1 Topographic Surface Anatomy. STUDY AIMS. At the end of your study, you should be able to: Identify the key landmarks. This approach is especially advantageous for stu- dents w h o begin their English studies with phonetics and phonology before taking other, more general. CAMBRIDGE English Phonetics and Phonology A practical course Peter Roach Fourth edition English Phonetics and Phonology A practical course English.
|Language:||English, Spanish, Hindi|
|ePub File Size:||20.45 MB|
|PDF File Size:||20.54 MB|
|Distribution:||Free* [*Regsitration Required]|
English Phonetics and Phonology. A practical course. Fourth edition. PETER ROACH. Frrieritus Prºfessor of Phonetics. University ºf Reading. º CAMBRIDGE. Library of Congress catalogue card number: 90– British Library cataloguing in publication data. Roach, Peter English phonetics and phonology: a. English Phonetics and Phonology: A practical course by Peter Roach has been a leading coursebook English phonetics and phonology.
The notion was later also introduced into morphology, where it is now an essential element. The only difference b e t w e e n them is that the English vowel is produced with the lips in a neutral shape, and the G e r - man vowel with rounded lips. The lips are slightly spread. Practise saying hahahaha - alternating between this state of the vocal folds and that described in iii below. IPA symbols representing phones or allophones , but also phonemic symbols i.
The difference in meaning between zeal and seal can be solely attributed to the dif- ference between their initial sounds, and it appears that the only difference between these two sounds is one of voicing.
Phonologically relevant features: Distinctive features Intensity of articulation 1: Voiced sounds are usually made with a relatively weak breath force, or little muscular tension. This is called a lenis articu- lation [Latin, 'soft'; ungespannt]. This is called a fortis articulation [Latin, 'hard'; ge- spannt]. You may occasionally hear the terms "soft" and " h a r d " in popular usage to describe speech sounds or their articulation as in "soft s" or "hard s" , but they have been abandoned in linguistics.
Whereas fortis sounds are indeed always voiceless in English, lenis sounds, which are usually voiced, may also occur as voiceless variants, i.
We have already seen one example in the discussion of zeal and seal above. Why, then, do we still perceive a difference between zeal and seal even when the words are whispered? There is another differ- ence, namely the intensity with which the initial sounds are articulated: Although the initial sound in zeal is made voiceless when whispered, it retains its lenis articulation.
In other words, the first sound in zeal is always articulated with a weaker breath force than the first sound in seal, no matter whether the words are whispered or not. Some lenis sounds can also be devoiced in certain environments. For example, they are partly devoiced in word-initial position, and almost entirely devoiced word-finally [Auslautverhartung], as in the words cab and serve.
The final sounds, here, are devoiced, but we still perceive them as the same voiced sounds because of their lenis articulation. We still hear the words cab and serve. If, however, we increase the breath force, or muscular tension, when producing the final sounds, i. Place of articulation!
We already know that virtually all English sounds are made with air that is pushed up from the lungs. In the production of approximately t w o thirds of these sounds, the air-stream is obstructed in the throat, technically called the p h a r y n g e a l cavity or p h a r - ynx [Rachenraum, Racben], or in the vocal tract before it leaves the body through the mouth or nose. An important feature for the de- scription of consonants is the exact place where the air-stream is obstructed.
The place of articulation [Artikulationsstelle, Artikulationsort] names the speech organs that are primarily involved in the production of a particular sound.
To produce a consonant, there is usually one active, mobile, lower speech organ that moves and makes contact with a passive, immobile, upper speech organ. For example, in the articulation of the last sound in the w o r d surf, the air-stream is obstructed by the contact of the lower lip with the upper teeth. This sound is therefore called a "labiodental consonant", or simply a "labiodental" [from Latin labialis, 'of the lips', and dentalis, 'of the teeth'].
You will find a diagram showing all the speech organs re- ferred to in this manual on the inside front cover. Thus the manner of articulation [Artikn- lationsart] refers mainly to the degree to which the air-stream is obstructed at the p l a c e of articulation of consonants. When pronouncing the last sound in surf, for e x a m p l e , the gap between the lower lip and the upper teeth is narrowed to the point w h e r e friction is caused as the air passes through.
The resultant consonant is therefore called a "fricative". If we wanted to describe this sound using all three distinctive features, w e would say that it is a "fortis labiodental fricative". There is no other sound in the E n g - lish sound system that fits this description. The places and manners of articulation will be discussed in greater detail in the n e x t lesson, and we shall return to the intensity of articulation in Lesson Four. Exercises 15 EXERCISES These discovery exercises are designed to help you discover the way the sounds of Eng- lish are made, and h o w some of the features discussed in this lesson contribute to the distinction of word meaning.
Exercises 2. Try the exercises at least twice to really familiarise yourself with the feature that is being considered. If you place stress at the beginning of some words, or alternatively towards the end, their meaning changes.
Try the words below, noting whether they are verbs or nouns ac- cording to where the stress is placed. Can you identify any regularity? She did it. Have you finished? It's your turn to pay.
Actually, I don't like whisky. We're not going there again. These exercises illustrate the function of loudness and pitch as suprasegmental features. Although they can change meaning, it is important to realise that this occurs on a suprasegmental, and not a segmental, level.
In each pair, identify which word has the longer middle sound. In the example just given, it is cheap. N o w try to reverse the lengths. So make the middle sound in cheap short, and the one in chip long. Does the swapping of length swap the meaning of the words, or is the difference in the quality of the middle sounds the cue which distinguishes these aberrant forms from each other?
There are two sets of words focusing o n e i - ther the sound at the beginning or the sound at the end of the word. For example, t h e initial sound in do is voiced whereas the one in to is voiceless; the final sound in bag i s voiced whereas the one in back is voiceless.
Lenis and fortis Try saying the following pairs of words while holding your hand in front of y o u r mouth, or placing a small, light, flat object, such as a dried leaf or a feather, on the p a l m of y o u r hand held near y o u r mouth: Which ones are they? This difference can still be noticed e v e n when you whisper the words. This exercise demonstrates the difference in the i n t e n s i t y of the articulation, i.
Notice how there is a build-up of air pressure in the mouth. Which parts o f y o u r m o u t h are doing the most work to hold back the pressure? When you release t h e sound and say the word, which parts move? Try the same with tan and can.
N o t i c e where the sound is made in your mouth. Where d o you feel a stream of air passing through? Exercises 17 2. C o m - pare this with the prolonged final sounds in ram, ran, rang. Think about the places where you can feel the sounds being made. What parts of the vocal tract are in contact or close proximity?
These three exercises show you some of the variety of place of articulation. There are many other places of articulation, which we shall discuss in the next lesson. Notice with each set that the sounds are made in approximately the same place, but your mouth is doing different things, i. Can you describe - even if only impressionistically - what is happening in these different articulations? Think about whether the articulators are in contact, or are close together.
Where is the air-stream flowing through? N o w hold your articulators in these positions, and try to blow out some extra air without either open- ing or closing your mouth.
You should notice that the air in the words mood and nude passes through the nose, while in lewd and rude it passes through the mouth. In fact, with the latter pair, you can also feel that the air passes over the side of the tongue in lewd, and over the centre of the tongue in rude.
In the next lesson, we shall discuss h o w these and other articulations can be described formally, and give them their technical names.
The phoneme is a concept used in phonology, which is why the subdiscipline is sometimes calledphonemics in the US. We saw in Lesson Two that a single phoneme can distinguish the word cab from cap, serve from surf, fool from full, and zeal from seal.
A phoneme can therefore be defined as the smallest distinctive, or contrastive, unit [kleinste bedeutungsunterscheiden. To put it in other words, a phoneme con- trasts meaningfully with other speech sounds. We also saw in Lesson Two that the con- trast between two phonemes lies solely in those characteristics that are phonologically relevant, and that it is therefore sufficient to describe phonemes only in terms of their distinctive features.
Thus a phoneme has also been defined as a bundle of distinctive features illustrated nicely by labels such as "fortis labiodental fricative". The various definitions emphasise different aspects of the phoneme, but they all mean more or less the same thing. It is important to remember that phonemes are abstract, idealised sounds that are never pronounced and never heard. Actual, concrete speech sounds can be regarded as the realisation of phonemes by individual speakers, and are referred to as phones [from Greek phone, 'voice'].
The phone, then, is a concept used in phonetics. Strictly speaking, they are then phonemic symbols, rather than phonetic symbols, but unfortunately this terminological distinction is not always observed.
Phones, the true phonetic symbols, occur in square brackets, [ ]. If we want to establish what phonemes there are in a sound system, also called a phonemic system or phoneme inventory [Phoneminventar], we need to find pairs of words that differ in meaning and in only-one sound.
Linguists do this, for example, when they record a previously unknown language. Each of the two contrasting sounds in such a minimal pair [Minimalpaar] is a distinct phoneme. We have shown, then, that the final sounds in cab and cap are phonemes because the two words are a minimal pair.
N o t e that orthography is absolutely irrelevant here: T h e spelling of week and weak, o n the other hand, differs in only one letter, but the two words are pronounced i d e n t i c a l l y and are therefore not a minimal pair. Every language has its o w n phoneme inventory, of course, but the phonemes s o m e - times also vary from dialect to dialect or from accent to accent. The phoneme i n v e n t o r y of most American English accents, for example, does not include the sound that m o s t British speakers pronounce as the second sound in the word shop.
Instead, A m e r i c a n speakers usually use sounds with the quality of the second sound in father or the s e c - ond sound in saw. T h e English c o n s o n a n t p h o n e m e s We already k n o w that consonants are sounds that are produced by an obstruction of a n air-stream either in the pharynx or in the vocal tract.
There are 24 consonant p h o n e m e s in R P and in most other accents of English. T h e aim of this lesson is to describe t h e s e consonant phonemes to a point where there is only one phoneme in the sound s y s t e m of English that fits each description, as we did with the final sound in surf'm L e s s o n Two.
We have learnt that loudness, pitch, tone of voice, and duration are suitable c r i t e - ria only to describe concrete speech sounds, or phones, but these features cannot d i s - tinguish phonemes. Length and voicing have proved to be rather unreliable f e a t u r e s , and since all English sounds are made with egressive pulmonic air, the air-stream m e c h - anism is not a suitable criterion either.
For the description of English consonant p h o - nemes, we therefore use only the distinctive features: Place of articulation II There are thirteen possible places of articulation in the languages of the world, but n o t all of them are utilised in English.
They are usually labelled according to the i m m o b i l e , upper speech organ used in their production.
The mobile, lower speech organ a l w a y s lies directly opposite. In this section, all the places of articulation, progressing from t h e lips to the glottis, are briefly described, and the relevant English consonant p h o n e m e s for each category are given as IPA symbols.
T h e underlined letters in the e x a m p l e words illustrate the sound a symbol represents. Remember that there is a list of all t h e phonetic symbols used in this manual and a diagram showing all the relevant speech o r - gans on the inside front cover.
Consonants 21 The bilabials and labiodentals form one larger group, the labials, because they all make use of the lips. The two dentals in English are often popularly called "teeaitch" because of their spelling. There is only one postalveolar in English, namely the lenis Ixl as in red. There are no retroflex phonemes in R P or any other accent of English.
There is, however, a retroflex pronunciation variant pertaining to parole or performance of the Ixl phoneme in most American accents, in Irish English, and in ac- cents of south-west England in words like worse and hard.
The lenis 1 1 as inj. A n example from another language is the final sound in the H i g h German pronunciation of the word ich, transcribed as [c]. An example from another language is the last sound in the High German word ach, transcribed as [x].
The lips, then, are a secondary place of articulation. A place of articulation which adds some quality to the main articulation i s called secondary articulation. Some linguists speak of coarticulation or double a r t i c - ulation, but these terms are usually reserved for the simultaneous use of two places o f obstruction of equal importance, which hardly ever occurs in English.
There are n o uvular phonemes in English. There are no pharyngeal consonant phonemes in English. T h e palatal, the velar, the uvular, and the pharyngeal sounds are grouped together a s dorsal sounds because they all use the b o d y of the tongue [from Latin dorsalis, 'of t h e back']. Such sounds do not exist in English. T h e glottal stop, [? In some non-standard British accents, it is a pronunciation variant of the Itl phoneme in certain phonetic environments, as in t h e words better and butter.
Manner of articulation II T h e manner of articulation, as we learnt in Lesson Two, refers to the type or degree o f closure of the speech organs. There are eight different manners of articulation that u s e an egressive pulmonic air-stream mechanism, plus an additional one that has b e c o m e a conventional category in the description of some languages, including English.
In t h i s section, we look at all nine categories, progressing from total closure of the speech o r - gans to a fairly wide opening between them. The first element is a plosive. This means that affricates, too, require a complete closure in the vocal tract, b u t the air is released slowly enough to produce friction, which we hear as a hissing s-like sound.
This second element is articulated in the same place, i. There are t w o affricates in English: An affricate, then, is represented by one symbol consisting of two characters. There is no compelling reason why we should analyse the two elements of an affricate as a single phoneme.
They stand out from all other English phonemes, however, in that the velum, or soft palate, is lowered, so that air escapes through the nose. In the production of English nasals, usually all the air escapes through the nose.
O t h e r languages have nasals where some air also passes through the mouth, as in the final sound in the French word bon. T h e three English nasals are all lenis sounds: All other English phonemes are usually produced with the velum raised, so that the passage to the nasal cavity [Nasenraum] is blocked, and the air escapes only through the mouth.
In order to distinguish them from nasals, these sounds are sometimes called orals. We cannot actually feel our velum moving, but there is a simple test that shows us whether the velum is lowered or raised, and the effect that the position has on the sound quality: While you are pronouncing one of the three English nasals, stop your nose and release it again.
You will hear h o w the quality changes when the air-stream through the nose is blocked. Try the same with any other English sound and you will find that the sound quality does not change a bit. Because in oral sounds, the passage through the nose is already blocked at the velum. Another terminological distinction that is sometimes made and that we should mention at this point is the contrast between non-continuant and continuant sounds. Non-continuants are produced with a complete closure of the speech organs.
Plosives and affricates are non-continuants because the passage through both the mouth and the nose is blocked. By contrast, all other speech sounds, including nasals, are made with- out a complete closure of the speech organs, and are therefore called continuants [Dauerlaute].
Some linguists count English nasals among the non-continuants, h o w - ever, because the passage through the mouth is always closed. Rolls are produced when one articulator vibrates against another. There are no rolled phonemes in R P or any other accent of English, but some dialects have a rolled pronunciation variant of the Ixl p h o - neme.
The typical Scottish Ixl, for example, is produced by a vibration of the tongue against the alveolar ridge. Such an alveolar roll sometimes also occurs in stylised speech, for example on stage. The phonetic symbol for this pronunciation variant is the same as the one for the underlying phoneme, namely [r].
For example, in s o m e accents of British English, including RP, the Ixl phoneme in words like very is s o m e - times realised as a single flap of the tongue tip against the alveolar ridge.
Such an alveolar flap, transcribed as [r], is also very c o m m o n i n American English, where it is a pronunciation variant of the III and Idl p h o n e m e s i n words like latter and ladder, which are then pronounced identically. The fricatives fall i n t o two subcategories, slit fricatives and groove fricatives, according to the w i d t h a n d depth of the air passage.
There are four slit fricatives [slit, German 'Schlitz'] in English: The fortis Ihl a s i n hat is usually also grouped together with the slit fricatives, but is best described a s a cavity fricative as it has no point of narrowing. Groove fricatives are made by forming a groove [German 'Rinne, Furche'] a l o n g the front part of the tongue [Vorderzunge].
They are made with more intensity t h a n t h e slit fricatives, and have a sharper, 5-like sound. They are therefore also called s i b i l a n t s [from Latin sibilans, 'hissing']. The second element in affricates, which we d i s c u s s e d above, is a groove fricative, or sibilant. There are no lateral fricatives in R P 0 r any other accent of English. The three English a p p r o x i m a n t s are all lenis phonemes: It should be noted here that no other consonant phoneme of English is as variable i n its actual pronunciation as the Ixl phoneme.
It has several different realisations, t h r e e o f which we have already encountered in this lesson. O n l y one of these three, the r e t r o f i t is also an approximant, like the underlying phoneme. The manner of the a r t i c u l a - tion of the other two pronunciation variants is not the same as that of the u n d e r l y i n phoneme: As we already know, [r] is a roll, and [r] is a flap.
The consonant table We can now use the three distinctive features to describe all English consonant p h o - nemes. It is easier, however, if we first arrange the relevant phonetic symbols in a coor- dinate system with the places of articulation on the horizontal axis, the manners of ar- ticulation along the vertical axis, fortis sounds positioned on the left side of the grid squares, and lenis sounds on the right.
You will find such a table containing all English consonant phonemes on the inside back cover of this manual. But the table reveals much more than that. It shows that sixteen consonant phonemes form pairs within which the only distinguishing feature is the intensity of the articulation.
Furthermore, the consonant table makes it easy to see that other sounds are differentiated by two distinctive features. Finally, the table also shows that there are sounds that do not share a single distinctive feature. The p r o b l e m c a s e s Of semi-vowels, contoids, and vocoids All consonants generally have two things in common: By contrast, the sounds that a are produced without any obstruction of air, and b usually occur at the centre of syllables are called vowels.
The English frictionless continuants, i. We have so far regarded them as con- sonants because they always appear at the margins, and never at the centre, of syllables. This can be illustrated by words like lot, car, yes, and wax, and by the fact that tip or trp, for example, are not possible words in English.
We learnt in Lesson O n e that phonol- ogy, more precisely segmental phonology, is concerned with the function and possible combinations of sounds. O n the other hand, they are produced w i t h a l - most n o obstruction of air. F r o m a purely phonetic point of view, then, the frictionless c o n t i n u a n t s are almost vowels. In order to reflect their intermediate status, all frictionless continuants may b e r e - garded as semi-vowels [Halbvokale] or semi-consonants, of course, but this t e r m i s hardly ever used.
They are sometimes also called glides [Gleitlaute] because, w h e n a r - ticulating these sounds as parts of actual words, the tongue moves in gliding fashion e i - ther towards or away from a neighbouring vowel. Alternatively, we can reserve the traditional terms consonant and vowel t o r e f e r only to the phonological properties of sounds. This would mean that we refer t o t h o s e sounds that typically occur at the margins of syllables as consonants, and to those t h a t usually occur at the centre as vowels.
In order to refer to the phonetic properties, w e then need another set of terms. We can call sounds that are produced with an o b s t r u c - tion of air c o n t o i d s rather than consonants , and those that are produced w i t h o u t a n y obstruction vocoids rather than vowels. Thus all consonants except the f r i c t i o n l e s s continuants are contoids, all frictionless continuants are vocoids, and so are all v o w e l s. We should note, however, that this terminology has not been universally a d o p t e d.
Most linguists use the concept of semi-vowel, or glide, and subsume this category u n - der the broader consonant category, as we did above.
This is done in order to make it e a s - ier f o r ' y o u to read a wide range of linguistic texts with varying theoretical a n d terminological approaches. It is important to be aware that not all linguists use all t h e terms introduced in this lesson in the same way. I n a broader sense, the term sometimes covers the same group of sounds that w e c a l l frictionless continuants, i. It is a traditional term that is not o f t e n used anymore, and should best be avoided.
Pay attention to the sensation of the "extra" air forced out as you say the sounds represented by the underlined letters. The first w o r d in each pair has a lenis fricative, with less in- tense articulation. Pairing of the words, and thus comparing individual sounds in those words, demonstrates the difference in intensity of articulation between lenis and fortis sounds.
In exercise 3. Which parts of your m o u t h are doing the most w o r k to hold back the pressure? When you do release the sound and say the word? Which parts move?
Try the same with [t] in tin and [k] in king. These will show you how the place where an obstruction is made varies. Describe the places of ar- ticulation of these three sounds. Where do you feel a stream of air passing through?
Try saying the sounds  as in the, and [z] as in zip. Where do you feel the stream of air now? This again shows variation in place of articulation. Describe the place of articulation of these three sounds. G r o u p the sounds represented by the underlined let- ters in the words below according to the degree of obstruction, using the terms you have learnt in this lesson. When you have completed this task, compare y o u r arrangment with the solution at the end of this manual, and then with the English c o n - sonant table on the inside back cover.
Which features intensity, place, or manner distinguish t h e sounds and hence the words from each other? Write the IPA symbol for each of t h e s e sounds.
Think about how the sounds d i f f e r and write the IPA symbol for the sounds. Can you think of any other sets of m i n i m a l pairs which show h o w sounds contrast meaning? F o r example, dogs should be transcribed [dgz].
D o a set, then check y o u r a n - swers before you do the next set. Thus we already k n o w that, a phonetically, vowels are produced without any obstruction of air, and, b phonologically, vowels usually occupy the centre of a sylla- ble.
Even though most languages have over twice as many consonants as vowels, in a way, vowels can be seen as predominant: T h e y carry most of the loudness, pitch, and tone of voice that we perceive in concrete utterances, and since their sound quality var- ies considerably from region to region, vowels make up most of the characteristics that distinguish different accents of the same language.
Different accents may even have dif- ferent vowel systems. T h e typical Scottish pronunciation of English, for example, has only 10 vowel phonemes whereas R P has T h e predominance of vowels is also re- flected in the origin of the w o r d consonant: It comes from the Latin w o r d consonans, which means 'sounding together'.
This implies that consonants d o not comfortably oc- cur alone. They usually sound together with vowels. What are the criteria that we use to describe English vowel phonemes? We noted in Lesson Two that loudness, pitch, tone of voice, and duration are suitable criteria only to describe concrete speech sounds, or phones, b u t these features cannot distinguish phonemes.
Length has proved to be a rather unreliable feature, and so has voicing. It is important to note, here, that all English vowels are typically voiced, so that voicing would not count as a distinctive feature anyway.
T h e same is true of the air-stream mechanism since all English sounds are made with egressive pulmonic air. Even the place of articulation, which is one of the three distinctive features for the description of consonant phonemes, is of no relevance here because the air-stream is not obstructed by the speech organs. The intensity of articulation, on the other hand, does contribute to the distinction between vowel phonemes, as we shall see later, but it is not nearly as important as it is for the description of consonants, and most linguists therefore neglect this feature altogether.
What we are left with, then, and what we solely rely on is the manner in which the English vowels are articulated. In our discussion of the manners of articulation of con- sonants in the previous lesson, we moved from total closure of the speech organs to a narrowing between them. Since vowels are produced without any obstruction of air, none of the nine categories in that progression applies here.
It follows that all English vowels are orals, and that this feature can t h e r e f o r e not distinguish different vowel phonemes either. But apart from the raising of t h e v e - lum, the different manners of vowel articulation also involve slight movements of t h e tongue and lips, as we shall see below.
These are the three criteria for the description of vowel phonemes. If the tongue is high, as in the last sound of the word bee, it is close t o t h e palate, and we therefore speak of a close vowel [geschlossener Vokal]. If the t o n g u e i s low, as in the third sound of the word starling, the gap between it and the palate is m o r e open, and we speak of an open vowel [offener Vokal].
Between these extremes, t h e r e are three intermediate levels: If the tongue is in a mid-high position, i. If it is mid-low, i. A vowel that is made with a tongue height s o m e - where between mid-high and mid-low is simply called a mid vowel. If the back of t h e tongue is raised highest in which case the b o d y of the tongue is pulled back , as in t h e middle sound in goose, the resultant sound is a back vowel [Hinterzungenvokal]. Be- tween these extremes, we recognise one intermediate position: If the centre of t h e tongue is raised highest, as in the second sound of the word bird, we speak of a c e n t r a l vowel [Mittelzungenvokal].
English does not u t i - lise this contrast very much. As in most other languages, the spreading of the lips u s u - ally correlates with frontness, and lip-rounding with backness.
This means that t h e r e are n o t w o vowel phonemes in English that differ only in the shape of the lips. M a n y linguists therefore d o not regard this criterion as relevant in English. Both sounds are mid central vowels, i. The only difference b e t w e e n them is that the English vowel is produced with the lips in a neutral shape, and the G e r - man vowel with rounded lips. Vowels 33 The vowel chart I Since the spreading or rounding of the lips cannot distinguish vowel phonemes in Eng- lish, there are only two distinctive features for the description of all English vowels: Just as we have arranged the symbols for consonant phonemes in a table, we can also arrange the symbols for vowels schemati- cally on the basis of these two criteria.
Depending on the particular language under de- scription, the vowel symbols are superimposed on the space within a triangle or a quad- rilateral. Such a vowel c h a r t , or vowel diagram, was first devised by Daniel Jones. It reflects roughly the space in the centre of the m o u t h , where the vowels are articulated.
You will find a vowel chart containing the symbols for all RP vowel phonemes on the inside back cover of this manual, and it is recommended that you refer to it as you con- tinue reading this lesson.
The closeness or openness of a vowel is shown by the vertical position of the sym- bol in the vowel chart: The higher the symbol, the closer the tongue is to the palate when articulating the corresponding sound. Conversely, the lower the symbol, the more open the gap between the tongue and the palate. In other words, close vowels oc- cupy the upper part of the vowel chart, and open vowels the lower part. Two horizontal lines mark the mid-close and mid-open positions.
Frontness or backness is indicated by the horizontal position of the symbols: The further left the symbol, the more front the part of the tongue that is raised highest when articulating the corresponding sound. T h u s the symbols on the left of the vowel chart represent front vowels. The further right the symbol, the more back the part of the tongue involved. The symbols on the right, then, represent back vowels. It goes with- out saying that the vowels in the central area of the chart are central vowels.
T h e vowel systems of most languages of the world can be represented by symbols that are evenly distributed within the vowel chart. This phenomenon is called vowel dispersion. Most vowel systems are arranged within a triangle. English belongs to the less than 10 per cent of the languages whose vowel systems have a more or less quadri- lateral shape. The vowel chart is then sometimes called a vowel quadrilateral [Vokal- viereck].
More specifically, the English vowel chart has the form of a trapezium, which is reflected in the German term Vokaltrapez, but there is no equivalent term used in English. The cardinal vowels In order to describe the vowels of any given language, and compare the vowel systems of different languages more precisely than is possible by using only the distinctive fea- tures, Daniel Jones invented 18 reference vowels, called cardinal vowels [Kardinal- vokale].
They illustrate the extremes of vowel quality that the vocal tract is able to p r o - duce.
It is important to understand that the cardinal vowels are not sounds of a particular language. Phoneticians and only they have to learn to recognise and articu- late these artificial sounds, so that they can describe all natural vowels in relation to the nearest cardinal vowel. A vowel chart with all cardinal vowels is given in Figure 3.
They are numbered counter-clockwise, beginning in the upper left corner. The vowels 1 to 8 are the primary cardinal vowels. They can be described as close front, mid-close front, mid-open front, open front, open back, mid-open back, mid-close back, and close back.
We have already mentioned that, in English and most other lan- guages, the spreading of the lips usually correlates with frontness, and lip-rounding with backness. This is also roughly true of the primary cardinal vowels: Vowels 1 to 5 are produced with unrounded, and vowels 6 to 8 with rounded lips.
For this reason, the primary cardinal vowels generally sound somewhat familiar to speakers of European languages, and they are, of course, more relevant when describing these languages. The secondary cardinal vowels are the vowels 9 to They occupy the same posi- tions as the primary cardinal vowels in the vowel chart, but they sound less familiar to us because the shape of the lips is reversed: Vowels 9 to 13 are produced with rounded, and vowels 14 to 16 with unrounded lips.
The two remaining cardinal vowels 17 and 18 are close central vowels with unrounded and rounded lips, respectively. All cardinal vowels can be represented by phonetic symbols. Unfortunately, some of these symbols are identical with the symbols that we use to represent English vowels even though the quality of the sounds is quite different.
Because the cardinal vowels are not part of a sound system of a language, their symbols are usually enclosed in square brackets, hke any other concrete sound.
It is therefore important that we gather from the context whether a vowel symbol in square brackets represents a cardinal vowel or concrete vowel of a particular language. The back of the tongue is raised b e - tween mid-close and mid-open position, and the lips are rounded.
The back of the tongue is raised s o that it almost touches the palate, and the lips are moderately rounded. A close b a c k vowel. Short vowels We now do the same for the 7 short vowel phonemes: The part of the tongue between t h e front and the centre is raised to just above mid-close position, and the lips a r e slightly spread.
A mid-close front-central vowel. The front of the tongue is raised b e t w e e n mid-close and mid-open position, and the lips are slightly spread. T h e front of the tongue is raised b e - tween mid-open and fully open position, and the lips are slightly spread. A m i d - open-open front vowel. The centre of the tongue is raised between mid-open and fully open position, and the shape of the lips is neutral. A mid-open-open central vowel. The back of the tongue is lowered to a l - most fully open position, and the lips are slightly rounded.
An open back vowel. The part of the tongue b e t w e e n the centre and the back is raised to just above mid-close position, and the lips a r e rounded. A mid-close central-back vowel. T h e centre of the tongue is raised between mid-close and m i d - o p e n position, and the lips are in a neutral shape. A mid central vowel. This last vowel, hi, is called schwa.
It stands out from all other vowels, and requires some further comment. The term schwa comes originally from Hebrew, where it means 'emptiness' and designates a H e b r e w vowel of the same quality. You may have noticed that the schwa is articulated in practically the same manner as the long vowel h: This is the only excep- tion to the general observation that a difference in length is accompanied by a differ- ence in sound quality.
We should note, however, that there is considerable variation in the way speakers produce the schwa. For example, it is usually more open in word-final position. But even if the quality of the schwa and h: Whereas h: Most unstressed syllables contain a schwa, which makes this vowel the most frequently occurring sound in English. O n e study has shown that almost 11 per cent of the sounds uttered in an English conversation are schwas.
For comparison, the most frequent consonant is Inl with around 8 per cent. Such a frequency count shows the important function that the schwa fulfils in unstressed syllables, and underlines the predominance of vowels over conso- nants. Many languages do not have a sound that occurs in most unstressed syllables. The speakers of these languages may thus find it difficult to pronounce unstressed syllables when they learn English. It should be added that some linguists analyse the schwa as a pronunciation variant pertaining to parole or performance of all English vowel phonemes in unstressed syl- lables, rather than as a phoneme in its o w n right.
The schwa is therefore sometimes called a n e u t r a l vowel, or reduced vowel. The vowel chart II If we now look at the English vowel chart again, we can easily see h o w certain vowels are similar to each other, and how they differ.
We have already noted that The vowel chart also tells us, for example, that IV. I and In: When we compare the English vowel chart with the cardinal vowel chart in Figure 3 , we can also describe the English vowel phonemes in relation to the cardinal vowels.
For example, the articulation of English III is more close and more central than cardinal vowel number 2, or [e]. Intensity of articulation II: Lax and tense We mentioned at the beginning of this lesson that the intensity of articulation also con- tributes to the distinction between vowel phonemes.
It is not nearly as important here, however, as it is for the description of consonants, where it is one of the distinctive fea- tures.
In fact, the intensity of vowel articulation has never received much attention in British linguistics. These labels correspond to the terms lenis a n d fortis, which we use to describe the intensity of consonant articulation.
Since all English vowels are typically voiced, it goes without saying that there c a n - not be a correlation between the intensity of vowel articulation and voicing, as t h e r e i s with consonants.
There is, however, a correlation between the intensity of vowel a r t i c - ulation and vowel length: All lax vowels are short vowels, and all tense vowels are l o n g. A n o t h e r interesting regularity is that all five tense, long vowels have lax, short c o u n t e r - parts that are roughly articulated in the same manner. The difference between these two sounds, in terms o f articulatory phonetics, lies merely in their length and, as we now know, in the i n t e n s i t y of their articulation.
We also know, however, that a difference in length is normally a c - companied by a difference in sound quality. T h u s the members within the r e m a i n i n g four pairs are not articulated in exactly the same way. Diphthongs and triphthongs The quality of the English long and short vowels remains relatively constant while t h e y are being p r o n o u n c e d , i. These vowels are therefore called pure or plain vowels, or m o n o p h - thongs [from Greek monopbtbonggos, 'single sound'].
More precisely, they start with a m o n o p h t h o n g , and the quality then changes towards, but never quite reaches, another m o n o p h t h o n g through a gliding movement of the tongue.
These vowel se- quences are called gliding vowels, vowel glides, or diphthongs [from G r e e k diphthonggos, 'double sound'; Doppellaut, Zwielaut].
D i p h t h o n g s can be divided into three groups: It is often replaced with h: A n d [o] is articulated exactly like h: They d o not exist in English. Another categorisation of diphthongs divides t h e m into diminuendo, descending, or falling diphthongs and crescendo, ascending, or rising diphthongs [fallende u n d steigende Doppellaute].
In falling diphthongs, the first element is longer and louder than the second. All English diphthongs are usually falling diphthongs. Rising diph- thongs, where the second element is more prominent, rarely occur in English. It is important to note that a diphthong is conventionally analysed as one vowel phoneme.
We would say, then, that words like face, voice, and mouth consist of three phonemes. We should also note, however, that the c o n v e n t i o n a l analysis of a vowel sequence as a single vowel is to a certain extent arbitrary just l i k e the analysis of a homorganic consonant sequence as an affricate. For t h i s reason, we hardly ever say that English has 20 vowels. We usually say that it has 1 2 monophthongs and 8 diphthongs. English also has typical vowel sequences that consist of three sounds.
They a r e 5 called t r i p h t h o n g s [from Greek triphthonggos, 'tripple sound ]. Unlike diphthongs, however, triphthongs are n o t analysed as separate vowel phonemes. They are interpreted as closing diphthongs followed by a schwa. The word royal, then, consists of four phonemes: This is most noticeable if the s h o r t e n e d sound is a long vowel or a diphthong.
In the second member of each pair, the duration of the sound in question is shortened through the i n - fluence of the following fortis consonant. We learnt in that lesson that we still recog- nise these sounds and we may even wrongly perceive them as voiced because of their lenis articulation. Notice how [i: With [ae, a: You could also repeat the words he and who.
D o the same with the open sounds [33, a: This time, you could repeat the words bad and bard. Again, you should be able to notice the movement of the tongue from the front of your mouth to the back.
Then categorise them as either front, back, or central vow- els. In the case of upper, which has two monophthongs, transcribe the underlined sound only. The concrete, particular forms show noticeable variation among themselves - they are different from each other - but they do not affect the underlying linguistic unit's function. This can be illustrated by a simple example: In graphology, i. A letter may appear in several different ways, depending on such factors as position within a word or sentence, typeface chosen, and handwriting style.
That the function of the grapheme is not affected is illustrated by the fact that the meaning of the w o r d car stays the same even if it is written cAr or car although the latter two are, of course, not equally acceptable. The first relationship of this kind was established in the field of phonetics and phonology in the s. The notion was later also introduced into morphology, where it is now an essential element.
Yet the application of the concept to some such areas seems questionable and is n o t undisputed, to say the least. Allophone vs. In the same lesson, we learnt that a concrete form, or realisation, of a phoneme is called a phone. When t w o or more concrete forms, or realisations, of a phoneme are compared, each of these forms is referred to as an allophone, rather than a phone.
An allophone can therefore be defined as one realisation of a phoneme among others. The terms phone and allophone, then, pertain to phonetics because they are r e - lated to parole or performance, and the term phoneme pertains to phonology because i t is related to langue or competence.
It should be pointed out, here, that the concept of allophone is not entirely n e w t o us: We have already encountered allophones in previous lessons, but have so far simply- referred to them as "pronunciation variants" or "realisations" of a phoneme.
In L e s s o n Two, for example, we briefly touched on the fact that lenis sounds, which are usually- voiced, may also occur as voiceless variants; and in Lesson Three, we had the glottal stop, [? All these variants and realisations are allophones, and w e shall look at them more systematically later in this lesson and in Lesson Eleven.
The two allophone criteria We now turn to the question of how allophones can be identified. As a starting point, it is useful to recapitulate briefly the criterion used for identifying phonemes, which w e discussed in Lesson Three: A phoneme of a language is identified through a minimal pair, i. A minimal pair shows us that each of the two contrasting sounds can differentiate words, and is therefore a phoneme. More precisely, we should say that each of the two contrasting phones can differentiate words, and is therefore a reali- sation of a separate phoneme.
Conversely, we can deduce from the foregoing that, if no minimal pair can be found for two particular phones, these phones cannot differentiate words, and are therefore allophones of the same underlying phoneme. The single most important criterion for identifying allophones, then, is the absence of a minimal pair. N o minimal pair can be found for [h] and [rj] because the former occurs only before a vowel in stressed syllables as in behave while the latter never occurs before a vowel in stressed syllables as in sing.
Yet they cannot be considered allo- phones of the same phoneme because they are too dissimilar with respect to the intensity, place, and manner of their articulation, and the resultant sound quality. Replacing one with the other does endanger intelligibility, which shows that the function of the underlying phoneme is affected.
Because of this one case, phoneticians have introduced a second necessary criterion for identifying allophones, namely that they must exhibit articulatory similarity, or phonetic similarity. Allophones in free variation If two or more allophones can replace one another, i. Some of the allophones we have already encountered in previous lessons can be used to illustrate this phenomenon. In Lesson Two, for example, in the section on the intensity of articulation, we saw that the phoneme Izl in zeal can be realised as a voiceless, or devoiced, allophone when whispered, or as a voiced allophone when pronounced "normally".
And in Lesson Three, in the section on the place of articulation, we learnt that the phoneme Iii in butter can be realised as a glottal stop, [? The examples show that the choice of one allophone rather than another may depend on such factors as communicative situation, language variety, and social class, and the example of the lisped  in German shows that the choice can even be determined by a language handicap.
Yet when we consider the wide range of possible realisations of any given phoneme even by a single speaker , it becomes clear that we owe the vast majority of allophones in free variation to idiolects or simply to chance, and that the number of such allophones is virtually infinite. These are hollow and are attached to the top of the trachea; when we breathe, the air passes through the trachea and the larynx. This point is commonly called the Adam's Apple.
Inside the "box" made by these two cartilages are the vocal folds, which are two thick flaps of muscle rather like a pair of lips; an older name for these is vocal cords. Looking down the throat is difficult to do, and requires special optical equipment, but Fig.
At the front the vocal folds are joined together and fixed to the inside of the thyroid cartilage. At the back they are attached to a pair of Fig. The arytenoid cartilages are attached to the top of the cricoid cartilage, but they can move so as to move the vocal folds apart or together Fig. We use the word glottis to refer to the opening between the vocal folds. If the vocal folds are apart we say that the glottis is open; if they are pressed together we say that the glottis is closed.
This seems quite simple, but in fact we can produce a very complex range of changes in the vocal folds and their positions. These changes are often important in speech. Let us first look at four easily recognisable states of the vocal folds; it would be useful to practise moving your vocal folds into these different positions.
The vocal folds are wide apart for normal breathing and usually during voiceless consonants like p, f, s Fig. Your vocal folds are probably apart now. If air is passed through the glottis when it is narrowed as in Fig.
The sound is not very different from a whispered vowel. It is called a voiceless glottal fricative. Fricatives are discussed in more detail in Chapter 6. Practise saying hahahaha - alternating between this state of the vocal folds and that described in iii below.
When the edges of the vocal folds are touching each other, or nearly touching, air passing through the glottis will usually cause vibration Fig.
Air is pressed up from the lungs and this air pushes the vocal folds apart so that a little air escapes. This opening and closing happens very rapidly and is repeated regularly, roughly between two and three hundred times per second in a woman's voice and about half that rate in an adult man's voice. The vocal folds can be firmly pressed together so that air cannot pass between them Fig.
When this happens in speech we call it a glottal stop or glottal plosive, for which we use the symbol?. You can practise this by coughing gently; then practise the sequence a?
The normal way for this airflow to be produced is for some of the air in the lungs to be pushed out; when air is made to move out of the lungs we say that there is an egressive pulmonic airstream. All speech sounds are made with some movement of air, and the egressive pulmonic is by far the most commonly found air movement in the languages of the world. There are other ways of making air move in the vocal tract, but they are not usually relevant in the study of English pronunciation, so we will not discuss them here.
How is air moved into and out of the lungs? Knowing about this is important, since it will make it easier to understand many aspects of speech, particularly the nature of stress and intonation. The lungs are like sponges that can fill with air, and they are contained within the rib cage Fig. If we allow the rib cage to return to its rest position quite slowly, some of the air is expelled and can be used for producing speech sounds. If we wish to make the egressive pulmonic airstream continue without breathing in again - for example, when saying a long sentence and not wanting to be interrupted - we can make the rib cage press down on the lungs so that more air is expelled.
In talking about making air flow into and out of the lungs, the process has been described as though the air were free to pass with no obstruction. But, as we saw in Chapter 7, to make speech sounds we must obstruct the airflow in some way - breathing by itself makes very little sound. We obstruct the airflow by making one or more obstructions or strictures in the vocal tract, and one place where we can make a stricture is in the larynx, by bringing the vocal folds close to each other as described in the previous section.
Remember that there will be no vocal fold vibration unless the vocal folds are in the correct position and the air below the vocal folds is under enough pressure to be forced through the glottis. If the vocal folds vibrate we will hear the sound that we call voicing or phonation. There are many different sorts of voicing that we can produce - think of the differences in the quality of your voice between singing, shouting and speaking quietly, or think of the different voices you might use reading a story to young children in which you have to read out what is said by characters such as giants, fairies, mice or ducks; many of the differences are made with the larynx.
We can make changes in the vocal folds themselves - they can, for example, be made longer or shorter, more tense or more relaxed or be more or less strongly pressed together.
The pressure of the air below the vocal folds the subglottal pressure can also be varied. Three main differences are found: We produce voicing with high intensity for shouting, for example, and with low intensity for speaking quietly. If the vocal folds vibrate rapidly, the voicing is at high frequency; if there are fewer vibrations per second, the frequency is lower. We can produce different-sounding voice qualities, such as those we might call harsh, breathy, murmured or creaky.
The stricture is, then, total. This noise is called plosion. To give a complete description of a plosive consonant we must describe what happens at each of the following four phases in its production: We call this the closing phase. We call this the compression phase. This is the release phase.
The glottal plosive? The plosives have different places of articulation. The plosives p, b are bilabial since the lips are pressed together Fig. Normally the tongue does not touch the front teeth as it does in the dental plosives found in many languages.
The plosives k, g are velar; the back of the tongue is pressed against the area where the hard palate ends and the soft palate begins Fig. The plosives p, t, k are always voiceless; b, d, g are sometimes fully voiced, sometimes partly voiced and sometimes voiceless.
We will consider what b, d, g should be called in Section 7. All six plosives can occur at the beginning of a word initial position , between other sounds medial position and at the end of a word final position.
To begin with we will look at plosives preceding vowels which can be abbreviated as CV, where C stands for a consonant and V stands for a vowel , between vowels VCV and following vowels VC. We will look at more complex environments in later chapters. The closing phase for p, t, k and b, d, g takes place silently. During the compression phase there is no voicing in p, t, k; in b, d, g there is normally very little voicing - it begins only just before the release.
If the speaker pronounces an initial b, d, g very slowly and carefully there may be voicing during the entire compression phase the plosive is then fully voiced , while in rapid speech there may be no voicing at all.
The release of p, t, k is followed by audible plosion - that is, a burst of noise. There is then, in the post-release phase, a period during which air escapes through the vocal folds, making a sound like h. This is called aspiration. Then the vocal folds come together and voicing begins. The release of b, d, g is followed by weak plosion, and this happens at about the same time as, or shortly after, the beginning of voicing.
The most noticeable and important difference, then, between initial p, t, k and b, d, g is the aspiration of the voiceless plosives p, t, k. If English speakers hear a fully voiced initial plosive, they will hear it as one of b, d, g but will notice that it does not sound quite natural.
If they hear a voiceless unaspi- rated plosive they will also hear that as one of b, d, g, because it is aspiration, not voicing which distinguishes initial p, t, k from b, d, g.
Only when they hear a voiceless aspirated plosive will they hear it as one of p, t, k; experiments have shown that we perceive aspiration when there is a delay between the sound of plosion and the beginning or onset of voicing. In initial position, b, d, g cannot be preceded by any consonant, but p, t, k may be preceded by s. When one of p, t, k is preceded by s it is unaspirated. From what was said above it should be clear that the unaspirated p, t, k of the initial combinations sp, st, sk have the sound quality that makes English speakers perceive a plosive as one of b, d, g; if a recording of a word beginning with one of sp, st, sk is heard with the s removed, an initial b, d or g is perceived by English speakers.
The pronunciation of p, t, k and b, d, g in medial position depends to some extent on whether the syllables preceding and following the plosive are stressed. In general we can say that a medial plosive may have the characteristics either of final or of initial plosives. Final b, d, g normally have little voicing; if there is voicing, it is at the beginning of the compression phase; p, t, k are always voiceless. The plosion following the release of p, t, k and b, d, g is very weak and often not audible.
The difference between p, t, k and b, d, g is primarily the fact that vowels preceding p, t, k are much shorter. The shortening effect of p, t, k is most noticeable when the vowel is one of the long vowels or diphthongs. This effect is sometimes known as pre-fortis clipping. The description of them makes it clear that it is not very accurate to call them "voiced"; in initial and final position they are scarcely voiced at all, and any voicing they may have seems to have no perceptual importance.
Some phoneticians say that p, t, k are produced with more force than b, d, g, and that it would therefore be better to give the two sets of plosives and some other consonants names that indicate that fact; so the voiceless plosives p, t, k are sometimes called fortis meaning 'strong' and b, d, g are then called lenis meaning 'weak'. It may well be true that p, t, k are produced with more force, though nobody has really proved it - force of articulation is very difficult to define and measure.
On the other hand, the terms fortis and lenis are difficult to remember. Despite this, we shall follow the practice of many books and use these terms. The plosive phonemes of English can be presented in the form of a table as shown here: Each major type of consonant such as plosives like p, t, k, fricatives like s, z, and nasals like m, n obstructs the airflow in a different way, and these are classed as different manners of articulation.
Notes on problems and further reading 7. Chapters 6 and 2 ; Ashby and Maidment In classifying consonants it is possible to go to a very high level of complexity if one wishes to account for all the possibilities; see, for example, Pike The vowel length difference before final voiceless consonants is apparently found in many possibly all languages, but in English this difference - which is very slight in most languages - has become exaggerated so that it has become the most important factor in distinguishing between final p, t, k and b, d, g; see Chen Some phonetics books wrongly state that b, d, g lengthen preceding vowels, rather than that p, t, k shorten them.
It is necessary to consider how one could measure "force of articulation"; many different laboratory techniques have been tried to see if the articulators are moved more energetically for fortis consonants, but all have proved inconclusive. The only difference that seems reasonably reliable is that fortis consonants have higher air pressure in the vocal tract, but Lisker has argued convincingly that this is not conclusive evidence for a "force of articulation" difference.
It is possible to ask phonetically untrained speakers whether they feel that more energy is used in pronouncing p, t, k than in b, d, g, but there are many difficulties in doing this. A useful review of the "force of articulation" question is in Catford Your description should start and finish with the position for normal breathing.
Here is a description of the pronunciation of the word 'bee' bi: Starting from the position for normal breathing, the lips are closed and the lungs are compressed to create air pressure in the vocal tract.
The tongue moves to the position for a close front vowel, with the front of the tongue raised close to the hard palate. The vocal folds are brought close together and voicing begins; the lips then open, releasing the compressed air. Voicing continues for the duration of an i: Then the lung pressure is lowered, voicing ceases and the articulators return to the normal breathing position.
Words to describe: It is now necessary to consider some fundamental theoretical questions. What do we mean when we use the word "sound"? How do we establish what are the sounds of English, and how do we decide how many there are of them? When we speak, we produce a continuous stream of sounds. In studying speech we divide this stream into small pieces that we call segments. The word 'man' is pronounced with a first segment m, a second segment a; and a third segment n.
It is not always easy to decide on the number of segments. To give a simple example, in the word 'mine' the first segment is m and the last is n, as in the word 'man' discussed above.
But should we regard the aI in the middle as one segment or two? We will return to this question. As well as the question of how we divide speech up into segments, there is the question of how many different sounds or segment types there are in English.
Chapters 7 and 7 introduced the set of vowels found in English. Each of these can be pronounced in many slightly different ways, so that the total range of sounds actually produced by speakers is practically infinite. Yet we feel quite confident in saying that the number of English vowels is not greater than twenty.
Why is this? The answer is that if we put one of those twenty in the place of one of the others, we can change the meaning of a word. For example, if we substitute as for e in the word 'bed' we get a different word: But in the case of two slightly different ways of pronouncing what we regard as "the same sound", we usually find that, if we substitute one for the other, a change in the meaning of a word does not result. If we substitute a more open vowel, for example cardinal vowel no. The principles involved here may be easier to understand if we look at a similar situation related to the letters of the alphabet that we use in writing English.
The letter of the alphabet in writing is a unit which corresponds fairly well to the unit of speech we have been talking about earlier in this chapter - the segment. In the alphabet we have five letters that are called vowels: If we choose the right context we can show how substituting one letter for another will change meaning.
Thus with a letter 'p' before and a letter 't' after the vowel letter, we get the five words spelt 'pat', 'pet', 'pit', 'pot', 'put', each of which has a different meaning. We can do the same with sounds. If someone who knew nothing about the alphabet saw these four characters: They would quickly discover, through noticing differences in meaning, that 'u' is a different letter from the first three. What would our illiterate observer discover about these three?
They would eventually come to the conclusion about the written characters 'a' and 'a' that the former occurs most often in printed and typed writing while the latter is more common in handwriting, but that if you substitute one for the other it will not cause a difference in meaning.
If our observer then examined a lot of typed and printed material they would eventually conclude that a word that began with 'a' when it occurred in the middle of a sentence would begin with 'A', and never with 'a', at the beginning of a sentence.
They would also find that names could begin with 'A' but never with 'a'; they would conclude that 'A' and 'a' were different ways of writing the same letter and that a context in which one of them could occur was always a context in which the other could not.
As will be explained below, we find similar situations in speech sounds. If you have not thought about such things before, you may find some difficulty in understanding the ideas that you have just read about. The principal difficulty lies in the fact that what is being talked about in our example of letters is at the same time something abstract the alphabet, which you cannot see or touch and something real and concrete marks on paper. The alphabet is something that its users know; they also know that it has twenty-six letters.
But when the alphabet is used to write with, these letters appear on the page in a practically infinite number of different shapes and sizes. Now we will leave the discussion of letters and the alphabet; these have only been introduced in this chapter in order to help explain some important general principles. Let us go back to the sounds of speech and see how these principles can be explained.
As was said earlier in this chapter, we can divide speech up into segments, and we can find great variety in the way these segments are made. But just as there is an abstract alphabet as the basis of our writing, so there is an abstract set of units as the basis of our speech. These units are called phonemes, and the complete set of these units is called the phonemic system of the language. The phonemes themselves are abstract, but there are many slightly different ways in which we make the sounds that represent these phonemes, just as there are many ways in which we may make a mark on a piece of paper to represent a particular abstract letter of the alphabet.
For example, the b at the beginning of a word such as 'bad' will usually be pronounced with practically no voicing. Sometimes, though, a speaker may produce the b with full voicing, perhaps in speaking very emphatically. If this is done, the sound is still identified as the phoneme b, even though we can hear that it is different in some way.
We have in this example two different ways of making b - two different realisations of the phoneme. One can be substituted for the other without changing the meaning. We also find cases in speech similar to the writing example of capital 'A' and little 'a' one can only occur where the other cannot. For example, we find that the realisation of t in the word 'tea' is aspirated as are all voiceless plosives when they occur before stressed vowels at the beginning of syllables.
In the word 'eat', the realisation of t is unaspirated as are all voiceless plosives when they occur at the end of a syllable and are not followed by a vowel.
The aspirated and unaspirated realisations are both recognised as t by English speakers despite their differences. But the aspirated realisation will never be found in the place where the unaspirated realisation is appropriate, and vice versa. When we find this strict separation of places where particular realisations can occur, we say that the realisations are in complementary distribution. One more technical term needs to be introduced: In the last example, we were studying the aspirated and unaspirated allophones of the phoneme t.
Usually we do not indicate different allophones when we write symbols to represent sounds. Basically the symbols are for one of two purposes: We will look first at phonemic symbols. The most important point to remember is the rather obvious-seeming fact that the number of phonemic symbols must be exactly the same as the number of phonemes we decide exist in the language.
It is rather like typing on a keyboard - there is a fixed number of keys that you can press. One of the traditional exercises in pronunciation teaching by phonetic methods is that of phonemic transcription, where every speech sound must be identified as one of the phonemes and written with the appropriate symbol.
There are two different kinds of transcription exercise: In a phonemic transcription, then, only the phonemic symbols may be used; this has the advantage that it is comparatively quick and easy to learn to use it.
The disadvantage is that as you continue to learn more about phonetics you become able to hear a lot of sound differences that you were not aware of before, and students at this stage find it frustrating not to be able to write down more detailed information.
The phonemic system described here for the BBC accent contains forty-four phonemes. We can display the complete set of these phonemes by the usual classificatory methods used by most phoneticians; the vowels and diphthongs can be located in the vowel quadrilateral - as was done in Chapters 7 and 7 - and the consonants can be placed in a chart or table according to place of articulation, manner of articulation and voicing.
Human beings can make many more sounds than these, and phoneticians use a much larger set of symbols when they are trying to represent sounds more accurately.
The best- known set of symbols is that of the International Phonetic Association's alphabet the letters IPA are used to refer to the Association and also to its alphabet. The vowel symbols of the cardinal vowel system plus a few others are usually included on the chart of this alphabet, which is reproduced at the beginning of the book p. It is important to note that in addition to the many symbols on the chart there are a lot of diacritics - marks which modify the symbol in some way; for example, the symbol for cardinal vowel no.
It would not be possible in this course to teach you to use all these symbols and diacritics, but someone who did know them all could write a transcription that was much more accurate in phonetic detail, and contained much more information than a phonemic transcription. Such a transcription would be called a phonetic transcription; a phonetic transcription containing a lot of information about the exact quality of the sounds would be called a narrow phonetic transcription, while one which only included a little more information than a phonemic transcription would be called a broad phonetic transcription.
One further type of transcription is one which is basically phonemic, but contains additional symbolic information about allophones of particular symbols: As an example of the use of allophonic transcription, in this course phonetic symbols are used occasionally when it is necessary to give an accurate label to an allophone of some English phoneme, but we do not do any phonetic transcription of continuous speech: A widely-used convention is to enclose symbols within brackets that show whether they are phonemic or phonetic: While this convention is useful when giving a few examples, there is so much transcription in this book that I feel it would be an unnecessary distraction to enclose each example in brackets.
It should now be clear that there is a fundamental difference between phonemic symbols and phonetic symbols. Since the phonemic symbols do not have to indicate precise phonetic quality, it is possible to choose among several possible symbols to represent a particular phoneme; this has had the unfortunate result that different books on English pronunciation have used different symbols, causing quite a lot of confusion to students.
In this course we are using the symbols now most frequently used in British publishing. It would be too long a task to examine other writers' symbols in detail, but it is worth considering some of the reasons for the differences. Some writers have concentrated on producing a set of phonemic symbols that need the minimum number of special or non-standard symbols.
Others have thought it important that the symbols should be as close as possible to the symbols that a phonetician would choose to give a precise indication of sound quality. To use the same example again, referring to the vowel in 'cat', it could be argued that if the vowel is noticeably closer than cardinal vowel no. There can be disagreements about the most important characteristics of a sound that a symbol should indicate: This is the approach taken in this course.
When we talk about how phonemes function in language, and the relationships among the different phonemes - when, in other words, we study the abstract side of the sounds of language, we are studying a related but different subject that we call phonology. Only by studying both the phonetics and the phonology of English is it possible to acquire a full understanding of the use of sounds in English speech. Let us look briefly at some areas that come within the subject of phonology; these areas of study will be covered in more detail later in the course.
In chess, for example, the exact shape and colour of the pieces are not important to the game as long as they can be reliably distinguished. But the number of pieces, the moves they can make and their relationship to all the other pieces are very important; we would say that if any of these were to be changed, the game would no longer be what we call chess.
Similarly, playing cards can be printed in many different styles and sizes, but while changing these things does not affect the game played with them, if we were to remove one card from the pack or add one card to it before the start of a game, nobody would accept that we were playing the game correctly.
In a similar way, we have a more or less fixed set of "pieces" phonemes with which to play the game of speaking English. There may be many slightly different realisations of the various phonemes, but the most important thing for communication is that we should be able to make use of the full set of phonemes. Phoneme sequences and syllable structure In every language we find that there are restrictions on the sequences of phonemes that are used.
In phonology we try to analyse what the restrictions and regularities are in a particular language, and it is usually found helpful to do this by studying the syllables of the language. Suprasegmental phonology Many significant sound contrasts are not the result of differences between phonemes. For example, stress is important: Intonation is also important: These examples show sound contrasts that extend over several segments phonemes , and such contrasts are called suprasegmental.
We will look at a number of other aspects of suprasegmental phonology later in the course. Notes on problems and further reading This chapter is theoretical rather than practical. There is no shortage of material to read on the subject of the phoneme, but much of it is rather difficult and assumes a lot of background knowledge. For basic reading I would suggest Katamba Chapter 7 , Cruttenden Chapter 8, Section 7 or Giegerich There are many classic works: Jones ; first published is widely regarded as such, although it is often criticised nowadays for being superficial or even naive.
The subject of symbols is a large one: Chapter 2. The IPA has tried as far as possible to keep to Roman-style symbols, although it is inevitable that these symbols have to be supplemented with diacritics extra marks that add detail to symbols - to mark the vowel [e] as long, we can add the length diacritic: There is a lot of information about symbol design and choice in Pullum and Ladusaw Some phoneticians working at the end of the nineteenth century tried to develop non-alphabetic sets of symbols whose shape would indicate all essential phonetic characteristics; these are described in Abercrombie We have seen that one must choose between, on the one hand, symbols that are very informative but slow to write and, on the other, symbols that are not very precise but are quick and convenient to use.
Pike presents at the end of his book an "analphabetic notation" designed to permit the coding of sounds with great precision on the basis of their articulation; an indication of the complexity of the system is the fact that the full specification of the vowel [o] requires eighty-eight characters.
On the opposite side, many American writers have avoided various IPA symbols as being too complex, and have tried to use as far as possible symbols and diacritics which are already in existence for various special alphabetic requirements of European languages and which are available on standard keyboards. The widespread use of computer printers and word processing has revolutionised the use of symbols, and sets of phonetic fonts are widely available via the Internet.
We are still some way, however, from having a universally agreed set of IPA symbol codes, and for much computer-based phonetic research it is necessary to make do with conventions which use existing keyboard characters. Note for teachers It should be made clear to students that the treatment of the phoneme in this chapter is only an introduction.
It is difficult to go into detailed examples since not many symbols have been introduced at this stage, so further consideration of phonological issues is left until later chapters. Written exercises The words in the following list should be transcribed first phonemically, then in square brackets phonetically. In your phonetic transcription you should use the following diacritics: Use the same mark for diphthongs, placing the diacritic on the first part of the diphthong. Example spelling: Most languages have fricatives, the most commonly- found being something like s.
Fricatives are continuant consonants, which means that you can continue making them without interruption as long as you have enough air in your lungs. Plosives, which were described in Chapter 7, are not continuants. You can demonstrate the importance of the narrow passage for the air in the following ways: The hissing sound will stop as the air passage gets larger. Notice how the hissing sound of the air escaping between teeth and lip suddenly stops.
Affricates are rather complex consonants. They begin as plosives and end as fricatives. A familiar example is the affricate heard at the beginning and end of the word church'. So the plosive is followed immediately by fricative noise. However, the definition of an affricate must be more restricted than what has been given so far. We would not class all sequences of plosive plus fricative as affricates; for example, we find in the middle of the word 'breakfast' the plosive k followed by the fricative f.
It is usually said that the plosive and the following fricative must be made with the same articulators - the plosive and fricative must be homorganic. We could also consider tr, dr as affricates for the same reason. They can be seen in the table below: This is similar to what was seen with the plosives.
The fortis fricatives are said to be articulated with greater force than the lenis, and their friction noise is louder. The lenis fricatives have very little or no voicing in initial and final positions, but may be voiced when they occur between voiced sounds.
The fortis fricatives have the effect of shortening a preceding vowel in the same way as fortis plosives do see Chapter 7, Section 7.
Thus in a pair of words like 'ice' aIs and 'eyes' aIz, the aI diphthong in the first word is considerably shorter than aI in the second. Since there is only one fricative with glottal place of articulation, it would be rather misleading to call it fortis or lenis which is why there is a line on the chart above dividing h from the other fricatives.
The fricative noise is never very strong and is scarcely audible in the case of v. T, D example words: The air escapes through the gaps between the tongue and the teeth.
As with f, v, the fricative noise is weak. The air escapes through a narrow passage along the centre of the tongue, and the sound produced is comparatively intense. The tongue position is shown in Fig. This means that the narrowing that produces the friction noise is between the vocal folds, as described in Chapter 7.
If you breathe out silently, then produce h, you are moving your vocal folds from wide apart to close together. However, this is not producing speech.
When we produce h in speaking English, many different things happen in different contexts. In the word 'hat', the h is followed by an as vowel. The same is found for all vowels following h; the consonant always has the quality of the vowel it precedes, so that in theory if you could listen to a recording of h-sounds cut off from the beginnings of different vowels in words like 'hit', 'hat', 'hot', 'hut', etc.
One way of stating the above facts is to say that phonetically h is a voiceless vowel with the quality of the voiced vowel that follows it. Phonologically, h is a consonant. It is usually found before vowels. As well as being found in initial position it is found medially in words such as 'ahead' shed, 'greenhouse' gri: It is noticeable that when h occurs between voiced sounds as in the words 'ahead', 'greenhouse' , it is pronounced with voicing - not the normal voicing of vowels but a weak, slightly fricative sound called breathy voice.
It is not necessary for foreign learners to attempt to copy this voicing, although it is important to pronounce h where it should occur in BBC pronunciation. Many English speakers are surprisingly sensitive about this consonant; they tend to judge as sub-standard a pronunciation in which h is missing.
In reality, however, practically all English speakers, however carefully they speak, omit the h in non-initial unstressed pronunciations of the words 'her', 'he', 'him', 'his' and the auxiliary 'have', 'has', 'had', although few are aware that they do this.
There are two rather uncommon sounds that need to be introduced; since they are said to have some association with h, they will be mentioned here.
The first is the sound produced by some speakers in words which begin orthographically i. The phonetic symbol for this voiceless fricative is AY. We can find pairs of words showing the difference between this sound and the voiced sound w: It is therefore rather surprising to find that practically all writers on the subject of the phonemes of English decide that this answer is not correct, and that the sound AY in 'which', 'why', etc. We do not need to worry much about this problem in describing the BBC accent.
However, it should be noted that in the analysis of the many accents of English that do have a "voiceless w" there is not much more theoretical justification for treating the sound as h plus w than there is for treating p as h plus b. Whether the question of this sound is approached phonetically or phonologically, there is no h sound in the "voiceless w". A very similar case is the sound found at the beginning of words such as 'huge', 'human', 'hue'.
However, it is usual to treat this sound as h plus j the latter is another consonant that is introduced in Chapter 2 - it is the sound at the beginning of 'yes', 'yet'. Again we can see that a phonemic analysis does not necessarily have to be exactly in line with phonetic facts. We will follow the usual practice of transcribing the sound at the beginning of 'huge', etc.
Since the remaining consonants to be described are not paired in this way, a few points that still have to be made about fortis consonants are included in this chapter. The first point concerns the shortening of a preceding vowel by a syllable-final fortis consonant. As was said in Chapter 7, the effect is most noticeable in the case of long vowels and diphthongs, although it does also affect short vowels.
What happens if something other than a vowel precedes a fortis consonant? The effect on those continuant consonants is the same as on a vowel: Fortis consonants are usually articulated with open glottis - that is, with the vocal folds separated. This is always the case with fricatives, where airflow is essential for successful production. However, with plosives an alternative possibility is to produce the consonant with completely closed glottis. This type of plosive articulation, known as glottalisation, is found widely in contemporary English pronunciation, though only in specific contexts.
It normally happens when the plosive is followed by another consonant or a pause; for example: This consonant often shows so little friction noise that on purely phonetic grounds it seems incorrect to class it as a fricative. It is more like a weak lenis dental plosive. This matter is discussed again in Chapter 97, Section On the phonological side, I have brought in a discussion of the phonemic analysis of two "marginal" fricatives AY, which present a problem though not a particularly important or fundamental one: I feel that this is worth discussing in that it gives a good idea of the sort of problem that can arise in analysing the phonemic system of a language.
The other problem area is the glottalisation described at the end of the chapter. There is now a growing awareness of how frequently this is to be found in contemporary English speech; however, it not easy to formulate rules stating the contexts in which this occurs. There is discussion in Brown Section 1.
All of these seven consonants are continuants and usually have no friction noise, but in other ways they are very different from each other. For this to happen, the soft palate must be lowered; in the case of all the other consonants and vowels of English, the soft palate is raised and air cannot pass through the nose. In nasal consonants, however, air does not pass through the mouth; it is prevented by a complete closure in the mouth at some point.
If you produce a long sequence dndndndndn without moving your tongue from the position for alveolar closure, you will feel your soft palate moving up and down. The three types of closure are: The consonants m, n are simple and straightforward with distributions quite similar to those of the plosives. There is in fact little to describe. However, N is a different matter.
It is a sound that gives considerable problems to foreign learners, and one that is so unusual in its phonological aspect that some people argue that it is not one of the phonemes of English at all. The place of articulation of N is the same as that of k, g; it is a useful exercise to practise making a continuous r sound.
If you do this, it is very important not to produce a k or g at the end - pronounce the N like m or n. For example, in BBC pronunciation we find the following: What is the difference between A and B? The important difference is in the way the words are constructed - their morphology. The words of column B can be divided into two grammatical pieces: These pieces are called morphemes, and we say that column B words are morphologically different from column A words, since these cannot be divided into two morphemes.
Let us now look at the ends of words ending orthographically with 'ng'. We do not need a separate explanation for this: If this point seems difficult, think of the comparable case of sentences and words: Unfortunately, rules often have exceptions. The main exception to the above morpheme-based rule concerns the comparative and superlative suffixes '-er' and '-est'.
It is important to remember that English speakers in general apart from those trained in phonetics are quite ignorant of this rule, and yet if a foreigner uses the wrong pronunciation i. The velar nasal consonant N is, in summary, phonetically simple it is no more difficult to produce than m or n but phonologically complex it is, as we have seen, not easy to describe the contexts in which it occurs. This is a consonant in which the passage of air through the mouth does not go in the usual way along the centre of the tongue; instead, there is complete closure between the centre of the tongue and the part of the roof of the mouth where contact is to be made the alveolar ridge in the case of l.
Because of this complete closure along the centre, the only way for the air to escape is along the sides of the tongue.
The lateral approximant is therefore somewhat different from other approximants, in which there is usually much less contact between the articulators. If you make a long l sound you may be able to feel that the sides of your tongue are pulled in and down while the centre is raised, but it is not easy to become consciously aware of this; what is more revealing if you can do it is to produce a long sequence of alternations between d and l without any intervening vowel.
If you produce dldldldldl without moving the middle of the tongue, you will be able to feel the movement of the sides of the tongue that is necessary for the production of a lateral. It is also possible to see this movement in a mirror if you open your lips wide as you produce it. Finally, it is also helpful to see if you can feel the movement of air past the sides of the tongue; this is not really possible in a voiced sound the obstruction caused by the vibrating vocal folds reduces the airflow , but if you try to make a very loud whispered l, you should be able to feel the air rushing along the sides of your tongue.
We find l initially, medially and finally, and its distribution is therefore not particularly limited. In BBC pronunciation, the consonant has one unusual characteristic: The sound in 'eel' is what we call a "dark l"; it has a quality rather similar to an [u] vowel, with the back of the tongue raised. We can therefore predict which realisation of l clear or dark will occur in a particular context: We can say, using terminology introduced in Chapter 8, that clear l and dark l are allophones of the phoneme l in complementary distribution.
Most English speakers do not consciously know about the difference between clear and dark l, yet they are quick to detect the difference when they hear English speakers with different accents, or when they hear foreign learners who have not learned the correct pronunciation.
You might be able to observe that most American and lowland Scottish speakers use a "dark l" in all positions, and don't have a "clear l" in their pronunciation, while most Welsh and Irish speakers have "clear l" in all positions.
Another allophone of l is found when it follows p, k at the beginning of a stressed syllable. The I is then devoiced i. The situation is as explained in Chapter 7 similar to the aspiration found when a vowel follows p, t, k in a stressed syllable: As far as the articulation of the sound is concerned, there is really only one pronunciation that can be recommended to the foreign learner, and that is what is called a post-alveolar approximant.
An approximant, as a type of consonant, is rather difficult to describe; informally, we can say that it is an articulation in which the articulators approach each other but do not get sufficiently close to each other to produce a "complete" consonant such as a plosive, nasal or fricative.
The difficulty with this explanation is that articulators are always in some positional relationship with each other, and any vowel articulation could also be classed as an approximant - but the term "approximant" is usually used only for consonants.
The important thing about the articulation of r is that the tip of the tongue approaches the alveolar area in approximately the way it would for a t or d, but never actually makes contact with any part of the roof of the mouth. You should be able to make a long r sound and feel that no part of the tongue is in contact with the roof of the mouth at any time.
This is, of course, very different from the "r-sounds" of many other languages where some kind of tongue-palate contact is made. The tongue is in fact usually slightly curled backwards with the tip raised; consonants with this tongue shape are usually called retroflex. If you pronounce an alternating sequence of d and r drdrdrdrdr while looking in a mirror you should be able to see more of the underside of the tongue in the r than in the d, where the tongue tip is not raised and the tongue is not curled back.
A rather different r sound is found at the beginning of a syllable if it is preceded by p, t, k; it is then voiceless and fricative. This pronunciation is found in words such as 'press', 'tress', 'cress'. One final characteristic of the articulation of r is that it is usual for the lips to be slightly rounded; learners should do this but should be careful not to exaggerate it.
If the lip-rounding is too strong the consonant will sound too much like w, which is the sound that most English children produce until they have learned to pronounce r in the adult way. The distributional peculiarity of r in the BBC accent is very easy to state: No one has any difficulty in remembering this rule, but foreign learners most of whom, quite reasonably, expect that if there is a letter 'r' in the spelling then r should be pronounced find it difficult to apply the rule to their own pronunciation.
There is no problem with words like the following: But in the following words there is no r in the pronunciation: Those accents which have r in final position before a pause and before a consonant are called rhotic accents, while accents in which r only occurs before vowels such as BBC are called non-rhotic.
They are known as approximants introduced in Section 2. The most important thing to remember about these phonemes is that they are phonetically like vowels but phonologically like consonants in earlier works on phonology they were known as "semivowels".
From the phonetic point of view the articulation of j is practically the same as that of a front close vowel such as [i], but is very short. In the same way w is closely similar to [u]. If you make the initial sound of 'yet' or 'wet' very long, you will be able to hear this.
But despite this vowel-like character, we use them like consonants. For example, they only occur before vowel phonemes; this is a typically consonantal distribution.
We can show that a word beginning with w or j is treated as beginning with a consonant in the following way: If a word beginning with w or j is preceded by the indefinite article, it is the 'a' form that is found as in 'a way', 'a year'. Another example is that of the definite article. This evidence illustrates why it is said that j, w are phonologically consonants.
However, it is important to remember that to pronounce them as fricatives as many foreign learners do , or as affricates, is a mispronunciation.
Only in special contexts do we hear friction noise in j or w; this is when they are preceded by p, t, k at the beginning of a syllable, as in these words: This means that the beginning of a vowel is voiceless in this context.
However, when p, t, k are followed not by a vowel but by one of l, r, j, w, these voiced continuant consonants undergo a similar process, as has been mentioned earlier in this chapter: Consequently, if for example 'tray' were to be pronounced without devoicing of the r i. This completes our examination of the consonant phonemes of English. It is useful to place them on a consonant chart, and this is done in Table l.
On this chart, the different places of articulation are arranged from left to right and the manners of articulation are arranged from top to bottom. When there is a pair of phonemes with the same place and manner of articulation but differing in whether they are fortis or lenis voiceless or voiced , the symbol for the fortis consonant is placed to the left of the symbol for the lenis consonant.
Notes on problems and further reading The notes for this chapter are devoted to giving further detail on a particularly difficult theoretical problem. Since the velar nasal is introduced in this chapter, I have chosen to attempt this here.
However, it is a rather complex theoretical matter, and you may prefer to leave consideration of it until after the discussion of problems of phonemic analysis in Chapter There are brief discussions of the phonemic status of N in Chomsky and Halle Everyone agrees that English has at least two contrasting nasal phonemes, m and n. Sapir said that "no native speaker of English could be made to feel in his bones" that N formed part of a series with m, n.
This is, of course, very hard to establish, although that does not mean that Sapir was wrong. We need to look at point i in more detail and go on to see how this leads to the argument against having N as a phoneme. Please note that I am not trying to argue that this proposal must be correct; my aim is just to explain the argument.
The whole question may seem of little or no practical consequence, but we ought to be interested in any phonological problem if it appears that conventional phoneme theory is not able to deal satisfactorily with it. For example: Neither m nor n can occur in this environment. The phonetic realisation of the n phoneme as a velar nasal will be accounted for by a general rule that we will call Rule 9: Rule 0: As explained in Section 2.
After establishing these "background facts", we can go on to state the argument as follows: Rule 2: However, the rule does apply to all the others, hence the final phonetic forms: The important point, however, is that if one is prepared to use the kind of complexity and abstractness illustrated above, one can produce quite far- reaching changes in the phonemic analysis of a language.
The other consonants - l, r, w, j - do not, I think, need further explanation, except to mention that the question of whether j, w are consonants or vowels is examined on distributional grounds in O'Connor and Trim Written exercises a.
List all the consonant phonemes of the BBC accent, grouped according to manner of articulation. Transcribe the following words phonemically: Describe what movements are carried out by the soft palate in the pronunciation of the following words: Most people seem to believe that, even if they cannot define what a syllable is, they can count how many syllables there are in a given word or sentence.
If they are asked to do this they often tap their finger as they count, which illustrates the syllable's importance in the rhythm of speech.
As a matter of fact, if one tries the experiment of asking English speakers to count the syllables in, say, a recorded sentence, there is often a considerable amount of disagreement.
We find a similar situation with the syllable, in that it may be defined both phonetically and phonologically. Phonetically i. We will now look at some examples: These are preceded and followed by silence. Looking at them from the phonological point of view is quite different.
What this involves is looking at the possible combinations of English phonemes; the study of the possible phoneme combinations of a language is called phonotactics. It is simplest to start by looking at what can occur in initial position - in other words, what can occur at the beginning of the first word when we begin to speak after a pause.
We find that the word can begin with a vowel, or with one, two or three consonants. No word begins with more than three consonants. In the same way, we can look at how a word ends when it is the last word spoken before a pause; it can end with a vowel, or with one, two, three or in a small number of cases four consonants.
No current word ends with more than four consonants. We now look at syllables beginning with two consonants. When we have two or more consonants together we call them a consonant cluster.
Initial two-consonant clusters are of two sorts in English. The s in these clusters is called the pre-initial consonant and the other consonant t, w, m in the above examples the initial consonant.
These clusters are shown in Table 7. We call the first consonant of these clusters the initial consonant and the second the post-initial. There are some restrictions on which consonants can occur together. This can best be shown in table form, as in Table 7. When we look at three-consonant clusters we can recognise a clear relationship between them and the two sorts of two-consonant cluster described above; examples of three- consonant initial clusters are: The s is the pre-initial consonant, the p, t, k that follow s in the three example words are the initial consonant and the l, r, w are post-initial.
In fact, the number of possible initial three-consonant clusters is quite small and they can be set out in full words given in spelling form: Two-consonant clusters of s plus l, w, j are also possible e. These clusters can be analysed either as pre-initial s plus initial l, w, j, r or initial s plus post-initial l, w, j, r. There is no clear answer to the question of which analysis is better; here they are treated in the latter way, and appear in Table 3.
Table 3 Two-consonant clusters with post-initial l, r, w, j Notes in doubtful cases: Sri Lanka. Many Welsh names including some well known outside Wales - such as girls' names like Gwen and place names like the county of Gwent - have initial gw and English speakers seem to find them perfectly easy to pronounce.
This is, however, a very infrequent cluster for English.