shy shell ship wish pressure Š beige measure ž chat chai check chip witch etcher č jack giant gender gin wedge edger ȷ̌ k cat kite Ken kin wick wrecker g gap guy gecko wig mugger ŋ wing singer h hat high hen hip ahead w whack why when win away j yak yen yip y 20   Basics of Articulation Table 2.2 A simplified IPA chart for the consonants of English. bilabial labio-dental dental alveolar post-alveolar palatal velar glottal plosive pb td kg fricative fv θð sz h affricate nasal m n ŋ approximant (w) l r j (w) these additional details is known as narrow transcription. Neither narrow nor broad tran- scription is better or more correct; they are just more or less detailed. Broad transcription focuses on aspects of pronunciation that are contrastive: that is, that make a difference in the meaning of a word (like the difference between [p] and [b] in “pit” vs. “bit.” Narrow tran- scription includes both contrastive and non-contrastive details. Table 2.2 provides the same symbols in the form of a chart. In an IPA chart, place of articulation is written across the top, manner down the side. Thus you can see that any given sound is a combination of a certain place and a certain manner, and place and manner combine relatively freely. (Additional place and manner combinations are shown in the full IPA chart in Figure 3.1.) If there are two symbols in a cell, the one on the left is voiceless, the one on the right is voiced. All these symbols may seem like a lot to learn at first, but in the long run it’s a lot easier than trying to describe sounds using English spelling or descrip- tion. As you learn the definitions of the different places and manners of articulation, you should be able to combine them to figure out the pronunciations of unfamiliar symbols in the chart, even if you cannot yet produce them perfectly. The next section of the chapter now turns to describing the sounds in the chart, based on their articulatory characteristics. Section 2.3.1 addresses airstream mechanism, laryngeal state, and nasality. Section 2.3.2 discusses manner, and Section 2.3.3, the longest and most detailed, describes place. 2.3 the building blocks of speech 2.3.1 airstream, larynx, and velum All speech sound is the movement of air made audible. Therefore, the first defining compo- nent of a speech sound is that of airstream mechanism: How will we get the air moving in the first place? As noted in Chapter 1, for English, this choice is invariably pulmonic egres- sive, that is, air moving out from the lungs. All the sounds in Table 2.1 are pulmonic egressive. Even in languages that use other airstream mechanisms for certain consonants, pulmonic egressive will be used for the majority of consonants, and for all the vowels. Because this is the most common state of affairs, it generally goes without saying: Unless otherwise stipu- lated, assume the airsteam is pulmonic egressive. In linguistics, when we can assume one state of affairs holds generally, although an alternative may be specified in a more limited set of cases, we call the general case the default or unmarked state. The limited case is the marked option. Pulmonic egressive is the unmarked airstream mechanism. The second defining characteristic of a sound is the state of the vocal folds. In English, the folds may be held in the airstream at the correct tension to produce vibration, or they  Basics of Articulation   21 may be pulled out of the way so that the air passes freely between them, or they may be clamped down tight to close off the airflow completely. As was noted in Chapter 1, sounds produced with vocal fold vibration are voiced; sounds produced without vocal fold vibration are voiceless. Again, if you place your finger on your larynx and produce a sustained [z], you should be able to feel the vibration. If you switch to [s], a voiceless sound, the vibration ceases. Another pair of sounds that illustrate the voiced/voiceless distinction is [v] and [f]. Can you feel which is which? If a sound is produced as voiceless, the speaker must control how long and large the vocal fold opening will be. For some sounds, as in the initial [p] in “pat,” the vocal folds are held apart far enough and long enough to allow an extra “puff of air” to exit the mouth at the end of the [p]. This extra voicelessness is called aspiration. You can feel the extra release of air if you hold your fingertips an inch or so in front of your lips as you say “pat” or “pill,” but not as you say “bat” or “bill.” Another test for aspiration is to hold a piece of notebook paper flat and level in front of your lips, holding it a few inches from the end so that one edge of the paper hangs down. Saying “pat” will make the edge of the paper move. However, the paper will not move, and you will not feel any aspiration, for a [p] that occurs after [s] as in “spot” or “spill.” For these non-initial consonants, the vocal folds resume a position of voicing more quickly. Aspiration is indicated by a superscript “h” following the consonant [ph], though this detail is indicated for English only in narrow transcription. Voicing and aspiration are the two most important systematic laryngeal distinctions for English vowels and consonants, but speakers have control over various other laryngeal con- figurations. English uses these distinctions mostly on the periphery of its vocabulary, in a small set of expressions or exclamations, though they play a more central role in other lan- guages. Shutting the folds down very tight, completely cutting off the airflow at the larynx produces a glottal stop (IPA [ ]). A glottal stop occurs in the middle of the word (or expres- sion) usually written “uh-oh.” If you say this slowly, stopping after the first syllable, you should be able to feel the tension in your throat, and note that you are also holding your breath. Some English speakers produce glottal stops in place of [t] in certain positions (which positions vary by dialect). In American English, this is often at the end of the word or before an [n]. For example, many American English speakers pronounce [ ] instead of [t] in “button,” or in a not-too-careful pronunciation of the word “important” (IPA [ ]). Adjusting the tension on the vocal folds changes the quality of voicing, making it creaky (more tense) or breathy (more relaxed). A narrow IPA transcription indicates these differ- ences with subscripts: a tilde for creaky and a double dot for breathy. Voice quality may be the only difference between affirmative, breathy [ ], and negative, creaky [ ]. The third defining characteristic is whether the velum will be open or not. If the velum is open, so that air flows through the nasal cavity, the sound is nasal (like [m]). If the velum is closed, the sound is oral. 2.3.2 manner of articulation Manner of articulation defines the kind or degree of constriction that is made, independent of where in the vocal tract. As was noted in Chapter 1, there are three ways of making moving air audible: making it pop, making it turbulent, or making it resonate. The first two ways involve creating some obstruction to the flow of air, and sounds made in this way are called obstruents. Sounds that resonate are sonorants. The set of obstruent consonants includes plosives, fricatives, and affricates. The set of sonorant consonants includes nasal stops and approximants. Vowels are always sonorant. The consonantal manners of articulation are seen in the leftmost column of Table 2.2. Plosives may also be called oral stops. As the name implies, a stop manner of articulation brings the active and passive articulators together to make a complete closure that stops air 22   Basics of Articulation from exiting the mouth. Try saying the word “poppa” very slowly, and note that there is complete silence, with no air exiting the mouth, while the lips are closed for [p] in the middle of the word. You may even feel pressure building up behind the lips, as air continues flowing from the lungs and has nowhere to go. This pressure is released with a slight pop, or burst, when the lips are opened. The sounds [p], [t], and [k] are plosives. Can you the feel the closure between the different active and passive articulators in the syllables [ ] (lower and upper lips), [ ] (tongue front to alveolar ridge), and (tongue body to velum)? For a fricative, the articulators are brought close together but not quite closed completely, creating a narrow slit or groove through which the stream of air is forced. Just as a river will flow smoothly and noiselessly when its channel is wide, but will become faster, turbulent and noisy when forced to flow through a narrow channel, so the flow of air becomes turbu- lent and noisy when channeled through a narrow gap in the vocal tract. The sounds [f] and [s] are fricatives. Try drawing out each sound. You should both be able to hear the charac- teristic hissing noise, and to feel where the constriction is made. For [f], air is forced between the upper teeth and lower lip. For [s], air shoots down a groove in the center of the tongue, like water out of a pipe, and crashes into the teeth. If you think about the splash that can be created with water shooting out of a hose and splashing into a wall, you’ll understand why [s] is so loud. An affricate combines a sequence of stop plus fricative in a single sound. The sound usually written “ch” in English is an affricate. The IPA symbol is [ ]: the two letters symbolize the stop and the fricative, combined into a single symbol that counts as one sound. Say the word “achoo” as slowly as possible, paying attention to what your tongue is doing in between the “a” and “oo” sounds. You should feel that you first make a closure of the tongue front at or just behind the alveolar ridge, and then lower the tongue front to let the air out through a narrow constriction. As you can see from the IPA chart, plosives, fricatives, and affricates (that is, the class of obstruents) can be voiced or voiceless. For [p, t, k, f, θ, s, ∫], the vocal folds are pulled apart during the time the oral constriction is made, making them voiceless. For the voiced stops [b, d, g] and voiced fricatives [ ] the vocal folds are held together throughout, and voicing may continue through the constriction. There is, however, something of a contradic- tion between voicing and obstruents. Voicing requires airflow over the vocal folds to produce the vibration, but obstruents, by definition, block this flow to a greater or lesser extent. Pres- sure builds up behind the constriction in the oral cavity, and when air pressure in the mouth and lungs becomes equal, airflow stops. When airflow stops, vibration of the vocal folds stops. Therefore, while voiced stops and fricatives are possible, they are relatively difficult. The voiced plosives of English tend to devoice as vocal fold vibration dies out, and some languages (such as Finnish, Hawai’ian, Mandarin, and Nootka, to name a few) skip voiced obstruents alto- gether. Because they are both more difficult and rarer, voiced obstruents are marked. Because they are both easier and more common, voiceless obstruents are unmarked. Sonorants, on the other hand, are almost always voiced. You will see only a single sound in each sonorant block in the IPA chart. That is, for sonorants the reverse holds: voiced sonorants are unmarked. The definition of a sonorant is a sound in which there is no build- up of pressure in the supralaryngeal vocal tract. By definition, they allow the free flow of air, so there is no hissing or popping involved. Sonorants make audible sound not by obstructing the airflow, but by making the air resonate, that is, causing it to vibrate at different frequen- cies. Chapters 6 through 8 discuss vocal tract resonance in detail, including formulas for deriving different vowel sounds; here a few words on how resonance works in general should suffice to understand how sonorants are made. A sonorant speech sound is very much like the note of a clarinet. To play a clarinet, the musician blows air over the reed, causing it to vibrate. The vibration of the reed causes the air inside the body of the instrument to vibrate as well, producing a particular note. When  Basics of Articulation   23 the musician changes his fingering, opening and closing the valves on the body of the instrument, he effectively is changing the size of the column of air inside, which changes the note. The vocal tract works in a similar way. In the vocal tract, air passing over the vocal folds causes them to vibrate. The vibration of the folds then causes the air in the supra- laryngeal vocal tract to resonate. When the speaker moves her tongue, lips, and other articulators, she changes the shape of the vibrating body of air, and this changes the quality of sound that is produced. The shape of the vocal tract is more complex than the shape of the clarinet, so the sound that is produced is more complex as well. (Details can be found in Chapter 8.) Because vibration is the basis of resonance, sonorants are almost always voiced. It is possible to get sound without vocal fold vibration out of a vocal tract configured for resonance, but it requires a lot of airflow, as in whispering, which is both very effortful and not very loud. Sonorant consonants include the nasal stops, such as [m]. An [m] counts as a stop because the lips are completely closed, exactly as they are for [p] or [b]. No air exits the mouth. If you observe your lips in a mirror, or just feel their position, as you say phrases such as [ ] or [ ], you will see and feel the same lip position in both. The difference between a plosive and a nasal stop, however, is that for a nasal stop the velum is open and air flows freely out of the nose. There is thus no build-up of pressure, and no burst noise on release. The vibra- tion of the air in the mouth and nasal cavity produces the sound. You can probably feel this vibration by articulating a sustained [m] (humming). Now, while you’re humming, try pinching your nose closed for a moment. What happens? You can also make a voiceless nasal: try producing an [m], then making it voiceless. The only sound you’ll make is that of quiet breathing. You can make the voiceless nasal louder by forcing a lot of air through the nose (but proceed with caution if you have a cold). The next set of sonorants is the approximants. In an approximant, the active articulator moves to narrow the vocal tract, but not so much that fricative noise is created. Glides, such as the sounds at the beginning of the words “yell” and “well,” are approximants, as are the sounds [r] and [l]. There are many different r-sounds, called rhotics, in the languages of the world, and l-sounds, called laterals. Rhotics and laterals together are sometimes called liquids. The [r] sound in American English is one of those aspects that is unusual cross-linguis- tically: The body of the tongue is bunched up high, and the tongue tip may be raised or curled backwards, without touching any other surface of the vocal tract. (Try saying [rrrrr].) It is no surprise that non-native speakers of English have trouble with this sound. In narrow transcription, the symbol for the specific configuration of the English rhotic is [ ]. Since English has only one rhotic, however, in broad transcription the basic symbol [r] can be used, and that practice will be followed in this chapter. Chapter 3 discusses some other rhotic variants. The l-sounds are called laterals because air flows out over the sides of the tongue. Try drawing out the initial sound in the word “lateral”: [lllllllll…]. You should be able to feel the contact between tongue tip and upper teeth or alveolar ridge. But [l] is not a stop: air con- tinues to flow, exiting not through the center of the mouth but at the side. Without moving your tongue, take a deep breath in and out. Can you feel the air moving over the sides of the tongue? (For some people, air flows over the right side of the tongue only, for others only the left, and for others both. Which are you?) The final class of sounds to be considered is the class of vowels. For a vowel, the vocal tract is relatively wide open, with no significant constriction at any point. Different vowel sounds are made by changing the position of the tongue and lips, changing the shape of the resonating body of air. Different vocal tract shapes for different vowel sounds are discussed in Section 2.3.4 below. To summarize the manners of articulation covered so far: 24   Basics of Articulation Stop: complete closure cutting off airflow in the oral cavity: Oral stops (plosives) have a closed velum, and build-up of pressure behind the oral closure. Nasal stops have an open velum, with air escaping through the nose. Fricative: narrow constriction causing turbulent, noisy airflow. Affricate: sequence of stop + fricative in one sound. Approximant: some constriction in the vocal tract, more so than for vowels, but not so much as to produce frication: Glides. Laterals (l-sounds). Rhotics (r-sounds). Vowel: open vocal tract. 2.3.3 place of articulation for consonants Having covered airstream, nasality, voicing and manner of articulation, we move on in this section to place of articulation. With all of these descriptive tools in place, we are now able to work systematically through the IPA chart of English consonants in Table 2.2, covering each of the sounds represented. Table 2.1 provides example words for each sound. As noted above, the term “place of articulation” combines both the articulator that moves to make the constriction, the active articulator, and the location to which it moves, the passive articulator. We will organize our discussion of the consonants grouped by active articulator first, and passive articulator second. Generally, each active articulator can move to more than one location. The combinations of active and passive articulator that make up the places of articulation for English are shown in Table 2.3. Any sound made with the lower lip as an active articulator is termed labial. The lower lip can make constrictions at two different places. If the lower and upper lips come together, the sound is bilabial. The sounds [p], [b], and [m] are bilabials. Note that [p] is voiceless and [b] (and [m]) are voiced. Alternatively, the lower lip can make contact with the upper teeth to produce a labiodental sound. The fricatives [f] and [v] are labiodentals. English makes its labial stops at the bilabial place of articulation and its labial fricatives at the labi- odental place. The lower lip is rather limited in the places at which it can make a constriction. The next articulator, the tongue front, is the most versatile of the active articulators, moving to at least four different places of articulation. All sounds made with the tongue front are coronal. The tongue tip moves forward to the upper teeth for the sounds at the beginning of “thin” and “then.” Table 2.3 Active articulators, passive articulators, and place of articulation. These dental fricatives are written [θ] (voiceless) Active Articulator + Passive Articulator = Place of Articulation and [ð] (voiced) in IPA. Some English speakers protrude the tip of the tongue between the teeth lower lip upper lip bilabial in producing these sounds (making them inter- upper teeth labio-dental dental); other speakers are more circumspect and tongue front upper teeth dental keep the tongue behind the teeth. (Which are you?) alveolar ridge alveolar Sounds made with the tongue front against post-alveolar region retroflex (tip) the alveolar ridge comprise the alveolar place of post-alveolar region post-alveolar (blade) articulation. The alveolar sounds of English tongue body hard palate palatal include the plosives [t] and [d], fricatives [s] and soft palate velar [z], the nasal stop [n], and the lateral [l]. There is some interesting variation in the way [s] and larynx laryngeal [z] are made. Some English speakers raise the tip  Basics of Articulation   25 of the tongue to the alveolar ridge for [s] and [z]; others tuck the tip of the tongue behind the lower teeth and make the alveolar constriction using the blade of the tongue, a centimeter or so further back. In either case, for the alveolar fricatives, the tongue forms a narrow groove like a spout that shoots a stream of air against the teeth, producing a very high-pitched hissing sound. Though the point of narrowest constriction for these fricatives is alveolar, the front teeth are necessary to create the proper high-pitched hiss, as every child discovers when she loses her baby teeth. The fricatives [ ] and [ ] (as in the middle of “pressure” and “pleasure,” or the beginning of “ship”) are made further back, with the blade of the tongue making a constriction at the point where the roof of the mouth is steeply rising behind the alveolar ridge. This place of articulation is variably known as palato-alveolar, alveo-palatal, or post-alveolar. Here, we will follow the IPA chart and use post-alveolar. (If you produce an alternating sequence of [s – – s – ] you’ll feel how the constriction moves: [ ] is further back both on the tongue and on the roof of the mouth.) These fricatives also involve a grooved tongue shape that channels the air. Interestingly, [ ] never occurs in initial position in English, except in obvious bor- rowings, such as “genre” from French, or in names such as “Zsa Zsa” or “Zhivago.” (For the record, my own family name was pronounced with initial [ ] in the original Hungarian, like Zsa Zsa, but was Americanized to [z] three generations ago.) The affricates in “church” [ ] and “judge” [ ] are also post-alveolar. In other transcription systems commonly used in linguistics books and dictionaries, such as the American Phonetic Alphabet (APA), [ ], [ ], [ ], and [ ] are written [š], [ž], [č], and [ ]. The IPA convention for affricates, with its double symbols, emphasizes the sequenced aspect of the sound. The APA convention emphasizes the affricate’s unity. This book will follow IPA convention, but the APA convention is also common, and every student of linguistics should be familiar with it. Usually a post-alveolar constriction is made with the blade of the tongue. It is also pos- sible, however, for the tip of the tongue to curl back to make a constriction in this area. If the tip of the tongue curls back, the sound is called retroflex. For some English speakers (but not all), [r] is a retroflex approximant ([ ] in narrow transcription). You can determine whether your own tongue tip curls back in a word like “road” by drawing out the initial [r], then (gently!) inserting a toothpick between your teeth until it touches the surface of the tongue. If you feel the point of the toothpick on the underside of your tongue, then your [r] is a retroflex approximant. If you feel the point on the upper side of the tongue, then your [r] is more likely an alveolar or post-alveolar approximant. Moving further back in the vocal tract, the next place of articulation is palatal. The glide at the beginning of the words “you” and “yacht” is palatal. To make a palatal constriction, the whole middle section of the tongue, including blade and body, is pushed straight up to narrow the space between the tongue and hard palate. The IPA symbol for a palatal glide is [j]. (Think Scandinavian “ja.”) The IPA reserves the symbol [y] for a particular, compara- tively rare, vowel sound, and linguists describing languages that don’t have that vowel sound often use [y] for the palatal glide instead. Here, however, we’ll stick to IPA. Sounds for which the tongue body makes the primary constriction are termed dorsal. If the tongue body moves further back to make a constriction against the velum, high in the back of the mouth, the place of articulation is velar. The English sounds [k] and [g] are velar stops. The sequence of letters “ng,” as at the end of “song” or “ring,” usually indicates a velar nasal. If you pay attention to what your tongue is doing in the word “song,” you’ll realize you’re not making a sequence of alveolar nasal followed by velar stop (n-g), but a single nasal sound at the same place as [k] or [g]. (Notice how little your tongue moves in the word “king”: the tongue tip is not involved at all.) The IPA symbol for a velar nasal stop is [ŋ]. As with [ ], English uses [ŋ] only at the end of words (“song”) or in the middle (“singer”), never at the beginning, though other languages use [ŋ] as freely as [n] or [m], as in Thai [ŋa:] tusk, [ma:] come, [na:] rice paddy. (The colon indicates a long vowel.) 26   Basics of Articulation The glide [w], as in “well” combines a narrowing of the vocal tract at the velar place of articulation with rounding of the lips. It is thus a double-articulation, a labio-velar glide: in the chart, it is entered under both bilabial and velar place. The sound [h] is made with the larynx as the only articulator. An [h] consists of the noise of air rushing through the open vocal folds, and may be considered a glottal fricative. As was noted in Section 2.3, it is also possible to produce a glottal stop (IPA [ ]), by closing the vocal folds up tight, stopping the airflow at the larynx. It is perfectly possible to make consonantal constrictions in between velar place and laryngeal place: The tongue body can move to make constrictions against the uvula (uvular place), and the tongue root can move back to constrict the pharynx (pharyngeal place). Consonants at these places of articulation are not used in English, however. Uvular and pharyngeal consonants are discussed in Chapter 3. In summary, there are nine places of articulation used for English consonants: bilabial, labiodental, dental, alveolar, post-alveolar, retroflex, palatal, velar, and laryngeal. Keep in mind that each symbol refers to a specific combination of place, manner, voicing, and nasality. For example, [m] = bilabial nasal (voiced) stop, and [s] = alveolar (oral) voiceless fricative. Predictable values are listed in parentheses: one can leave out “voiced” in the description of [m] because all English nasals are voiced, and leave out “oral” in the description of [s] because no English fricatives are nasal. In most cases, however, exchanging one term for another creates a different sound: changing “bilabial nasal stop” to “alveolar nasal stop” changes [m] to [n]. Changing “alveolar voiceless fricative” to “alveolar voiced fricative” changes [s] to [z]. It is a useful exercise to keep these relationships in mind. The articulatory information provided in the preceding sections provides enough infor- mation to identify and distinguish the consonants of English. Much more can be said about the details of their articulation, relationships with each other, and distinctions from the consonants of other languages, and these topics will be covered in the following chapters. First, however, Section 2.3.4 turns to the (large) set of English vowels. 2.3.4 vowels Vowels are harder to describe than consonants. By definition, vowels have an open vocal tract, so the tongue does not make contact at any particular place. If you slowly repeat the sequence of words “he, who, ha,” drawing out the vowel sounds, you can feel that your tongue and lips move to different positions, but it is very difficult to feel exactly what position the tongue has taken. It was not until the advent of devices such as X-rays, which could actually image the whole of the tongue inside the vocal tract, that phoneticians had a very accurate idea of the exact articulatory positions for vowels. We are much more sensitive to the vari- ations in sound quality that these movements produce. Nonetheless, vowels are described in terms of the relative position of the tongue body – high, low, front, back – and of the lips – pursed, (that is, rounded) or not. If describing vowel systems in general is a difficult task, describing the vowels of English is even more so. As noted above, the vowel system of English is marked compared to that of most other languages. The most common number of vowels for a language to have is five. Spanish, Hawai’ian, Swahili, and Modern Standard Arabic, to name a few, use just five dif- ferent vowel qualities. Though English writers use just five letters to encode them (relics of an older system!), the English language uses more than a dozen different vowel sounds. The exact number depends on the dialect. Table 2.4 gives the IPA symbols, with example words, for the vowels of “General American” English, that is, English as it is more or less spoken in the central United States, or on national  Basics of Articulation   27 2.3 Another reason the English vowel system is difficult to describe: vowel qualities differ a lot from dialect to dialect, much more so than for the consonants. For example, the word “mate” as pronounced by a speaker from Perth sounds a lot (though not exactly) like the word “might” as pronounced by a speaker from Baltimore. The word “my” spoken by a native of Atlanta sounds similar to “mar” as pro- nounced in Boston, and “ma” as pronounced in Seattle. For most speakers on the East Coast of the United States, the words “caught” and “cot” have two different vowel sounds; for most speakers on the West Coast, the two words are pronounced the same. TV. (I say “more or less” since every dialect of English has Table 2.4 IPA symbols for the vowels of General American English. particular regional characteristics, and the ideal speaker i bead key he of pure “General American” doesn’t exist.) Your pronun- ciation of the example words may well differ from the bid kit one described. For example, if you’re from California, you e bade kate hey probably use a vowel in between [ ]and [ ] instead of two ε bed ketchup distinct vowels. If you speak a British variety of English, you may have an additional low vowel, a round version æ bad cat of [ ], in words like “lot.” (For more discussion of dialect u booed coot who variation in English vowels, see Chapters 4 and 19; Table book cook 19.1 compares vowel pronunciation in General American o bode coat hoe and in “BBC” English.) baud caught haw Table 2.4 (and Figure 2.1) use broad transcription, so certain details are left out. For example, the vowels in body cot ha “bade” and “bode” actually change in quality during the bud cut pronunciation of the word: For both vowels the tongue bide kite high body moves up towards the end of the vowel, and in a bout count how narrower transcription they would be represented as [ ] and [ ]. While this “off-glide” is a salient boy coy ahoy aspect of the pronunciation of these vowels in American English, pronouncing [e] or [o] without the offglide does not make a different word: it just makes you sound like you have a foreign accent. front central back In contrast, for the vowels in “bide,” “bout,” and “boy,” the movement upward at the end of the vowel is contrastive: “find” i u high [ ] is distinct from “found” [ ], which is distinct from “fond;” and “soy” [ ] is distinct from “saw.” Thus a separate symbol for the offglide is included in the transcription. These two-part sequenced vowels (the vocalic counterparts of affricates) are e o mid called diphthongs. (Vowels like [ ] and [ ] that have a non- ε contrastive offglide may be described as diphthongized.) As with the consonants, vowels can be described in terms of a set of choices made by the speaker. Vowels are almost always æ (a) low voiced; they may be nasalized or not. All vowels use the tongue body and lips, and sometimes the tongue root, as active articula- tors. By definition, all vowels have an open manner of articula- Figure 2.1 Chart of the vowels of General American English. tion. Because the tongue does not actually touch the upper Example words are given in Table 2.4. The vowel [a] is shown in surface of the vocal tract anywhere, the term place of articulation parentheses because it occurs in GA only in the first half of diph- is not really appropriate, but different vowels may be described thongs. Source: Based on a similar chart in Zsiga (2006). 28   Basics of Articulation 2.4 In Focus Note that the IPA symbols for the vowels do not correspond to their usual values in English spelling, nor to the names of the letters as they are taught to children. (Once I became used to phonetic transcription, I became absolutely useless in helping my kids with their spelling words.) The word “see,” for example, with its “long e” is transcribed as [si]. (The pronunciation of IPA [i] and [u] are actually much closer to their values in Spanish spelling, if that’s a help to you.) The reason for the discrepancy lies in the history of English. Five hundred years ago, when English spelling was getting established, the sounds of the English vowels were much closer to their mainland European counterparts: “see,” for example, was pronounced similarly to a long version of Spanish “se” (similar to present-day English “say”), and the difference between “made” (long “a”) and “mad” (short “a”) really was one of length. Over the next 500 years, however, “The Great Vowel Shift” moved the vowel qualities of English around like the tiles in those plastic puzzles where you have to put all the numbers in order: See Chapter 18 for more details. in terms of the ways in which the tongue body and lips move. Figure 2.1 charts the positions of the vowels relative to each other, based roughly on the position of the highest point of the tongue during that vowel. The terms we use to classify different vowels refer to this position. Vowels may differ with respect to the height of the tongue: the tongue body moves up for the high vowels [ ], down for the low vowels [ ], and stays in the middle for the mid vowels [ ]. The tongue moves forward in the mouth for the front vowels [ ] and backward for the back vowels [ ]. The vowels [ ] and [a] are central. In General American English, [a] occurs only as the first part of a diphthong: [ ] begins in a low central position and moves to high front, [ ] begins from the same low central position and moves to high back. Vowels also differ with respect to lip rounding. In General American, the back vowels [ ] are round, all other vowels are unround. Finally, English divides its vowels into two sets, tense and lax. The tense vowels [i, e, o, u] are longer, slightly higher, and produced with greater stiffening of the tongue root than their lax counterparts [ ]. (Linguists disagree over which set the low vowels belong to.) All these descriptive terms can be combined to pick out a specific vowel: [ ] is high, front, lax, unround; [o] is mid, back, tense, round. Can you write the symbol for the mid, front, lax, unround vowel? High, front, unround, tense? One more English vowel should be mentioned here: the reduced vowel [ ], called schwa. This symbol is used to mark the vowel quality in short, unstressed syllables, such as first syllable of “about” [ ] or the last syllable of “panda” [ ]. Because it is a short, mid, central vowel, the symbol [ ] means that the mouth is open for a vowel sound, but the tongue basically stays in a neutral position: neither high nor low, front nor back. For a mid, central vowel in a stressed syllable, the symbol [ ] (“wedge”) is used. Thus the word “abut” is tran- scribed [ ]. More details on describing, and more narrowly transcribing, English consonants and vowels are given in Chapters 3 and 4 respectively. But the symbols discussed here should be sufficient to unambiguously transcribe any word of English. Just as important as getting the symbols correct, however, is understanding the different articulatory choices that a speaker combines to create each sound.  Basics of Articulation   29 chapter summary Human language is discrete, combinatorial, unbounded, and recursive. Utterances are made up of words, which are made up of sounds, which are made up of articulatory gestures. English spelling is ill-suited for a general representation of the sounds of language. The International Phonetic Alphabet aims to provide one unambiguous symbol for any sound used in the languages of the world. A speaker makes choices to produce a speech sound. These include: How should I get the air moving? Generally, this will be pulmonic egressive: air forced out of the lungs. Which active articulator should I use: lips, tongue front, tongue body, tongue root, or larynx? What kind of constriction should I make: stop, fricative, affricate, approximant, vowel? Where should I make the constriction? For consonants, the choices are bilabial, labio­ dental, dental, alveolar, postalveolar, retroflex, palatal, velar, uvular, pharyngeal, and laryngeal; for vowels, the choices are high/mid/low, front/central/back, tense/lax, and round/unround. Should the velum be open or closed? What should I do with the larynx: voiced or voiceless, aspirated or unaspirated? The figures and charts provided should help in keeping these straight. The best way to learn the symbols and descriptions is to practice! further reading Any introductory phonology or phonetics book will cover place and manner in English. If you’d like to read about the topic from another source, I recommend: Ladefoged, Peter (2006). A Course in Phonetics. 5th edition. Independence, KY: Thomson Wadsworth. For more on writing systems: Rogers, Henry (2005). Writing Systems: A Linguistic Approach. Oxford: Blackwell. The idea of speech as a “dance of the articulators” is due to Louis Goldstein. The centrality of the articulatory gesture as the basic building block of speech is based on the joint work of Louis Goldstein and Cathe Browman (see Chapter 5 for more details on their theory of “Articulatory Phonology”).