Chapter 7
Sounding Out English Vowels
In This Chapter
Searching for (IPA) meaning in all the right places
Hearing vowels in full and reduced forms
Switching between British and North American English vowels
Keeping track of vowel quality over time
Vowels are a favorite subject of phoneticians because they play such an important role in perception, yet they pose so many mysteries about how speech is perceived and produced. Some vowels are quite easy to transcribe; some remain difficult. In this chapter, I highlight the commonalities among English vowels by describing the group’s tense and lax characteristics. I also talk about rhoticization (also referred to as r-coloring), which is important for many applications, including the description of various English accents and understanding children’s language development.
Cruising through the Vowel Quadrilateral
Making vowels is all about the tongue, lips, and jaw. However, the final product is acoustic (sound related), not articulatory (mouth related). Phonetics texts typically start out with articulatory instructions to get people started, but it becomes important to transfer this information to the ear — to the world of auditory information.
In articulatory phonetics, vowels are studied using the vowel quadrilateral, a trapezoid-like diagram that classifies vowels according to tongue height, advancement (front-back positioning), and lip rounding.
This section focuses on moving your tongue to known target regions and consciously getting used to what these regions sound like. In this way, sound anchors become familiar landmarks as you cruise through the land of vowels.
Looking up vowel sounds
Are you a book person, an Internet person, or both? The tools available for looking up sounds are becoming more extensive and convenient. As your transcription skills improve, you may find yourself wishing to compare pronunciations of certain words now and then. At such times, you may notice quite a bit of difference among various reference sources because vowels can be transcribed in a number of ways, depending on the exact needs of the transcription.
In the following, I compare some dictionary and Internet sources for the English words “cheap” and “chip” (American pronunciation) to see how they’re transcribed. This table can give you an idea of how transcription is handled in a range of available sources.
|
Source |
“cheap” |
“chip” |
|
English Pronouncing Dictionary (EPD) |
/ʧi:p/ |
/ʧɪp/ |
|
Longman Pronunciation Dictionary (LPD) |
/ʧi:p/ |
/ʧɪp/ |
|
Oxford Dictionary of Pronunciation for Current English (2003) |
/ʧi:p/ |
/ʧɪp/ |
|
American Heritage Dictionary 5th Edition (2011) |
(che̅p) |
(chĭp) |
I selected these sources because they’re some of the most authoritative print dictionary sources. The first three sources provide both British and American English pronunciations. Unlike the conventions used by most phonetics teachers (and used in this book), these first three sources are also more detailed in that they show both length and quality features for the “cheap/chip” distinction. That is, there are two related phonetic factors that contribute to the vowel difference:
Quality: The formants are different in the two words, requiring different symbols /i/ versus /ɪ/
Length: The vowel is longer in cheap than “chip,” requiring a length diacritic for cheap /i:/
The Oxford Dictionary uses an even larger set of symbols and gives more details than the first two sources.
A different approach has been taken in the American Heritage, where an alternative to the International Phonetic Alphabet (IPA) was used, presumably to be more user friendly. This, however, isn’t as handy or reliable for some people who’ve already learned the IPA.
This book follows a convention used by many phonetics instructors and transcribes vowel quality, assuming that length can be implied. Therefore, I transcript “cheap” as /ʧip/ and “chip” as /ʧɪp/. As long as you remember that quality and length go together, you should be able to appreciate other transcriptions (such as EPD and LPD) when you use them.
Sounding out front and back
Sound-based descriptions are especially important for vowels. For this reason, phoneticians have long relied on perceptual descriptions of vowels. For instance, front vowels were frequently called acute because they’re perceptually sharp and high in intensity. These vowels also trigger certain sound changes in language (notably, palatalization) and involve active tongue blade (coronal) participation. In contrast, back vowels were called grave because they have dull, low intensity.
You have made front vowels, but you have probably not spent that much time attending to what the vowels sound like. So here, you tune in to the sounds themselves. Begin by making an /i/ as in “heed,” then move to the following, one by one:
/ɪ/ as in “hid”
/e/ as in “hayed”
/ɛ/ as in “head”
/ӕ/ as in “hat”
You hold your tongue in a certain position for each vowel (although there is some wiggle room), and the tongue position need not be exact. Also, each vowel position can blend somewhat into the position of the next.
Now, try saying them all together in a sequence, /i ɪ e ɛ ӕ/. Notice that the vowels are actually in a continuum. Unlike consonants, vowels are made with the tongue relatively free in the articulatory space and the shaping of the whole vocal tract is what determines the acoustic quality of each sound.
Now try the same listening exercise with the back vowel series, beginning with /u/ as in who’d, and proceeding to the following:
/ʊ/ of “hood”
/o/ of “hope”
/ɔ/ of “law”
/ɑ/ of “dog”
In the back vowel series, you pass through the often-confused /ɔ/ and /ɑ/. There are many dialectal differences in the use of these two vowels. For instance, in Southern California (and most Western United States dialects), most talkers pronounce “cot” and “caught” with /ɑ/. In Northern regions, say Toronto, talkers use /ɒ/ for both words. This vowel /ɒ/ is a low back vowel similar to /ɑ/ but produced with slight lip rounding. However, elsewhere in the States (especially in the Mid-Atlantic States) talkers typically produce “cot” with /ɑ/ and “caught” with /ɔ/. You can easily tell the two apart by looking at your lips in a mirror. During /ɑ/, your lips are more spread than in /ɔ/, and in /ɔ/ the lips are slightly puckered. Compare your productions with the drawings in Figure 7-1.
Illustration by Wiley, Composition Services Graphics
Figure 7-1: Lip positions for /ɑ/ versus /ɔ/.
The point here is to be able to hear such differences. Try moving from an /o/ as in “hope” to an /ɔ/ in “law” to an /ɑ/ in “father.” Now try this while adding lip rounding to the /ɑ/. You should hear its quality change, sounding more like /ɔ/. As you get more fine-grained in your transcriptions, you need to be able to distinguish vowels better, including whether lip rounding occurs as a secondary articulation.
Stressing out when needed
In English, stress refers to a sound being longer, louder, and higher. Stress is a suprasegmental property, meaning it affects speech units larger than an individual vowel or consonant. I also discuss stress in Chapters 10 and 11.
In English, the amount of stress a syllable receives influences vowel quality. Stressed syllables tend to have a full vowel realization, while unstressed syllables have a centralized, reduced quality. Sometimes there is a more complicated situation, where a full vowel will appear in fully stressed syllables, but whether a vowel is reduced in unstressed syllables depends on the particular word involved. Take a look at Table 7-1, and try saying the English words.
You should have nice, full /i/, /u/, and /aɪ/ vowels in the words of the second column of Table 7-1. This should also be the case for the words in the third column.
The most variable response is the fourth column. The vowels produced here depend on your accent. These words contain unstressed syllables that some speakers produce with a fully realized vowel quality (for example, /ɹi/ for the first syllable of “resourceful”) while others use a reduced vowel instead (such as /ɹə/). If your vowel is a bit higher toward /ɪ/, it may qualify for being /ɨ/ (called barred-I in IPA), as is frequently heard in American productions of words such as “dishes” and “riches.”
By the way: If you find yourself almost forgetting what you normally sound like, please remember these rules:
Use a carrier phrase. A carrier phrase is a series of words you place your test word into so that it’s pronounced more naturally. For example, “I said ___ again.”
Have one or two repetitions and then move on. Natural speech is usually automatic and not consciously fixated on. If a word or phrase is repeated over and over, this natural, automatic quality may be lost.
Coloring with an “r”
Whether or not people produce an “r” quality in words like “further,” “father,” and “sir” is a huge clue to their English accent. Most speakers of North American English produce these vowels with rhoticization. This term, also referred to as r-coloring, means that the vowel (not the consonant) has an “r”-like sound. If the vowel is stressed, as in “further” or “sir,” then you use the mid-central stressed vowel /ɝ/ symbol for transcription. For unstressed syllables, such as the “er” of “father,” you instead use the IPA symbol schwar /ɚ/.
R-coloring is a perceptual quality that can be reached in a number of ways. R-coloring demonstrates the property of compensatory articulation, that a given acoustic goal can be reached by a number of different mouth positions.
R-coloring can differ substantially among individual speakers. Some make a retroflex gesture, putting the tip of the tongue against the rear of the alveolar ridge, while others hump the tongue in the middle of the mouth, sometimes called American bunched r. These vowel gestures are very similar to the consonant /ɹ/ in English and are described in detail in Chapter 6.
A useful series of r-colored vowels can be elicited in the context /fVɹ/ where V stands for a vowel. Table 7-2 contains many of these items and some others, including common North American English and British English words. Try these words out and see how much rhoticization (r-coloring) you use.
Different transcription systems may be used for non-rhotic forms of English, such as commonly found in parts of the United Kingdom, Ireland, South Africa, and the Caribbean. I give more detail on different accents in Chapter 18.
A symbol used to describe the central nonrhotic (stressed) vowel is /ɜ/ (reversed epsilon). You can find this vowel in Received Pronunciation (RP) British for words such as “fur” and “bird.”
Neutralizing in the right places
The vowels /o/ and /i/ make predictable changes in particular environments. Phoneticians have adopted conventions for transcribing these patterns. For example, take a look at these transcriptions (GAE accent):
|
sore |
/sɔɹ/ |
|
selling |
/ˈsɛlɪŋ/ |
Beginning transcribers are often puzzled as to why /ɔ/ is used in “sore” (instead of /o/), and why /ɪ/ is used before /ŋ/ in words that end with -ing, such as “selling.” The answer is that vowels are affected by their surrounding consonants. These effects are more pronounced with certain consonants, especially the liquids (/ɹ/ and /l/) and nasals (/m/, /n/, and /ŋ/). This results in neutralization, the merger of a contrast that otherwise exists. For example, /o/ and /ɔ/ sound quite distinct in the words “boat” and “bought” (at least in GAE). However, before /ɹ/ these vowels often neutralize, as the /ɹ/ has the effect of lowering and fronting the /ɔ/ toward the /o/. Front vowel examples include “tier” and “pier” (pronounced with /ɪ/). The same process can take place before /l/. Examples include “pill” and “peel,” both produced as /pɪɫ/ in some accents.
Say “running.” Do you really make a tense /i/ as in “beat” during the final syllable? Probably not. For that matter, you’re probably not making a very pure /ɪ/ either. You’re neutralizing, making something in-between. To label this sound, phoneticians lean toward the lax member and label it /ɪ/. Thus, /ˈɹʌnɪŋ/.
It’s the same principle with “sore.” You’re probably not using a tense /o/, such as in “boat.” Listen closely! The closest vowel that qualifies is /ɔ/, even though its quality is different when rhotic.
Tensing up, laxing out
This tense versus lax vowel difference is important for a number of applications in language instruction and clinical linguistics. Specifically, the tense-lax difference indicates whether a vowel can stand alone at the end of a stressed syllable (tense), or whether the syllable must be closed off by a consonant at the end (lax). Many languages (such as Spanish) don’t have any of the English lax vowels, and native speakers will therefore have difficulty learning them when studying English as a second language.
Take a look at these word pairs and pronounce them, one by one.
“beat” versus “bit”
“bait” versus “bet”
“Luke” versus “look”
Can you hear a systematic change in the sound of each pair? The first member of each pair is tense, and the second member, lax. This distinction was originally thought to result from how the vowels were made, muscularly. However, these differences are now understood as relating to English phonology (system of sound rules). Refer to Table 7-3 for examples.
The tense vowel /i/ can appear in a stressed open syllable word such as “bee,” or in a syllable closed with a consonant at the end, such as “beat.” If you try to leave a lax vowel in a stressed open syllable (such as the made-up word “bih”), you end up with something very un-English-like. You can pronounce such a word, but it will sound like something from another language. The same is true with /ɛ/, /ӕ/, /ʊ/, and /ʌ/. You can’t really go around saying “That is veh. I appreciate your geh very much.”
Because of this restriction of not being able to appear in stressed open syllables, /ɪ/, /ɛ/, /ӕ/, /ʊ/ and /ʌ/, as in “hid,” “head,” “had,” “hood,” and “mud” are called the lax vowels of English. Most phoneticians consider the vowels /ɑ/, /i/, /u/, /e/, and /o/ to be the tense vowels. These vowels are produced more at the edges of the vowel space (less centralized) than their lax counterparts. You can hear the difference between these tense vowels and their corresponding lax member in the pairs /i/ and /ɪ/, /e/ and /ɛ/, and /u/ and /ʊ/. If you say these in pairs, you should be able to hear both a difference in quality and quantity (with the lax member being shorter in duration). The /ɑ/ and /o/ tense vowels don’t really have a lax member to pair up with (oh, well — somebody has to stay single!).
Most forms of British English have one more lax vowel than American English, /ɒ/ called turned script a in IPA. This is an open, back rounded vowel, as in RP “cod” and “common.” It can’t appear in stressed open syllables and is lax.
Sorting the Yanks from the Brits
Phoneticians focus on the sound-based aspect of language and don’t fret about the spelling, syntax (grammar), or vocabulary differences between North American and British varieties. This helps narrow down the issues to the world of phonetics and phonology.
In terms of vowels, you need to consider other issues than just the presence or absence of rhotics. There are quality differences in monophthongs as well as different patterns of diphthongization depending on which side of the pond you live. These sections take a closer look at these differences.
Differentiating vowel sounds
For front vowels (ranging from /i/ to /ӕ/), both North American English and British English have sounds spaced in fairly equal steps (perceptually). You should be able to hear this spacing as you pronounce the words “heed,” “hid,” “head,” and “had.” Try it and see if you agree.
Things get testy, however, with the vowel /e/. English /e/ is transcribed as /eɪ/ by many phoneticians (especially in open syllables) because this vowel is typically realized as a diphthong, beginning with /e/ and ending higher, usually around /ɪ/. This is shown in a traditional vowel quadrilateral (Figure 7-2a). Overall, the amount of diphthongal change for American /eɪ/ is less than that found for the major English diphthongs /aɪ/, /aʊ/, and /ɔɪ/.
Illustration by Wiley, Composition Services Graphics
Figure 7-2: Vowel quadrilateral showing different offglides used for varieties of GAE (both a and b) and British English (c).
However, talkers vary with respect to where they really start from. Fine-grained studies of American English talkers suggest that many people start from lower vowel positions, producing words like “great” as /gɹɛɪt/. The trajectory of this diphthong is shown in Figure 7-2b. Forms of English spoken in the United Kingdom have different trajectory patterns. The direction of the /e/ diphthong for RP is similar to the direction of the GAE /eɪ/, but extends slightly further (not shown in figure).
Other British dialects have larger diphthong changes, including London accents sometimes called Estuary English (see Chapter 18). These upstarts (named for people living around the Thames, not birds), produce /seɪ/ “say” sounding more like /saɪ/. A panel showing the diphthong trajectories of this accent is shown in Figure 7-2c. Not to be outdone, the Scots arrive at a vowel like the Japanese, doing away with a diphthong altogether and instead producing a high monophthong that can be transcribed [e]:
“Which way (should we go to Lochwinnoch?) [ʍɪtʃ we:] . . .
There are even more North American versus British differences in the mid and back vowels. Starting with the mid vowel /ʌ/, British speakers produce the vowel lower than their North American counterparts. This is likely due to the fact that British talkers distinguish words like “bud” and “bird” by distinguishing between low /ʌ/ and the higher mid-central vowel /ɜ/. However, North American talkers use a rhotic distinction (/ʌ/ versus /ɝ/) and don’t require this height separation.
North American talkers show regional differences among the back vowels, particularly for the notorious pair /ɑ/ and /ɔ/. The tendencies are either to merge the two toward /ɑ/ (Southern California) or closer to /ɔ/ (Northern American dialects). Most speakers of British English have added another vowel to the mix: high back rounded /ɒ/.
Table 7-4 shows some examples of these British back vowel distinctions so you can get grounded in the differences. This may be especially helpful if you’re interested in working on accents for acting, singing, or other performance purposes. (I also include URLs where you can listen to audio files.)
These differences provide an insight into the challenges facing people trying to master new accents. Namely, it’s difficult moving from an accent with fewer distinctions (such as no difference between /ɑ/ and /ɔ/) to an accent with more distinctions. This is not only because the learner must use more sounds but also because the distribution of these sounds isn’t always straightforward.
For example, British RP accent uses an /ɑ:/ sound for many words that American English uses an /æ/ for. For instance, “glass” and “laugh” (/glɑ:s/ and /lɑ:f/). However, speakers of RP pronounce “gas” and “lamp” the same as in GAE, with /ӕ/. Thus, a common mistake for GAE speakers attempting RP is to overdo it, producing “gas” as /gɑ:s/. Actually, there is no easy way to know which RP words take /ɑ:/ and which take /ӕ/, except to memorize.
Notice that it’s not as tricky to go in the opposite direction, from more accent distinctions to less. For example, a British RP speaker trying to imitate a California surfer could simply insert an /ɑ/ vowel for “bomb,” “balm,” and “bought” and probably get away with it. But could that British person actually surf?
English has a diphthongal quality to the tense vowels /e/, /i/, /o/, and /u/, particularly in open syllables. For this reason, these vowels are often transcribed /eɪ/, /ij/, /oʊ/, and /uw/ (see also Chapter 2).
Dropping your “r”s and finding them again
Rhotic and non-rhotic accents are a bit more complicated than is indicated in the “Coloring with an ‘r’” section, earlier in the chapter. Many of the nonrhotic accents (they don’t pronounce an “r” at the end of a syllable) express an /ɹ/ under certain interesting circumstances.
A linking-r occurs if another morpheme beginning with a vowel sound closely follows nonrhotic sounds. This is typical of some British accents, but not American Southern States. Here are a couple examples.
|
Example Word |
British SE |
American Southern States |
|
care |
/keə/ |
/keə/ |
|
care about |
/ˈkeə˞ əbaʊt/ |
/ˈkeəɁ əbaʊt/ |
A similar-sounding process is intrusive-r, the result of sound rules trying to fix things that really aren’t broken. For these cases, such as law-r-and-order, an “r” is inserted either to fix the emptiness (hiatus) between two vowels in a row, or to serve as a linking-r that was never really there in the first place (for example, if “tuna oil” is pronounced “tuner oil”). Table 7-5 shows some examples. I also include URLs where you can listen to audio files.
Table 7-5 Examples of Linking-r
|
Phrase |
IPA |
URL |
|
Australia or New Zealand |
/ɒsˈtɚɪlɪɚ ɔ:nju: ˈzi:ln̩d/ |
|
|
There’s a comma after that. |
/ðəzə ˈkɒmɚ ɑ:ftə θæt/ |
|
|
Draw all the flowers |
/drɔ:ɹ ˈɔ:l ðə flaʊəz/ |
Noticing offglides and onglides
There are a number of different ways to describe the dynamic movement of sound within a vowel. One way, as I describe in Chapter 2, is to classify vowels as monophthongs, diphthongs, or triphthongs. This description takes into account the number of varying sound qualities within a vowel. Phoneticians also note which part of the diphthongs (the end or the beginning) is the most prominent (or unchanging). This distinction is commonly referred to as offglides and onglides:
Offglides: If the more prominent portion is the first vowel (as in /aɪ/), the second (nonsyllabic) part is the offglide. This idea of an offglide also provides a handy way to mark many types of diphthongs that you may find across different accents. For instance, in American Southern States accents, lax /ӕ/ becomes /ɛə/ or /eə/. That is, they are transcribed including a /ə/ offglide. Table 7-6 shows some examples with URLs to audio files.
Table 7-6 Vowels Produced with an Offglide
|
Example Word |
IPA |
URL |
|
lamp |
/leəmp/ |
|
|
gas |
/ɡeəs/ |
Some phoneticians denote an offglide with a full-sized character (such as /eə/), while others place the offglide symbol in superscript (such as /eə/).
Onglides: An onglide is a transitional sound in which the prominent portion is at the end of the syllable. These sounds begin with a constriction and end with a more open, vowel quality.
An example of an onglide in English would be the /j/ portion of /ju/. Some phoneticians treat this unit as a diphthong, while a more traditional approach is to consider this syllable a combination of an approximant consonant followed by a vowel.
Doubling Down on Diphthongs
American English and British English accents have in common a set of three major diphthongs, /aɪ/, /aʊ/, and /ɔɪ/. These are called closing diphthongs, because their second element is higher than the first (the mouth becomes more closed). You can see the three major diphthongs (similar in GAE and British English) in Figure 7-3a, and a minor diphthong (found in British English) in Figure 7-3b. The /aɪ/, /aʊ/, and /ɔɪ/ diphthongs are also called wide (instead of narrow) because they involve a large movement between their initial and final elements.
Illustration by Wiley, Composition Services Graphics
Figure 7-3: Diphthongs found in both GAE and British English (a), and in only British English (b).
Considering first /aʊ/, as in “cow,” a similar trajectory is seen in BBC broadcaster English as in GAE. The /aʊ/ diphthong is also called a backing diphthong because posterior tongue movement is involved when moving from /a/ to /ʊ/. As may be expected, there are many variants on this sound, especially in some of the London accents (which can sound like gliding through /ɛ/, /ʌ/, /u/ or /ӕ/, /ə/, and /ʊ/).
The /aɪ/ sound is a fronting diphthong. An important thing to remember about this sound is that few talkers will reach all the way up to a tense /i/ for the offglide; it’s usually /ɪ/. A second fronting diphthong found in British English and American English accents begins in the mid back regions. This is the diphthong /ɔɪ/, as in “boy,” “Floyd,” and “oil”.
An interesting diphthong found in British accents (but not in GAE) is the closing diphthong /əʊ/. Look at the dotted line in Figure 7-3b. This sound is found in place of the GAE tense vowel /o/. Because it doesn’t have much of a sound change, it would qualify as a narrow diphthong. Table 7-7 shows some examples. You can also check out the audio files.
Lengthening and Shortening: The Rules
This section concentrates on vowel length, namely how a given vowel’s length changes as a function of context. Such context-conditioned change is called allophonic variation (see Chapter 5 for more information).
If you’re an English speaker, you naturally carry out at least three subtle timing changes for vowels when you speak. Here, I note these processes formally as rules. This information can come in handy if you teach English as a second language, compare English to other languages, or engage in any work where you need to be able to explain what the English sound system is doing (instead of, say, stamping your feet and saying “because that’s just how it is!”).
Check out each rule and its examples:
Rule No. 1: Vowels are longest in open syllables, shorter in syllables closed by a voiced consonant and shortest when in syllables closed by a voiceless consonant. For example:
“bay” (/beɪ/)
“bayed” (/bed/)
“bait” (/bet/)
Rule No. 2: Vowels are longer in stressed syllables. For example:
“repeat” (/ɹəˈpit/
“to repeat” (/ˈɹipit/)
Here, “peat” (/pit/) should sound longer in the first than the second example.
Rule No. 3: Vowels get shorter as syllables are added to a word (up to three syllable-words). For example:
“zip” (/zɪp/)
“zipper” (/ˈzɪpɚ/)
“zippering” (/ˈzɪpɚɪŋ/)