Modifications of Consonants in Connected Speech

Hitherto, we have looked at sounds individually. But lan­guage in everyday use is not conducted in terms of isolated, sep­arate units; it is performed in connected sequences of larger units, in words, phrases and longer utterances. Now we shall be concerned with what happens to sounds not only within words, but also when the words are connected into larger units. There are actually some remarkable differences between the pronuncia­tion of a word in isolation and of the same word in a block of connected speech. These changes are mostly quite regular and predictable. Still the problem of defining the phonemic status of sounds in connected speech is by far too complicated because of the numerous modifications of sounds in speech. These modifica­tions are observed both within words and at word boundaries. As you may know from the practical course of phonetics, speech sounds influence each other in the flow of speech. As a result of the intercourse between consonants and vowels and within each class there appear such processes of connected speech as assimi­lation, accommodation, vowel reduction and elision which is sometimes termed deletion.

The adaptive modification of a consonant by a neighbouring consonant m the speech chain is known as assimilation, e. g. the alveolar [t] followed by the interdental [θ] becomes dental: eighth, at three.

The term accommodation is often used by linguists to denote the interchanges of "vowel + consonant type" or "consonant + vowel type", for instance, some slight degree of nasalization of vowels preceded or followed by nasal sonorants: never, men; or labialization of consonants preceding the vowels [o] and [y] in Russian: больно , конь , думать , лучше .

One of the wide-spread sound changes is certainly vowel re­duction. Reduction is actually qualitative or quantitative weak­ening of vowels in unstressed positions, e.g. board — blackboard, man — postman.

71

Elision or complete loss of sounds, both vowels and conso­nants, is often observed in English. Elision is likely to be minimal in slow careful speech and maximal in rapid relaxed colloquial forms of speech.

The processes involved cannot be neglected in defining the phonemic status of speech sounds. These phenomena manifest the economy of pronouncing efforts on the part of the speaker. The speaker and the listener are two participants of communica­tion. The addresser's aim is to inform the addressee of some­thing. The letter's wish is to comprehend the idea. The simplifications themselves go quite unnoticed by the listener, as they do not affect the meaning. The listener is mainly interested in the meaning the speaker aims to convey and not in the precise pho­netic organization of a block of connected speech. So long as the meaning is recognizable, the listener is satisfied. We have to re­gard the omissions and reductions then as a kind of economy on the part of the speaker, who aims not to give more information than necessary. The speaker assumes usually correctly that the listener will not notice the omissions. A question then arises whether such kind of simplification leads to excessive ambiguity. On the face of it, one would expect a great deal of ambiguity to arise, particularly in cases where a whole word is reduced (weakened) to a single sound, for example, [z] for the whole word has. The load carried by a single sound can become enor­mous, [z], for example, can represent the reduced forms of has, is and even does, the plural and possessive for nouns, the third per­son singular for verbs. The sound [ə] can represent the reduced forms of are, or, her and sometimes of (as in six o'clock), as well as the indefinite article a, the comparative degree of adjectives {shorter), the suffix of a noun (teach-er), etc. But in spite of the meaning load carried by the same sounds, ambiguity rarely arises because the syntactic functions are quite different and the context makes the intention clear. On hearing a sequence like [z'nik 'kʌmiŋ] the listener unmistakably reconstructs: Is Nick coming?; in [ð'ə 'boiz 'skeit] the sound [z] must be the plural form, while in [ð'ə 'boiz 'dʌn it] it must be derived from has.

Now let us see which qualitative features of consonant sounds may be changed in the process of their interrelation in a speech chain.

Consonants are modified according to the place of articula­tion. Assimilation takes place when a sound changes its charac­

72

ter in order to become more like a neighbouring sound. The characteristic which can vary in this way is nearly always the place of articulation, and the sounds concerned are commonly those which involve a complete closure at some point in the mouth that is plosives and nasals which may be illustrated as follows:

1. The dental [t], [d], followed by the interdental [θ], [ð] sounds (partial regressive assimilation when the influence goes backwards from a "later" sound to an "earlier" one), e.g. "eighth", "at the", 'breadth", "said that".

2. The post-alveolar [t], [d] under the influence of the post-alveolar [r] (partial regressive assimilation), e.g. "tree", "true", "that right word", "dry", "dream", "the third room".

3. The post-alveolar [s], [z] before [f] (complete regressive as­similation), e.g. horse-shoe ['hɔ:ʃʃu:], this shop [ðiʃ 'ʃɔp], does she ['dʌʃʃi:].

4. The affricative [t + j], [d + j] combinations (incomplete re­gressive assimilation), e.g. graduate ['græʤueit], congratulate [kən'græʧuleit], did you ['diʤu:], could you ['kuʤu:], what do you say ["wɔtʒu:'sei].

It is easy to see from the examples above that the sounds commonly changing their place of articulation are alveolar stops. Nasal consonants are not less susceptible to assimilation. The place of articulation of nasals also varies according to the conso­nant that follows, e.g.

In camp [m] remains bilabial before another bilabial as well as in man before a vowel.

Similarly in cent [n] is alveolar before another alveolar as well as in net.

But in "symphony" [m] is actually labio-dental followed by the labio-dental [f].

In "seventh" [n] becomes dental, before the interdental [θ].

In "pinch" [n] is palato-alveolar corresponding to the follow­ing affricate [ʧ].

In "thank" [n] assimilates to the velar consonant becoming velar [ŋ].

We should like to note here that by analogy with alveolar consonants nasal assimilation operates not only within the mor­pheme as in "thank" but also across syllable boundaries as in "symphony", across morpheme boundaries, for example, in pre­fixes in-, un- as in "incomplete", "ungrateful"; "impractical",

73

where [n] assimilates to [p] and becomes bilabial [m], in the stressed prefix con- as in "conquer". Assimilation of nasals seems to be also optional across word boundaries, e.g. in case, in fact.

The manner of articulation is also changed as a result of as­similation, which may be illustrated as follows:

1. Loss of plosion. In the sequence of two plosive consonants the former loses its plosion: glad to see you, great trouble, and old clock (partial regressive assimilations).

2. Nasal plosion. In the sequence of a plosive followed by a nasal sonorant the manner of articulation of the plosive sound and the work of the soft palate are involved, which results in the nasal character of plosion release: "sudden", "not now", "at night", "let me see" (partial regressive assimilations).

3. Lateral plosion. In the sequence of a plosive followed by the lateral sonorant |1] the noise production of the plosive stop is changed into that of the lateral stop: "settle", "table", "at last" (partial regressive assimilations). It is obvious that in each of the occasions one characteristic feature of the phoneme is lost.

The voicing value of a consonant may also change through assimilation. This type of assimilation affects the work of the vo­cal cords and the force of articulation. In particular voiced lenis sounds become voiceless fortis when followed by another voice­less sound, e.g.:

1.  Fortis voiceless/lenis voiced type of assimilation is best manifested by the regressive assimilation in such words as news-paper [news [z] + paper); gooseberry (goose [s] + berry). In casual informal speech voicing assimilation is often met, e.g. have to do it ['hæf tə 'du:], five past two ['faif pa:st 'tu:]. The sounds which assimilate their voicing are usually, as the examples show, voiced lenis fricatives assimilated to the initial voiceless fortis consonant of the following word. Grammatical items, in particular, are most affected: [z] of has, is, does changes to [s], and [v] of of, have be-
comes [f], e.g.

She's five. Of course.

She has fine eyes. You've spoiled it.

Does Pete like it?

2.  The weak forms of the verbs is and has are also assimilated
to the final voiceless fortis consonants of the preceding word thus
the assimilation is functioning in the progressive direction, e.g.

74

Your aunt's coming.

What's your name?

(partial progressive assimilation)

3. English sonorants [m, n, r, I, j, w] preceded by the fortis voiceless consonants [p, t, k, s] are partially devoiced, e.g. "smart", "snake", "tray", "quick", "twins", "play", "pride" (partial progressive assimilation).

The voiced/voiceless type of assimilation is well developed in the Russian language, e.g. сдавать , сбросить , французский , абсолютный . The positional devoicing of final consonants is es­pecially constant, e.g. клуб , снег , мороз .

It should be noted that the interference of the Russian voi­ced/voiceless regressive type of assimilation results in a typical mistake in English: "black dog", "this day", "gets dark", "much bet­ter", "let's go". In English assimilation usually results in changing voiced lenis consonants into voiceless fortis, e.g. of course [əf'kɔ:s]. The change of voiceless fortis consonants into voiced lenis as a result of assimilation is not typical. Thus teachers of English should be aware of it and be ready with special exercises to prevent the errors.

Lip position may be affected by the accommodation, the in­terchange of consonant + vowel type. Labialisation of conso­nants is traced under the influence of the neighbouring back vowels (accommodation), e.g. pool, moon, rude, soon, who, cool, etc. It is possible to speak about the spread lip position of conso­nants followed or preceded by front vowels [i], [i], e.g. tea — beat; meet — team; feat — leaf, keep — leak; sit — miss (accom­modation).

The position of the soft palate is also involved in the accom­modation. Slight nasalization as the result of prolonged lowering of the soft palate is sometimes traced in vowels under the influ­ence of the neighbouring sonants [m] and [n], e.g. and, morning, men, come in (accommodation).

To summarize so far, assimilation affecting the place of arti­culation is considered to be most typical of the English sound system and assimilation affecting the work of the vocal cords (voiced/voiceless type) is most typical of the Russian speech.

It is to be noted that the described allophonic realizations of phonemes are marked in Received Pronunciation as obligatory and stable for all the members of the speech community in every

75

phonetic style. It is perfectly natural that all sorts of sound adapta­tion are more frequent in informal colloquial flow of speech than in formal speech. This tendency is a matter of style, not correct­ness. In informal casual discourse assimilation involves the alveo­lar stops [t], [d] before another stop at border junctions, e.g.

that place [ðæp 'pleıs]         Or: hard problem ['hæb 'prɒbləm]

that book ['ðæp 'buk]          hard blow ['ha:b 'blɜu]

that kind ['ðæk 'kaınd]        hard case ['ha:g 'keıs]

that golfer ['ðæk 'gɒlfə]      hard ground ['hag 'graund]

In these examples [t] retains its voicelessness, [d] accordingly retains its voiced character, but both of them shift their articula­tion in symphony with the articulation of the following stop. It should be noted that the velar stops [k], [g] are not subjected to the assimilation of this kind.

The alveolars [s], [z] and [t], [d] assimilate in informal casual speech more often than in slow careful speech to palato-alveolars when followed by the palatal [j], e.g.

face your friend ['feıʃ jə 'frend]

as you like [əʒ ju 'laık]

can't you do it ['ka:ntʃə 'du: it]

on duty [ɒn 'ʤu:tı]

The examples above illustrate the changes affecting the place of articulation.

In informal casual speech complete type of assimilation is often observed, e.g.

ten minutes ['tern 'mınıts]

nice shoes ['naıʃ'ʃu:z]

one more ['wʌm 'mɔ:J

good-bye ['gub 'baı]

let me ['lem mi]

We would like to point out right here that elision or com­plete loss of sounds, both vowels and consonants, is observed in the structure of English words. It is typical of rapid colloquial speech and marks the following sounds:

1. Loss of [h] in personal and possessive pronouns he, his, her, him and the forms of the auxiliary verb have, has, had is wide­spread, e.g. What has he done?  ['wɒt əz i ˎdʌn].

2. [1] tends to be lost when preceded by [ɔ:], e.g. always ['ɔ:wız], already [ɔ:'redı], all right [ɔ:'rait].

76

ɜ. Alveolar plosives are often elided in case the cluster is fol­lowed by another consonant, e.g. next day ['neks 'dei], just one ['ʤʌs 'wʌn], mashed potatoes ['mæʃ pə'teıtɜuz]. If a vowel follows, the consonant remains, e.g. first of all, passed in time. Whole sylla­bles may be elided in rapid speech: library ['laıbrı], literary ['lıtrı].

Examples of historical elision are also known. They are initial consonants in write, know, knight, the medial consonant [t] in fasten, listen, whistle, castle.

In sum, we may say that in the process of speech the degree of sound modifications may be different, varying from partial as­similation, when one sound feature is modified, like in "tenth" (alveolar [n] becomes dental) to actual loss of a sound. For example: listen ['lısn], next day ['neks 'deı], complete sound adap­tation being the intermediate state: ten minutes [tem 'mınıts], nice shoe ['naıʃ'ʃu:].

Describing the interrelation of sounds in connected speech we would like to mention one more remarkable phenomenon.

While the elision is a very common process in connected speech, we also occasionally find sounds being inserted. When a word which ends in a vowel is followed by another word begin­ning with a vowel, the so-called intrusive "r" is sometimes pro­nounced between the vowels, e.g.

Asia and Africa ['eıʃər ənd 'æfrıkə]

the idea of it [ði:aı'dıər əv it]

ma and pa ['ma;r ənd 'pa:]

The so-called linking "r" is a common example of insertion, e.g. clearer, a teacher of English.

Thus it is clear that the linking and intrusive [r] are both part of the same phonetic process of [r] insertion.

When the word-final vowel is a diphthong which glides to [ı] such as [aı], [eı] the palatal sonorant [j] tends to be inserted, e.g saying ['seıjıŋ]; trying [trajıŋ].

In case of the [u]-gliding diphthongs [ɜu], [au] the bilabial sonorant [w] is sometimes inserted, e.g. going ['gɜuwıŋ], allowing [ə'lauwıŋ].

The process of inserting the sonorants [r], [j] or [w] may seem to contradict the tendency towards the economy of articulatory efforts. The explanation for it lies in the fact that it is apparently easier from the articulatory point of view to insert those sounds than to leave them out.

77

The insertion of a consonant-like sound, namely a sonorant, interrupts the sequence of two vowels (VV) to make it a more optional syllable type: consonant + vowel (CV). Thus, insertion occurs in connected speech in order to facilitate the process of articulation for the speaker, and not as a way of providing extra information for the listener.

Now by way of conclusion we should like to say that we un­derstand the sound quality as a set of characteristics which are in constant interrelation and compensation. In case one of the fea­tures of a phoneme is lost there remains a sufficient number of characteristics of a phoneme and its status and function are not lost. Thus modifications of sounds in a speech chain are of allophonic character, that is they are realizations of allophones of phonemes.

 

Vowels

Talking about vowels requires first to specify their articulatory and acoustic characteristics. As was mentioned earlier, vowels unlike consonants are produced with no obstruction to the stream of air, so on the perception level their integral characteris­tic is naturally tone, not noise.

It would be interesting to know that a minimum vowel sys­tem of a language is likely to take the form of

The most important characteristic of the quality of these vowels is that they are acoustically stable. They are known to be entirely different from one another both articulatorily and acoustically. Consequently they may well be said to form boundaries of "phonetic field of vowels" in a modern man's life. Thus they display the highest degree of unlikeness and so maxi­mum of abilities of people as regards to vowels. We could add that the commonest vowel system adds two other vowels to this minimum triangle to give a five vowels system of the type:

78

In the matter of the English language it would be fair to men­tion that due to various reasons it has developed a vocalic sys­tem of a much larger number of phonemes.

The quality of a vowel is known to be determined by the size, volume, and shape of the mouth resonator, which are modi­fied by the movement of active speech organs, that is the tongue and the lips. Besides, the particular quality of a vowel can de­pend on a lot of other articulatory characteristics, such as the rel­ative stability of the tongue, the position of the lips, physical du­ration of the segment, the force of articulation, the degree of tenseness of speech organs. So vowel quality could be thought of as a bundle of definite articulatory characteristics which are sometimes intricately interconnected and interdependent. For example, the back position of the tongue causes the lip round­ing, the front position of the tongue makes it rise higher in the mouth cavity, the lengthening of a vowel makes the organs of speech tenser at the moment of production and so on. From what we have said it follows that isolation and distinctions of the above-mentioned articulatory features are done only for the sake of analysis with the purpose of describing the vocalic system of the English language.

The analysis of the articulatory constituents of the quality of vowels allowed phoneticians to suggest the criteria which are conceived to be of great importance in classificatory description. First to be concerned here are the following criteria termed:

a) stability of articulation;

b) tongue position;

c) lip position;

d) character of the vowel end;

e) length;

f) tenseness.

In the part that follows, each of the above-mentioned princi­ples will be considered from phonological point of view.

Stability of articulation specifies the actual position of the ar­ticulating organ in the process of the articulation of a vowel. There are two possible varieties: a) the tongue position is stable; b) it changes, that is the tongue moves from one position to an­

79

other. In the first case the articulated vowel is relatively pure, in the second case a vowel consists of two clearly perceptible ele­ments. There exists in addition a third variety, an intermediate case, when the change in the tongue position is fairly weak. So according to this principle the English vowels are subdivided into:

a) monophthongs,

b) diphthongs,

c) diphthongoids.

Though the interpretation we have just given is an obvious matter for Soviet phoneticians it does not mean that this way of seeing the situation is shared by British phoneticians. A.C. Gimson, for example, distinguishes twenty vocalic phonemes which are made of vowels and vowel glides (57). Seven of them are treated as short phonemes: [ı], [e], [æ], [ɒ], [u], [ʌ], [ə] and thir­teen as long ones: [a:], [ɔ:], [ɜ:] [i:], [u:], [eı], [so], [aı], [au], [ɒu], [ıə], [ɛə], [uə] five of which are considered relatively pure: [a:], [ɔ:] [ɜ:], [i:], [u:]; the rest are referred to long phonemes with dif­ferent glides: [eı], [aı], [ɒı] vith a glide to [ı], [ɜu], [au] with a glide to [u]; and [iə], [ɛə], [uə] with a glide to [ə]. It is easy to see that this way of presenting the system does not reveal the actual difference between long monophthongs and long diphthongoids and consequently we could say that it fails to account adequate­ly for more delicate distinctions. Here we have to admit that though it is not a decisive difference this is the case when expli­cit information about distinguishing between different degrees of instability is practically useful far teaching purpose. For the learner of English it is important to know that the vowels [i:] and [u:] are diphthongized in modern English and the tendency for diphthongization is becoming gradually stronger.

At this point we are ready to consider the question of the phonemic status of English diphthongs. Diphthongs are complex entities just like affricates described in the previous section, so essentially similar complications are known to exist with them. The question is whether they are monophonemic or biphonemic units. It is not the lack of evidence that does not enable to an­swer it. We might say that now there is much available data ob­tained with the help of the computer equipment. Though the problem has been given a lot of attention up to now it has been neither completely discounted nor satisfactorily explained. The reason that accounts for the present situation could be formu­

80

lated in the following way: it is impossible to find a simple and logic criterion which might serve as a basis for a decision.

Soviet scholars grant the English diphthongs monophonemic status on the basis of articulatory, morphonological and syllabic indivisibility as well as the criteria of duration and commutability.

As to articulatory indivisibility of the diphthongs it could be proved by the fact that neither morpheme nor syllable boundary that separate the nucleus and the glide can pass within it, for example: ['seı-ıŋ] saying, ['kraı-ıŋ] crying, [in-'ʤɔı-ıŋ] enjoying, ['slɜu-ə] slower, ['plɜu-ıŋ] ploughing, ['klıə-rə] clearer, ['ɛə-rıŋ] airing, [puə-rə] poorer. The present study of the duration of diph­thongs shows that the length of diphthongs is the same as that that characterizes the English long monophthongs in the same phonetic context, cf. [saıt — set], [kɜut — kɔ:t]. Finally the applica­tion of commutation test proves the monophonemic status of diphthongs because any diphthong could be commutated with practically any vowel. It could be exemplified in the following oppositions:

[bait — bit] bite — bit

[bait —bʌt] bite —but

[bait —[ʧ] bɔ:t] bite — bought

and so on.

Monophonemic character of English diphthongs is proved by native speakers' intuition, who perceive these sound complexes as a single segment.

The above-mentioned considerations make Soviet linguists V. A. Vassilyev (79), L. R. Zinder (16) treat English diphthongs as monophonemic entities.

The suggestion that English diphthongs are monophonemic is necessary not only for linguistic purpose; accepting that fin­ding is also of practical importance in teaching English as a foreign language, since in Russian there are no diphthongs or diphthongoids as phonemic entities. Such combinations of sounds as [йа, йо, йу], [ой, ай], [ау, уа] (яд, йод, юг, рой, край, мяукать, вуаль), and others are biphonemic clusters, consisting either of a vowel and the Russian sonorant [й] or two vowels. Both elements in the clusters are equally energetic and distinct. So special attention should be given to the pronunciation of En­glish diphthongs which consist of two elements, the first of which, the nucleus, being strong and distinct and the second, the glide, being very weak and indistinct.

81

Another principle we should consider from phonological point of view is the position of the tongue. For the sake of con­venience the position of the tongue in the mouth cavity is cha­racterized from two aspects, that is the horizontal and vertical movement.

According to the horizontal movement Soviet phoneticians distinguish five classes of English vowels. They are:

1) front: [i:], [e], [eı], [æ], [ɛ(ə)];

2) front-retracted: [ı], [ı(ə)];

3) central: [ʌ] [ɜ:] [ə], [ɜ(u)], [ɛ(u)];

4) back [ɒ], [ɔ:], [u], [a:];

5) back-advanced: [u], [u(ə)].

A slightly different approach seems to have been taken by British phoneticians. They do not single out the classes of front-retracted and back-advanced vowels. So both [i:] and [ı] vowels are classed as front, and both [u:] and [u] vowels are classed as back. The latter point of view does not seem to be consistent enough. The point is that the vowels in these two pairs differ in quality which is partially due to the raised part of the tongue. So in this case a more detailed classification seems to be a more pre­cise one since it adequately reflects the articulatory distinction actually present in the language.

The other articulatory characteristic of vowels as to the tongue position is its vertical movement. The way British and Soviet phoneticians approach this aspect is also slightly different. British scholars distinguish three classes of vowels: high (or close), mid (or half-open), and low (or open) vowels. Soviet pho­neticians made the classification more detailed distinguishing two subclasses in each class, i.e. broad and narrow variations of the three vertical positions of the tongue. Thus the following six groups of vowels are distinguished:

1) close a) narrow: [i:] [u:];

b) broad: [ı], [u], [ı(ə)], [u(ə)];

2) mid a) narrow: [e], [a:], [ə], [e(ı)]_t [ɜ(u)];

b) broad: [ə], [ʌ];

3) open a) narrow: [ɛ(ə)], [ɔ:], [ɔ(ı)];

b) broad: [æ], [a(ı, u)], [ɒ], [a:]

The phonological relevance of the criterion under discussion can be easily discovered in the folloving oppositions:

82

[pen — pæn] pen — pan     [kæp — ka:p] cap — carp

[pen — pin] pen — pin        [kæp — cʌp] cap — cup

[bin — bi:n] bin — been      [bʌn — ba:n] bun — barn

Another feature of English vowels which is sometimes in­cluded into the principles of classification is lip rounding. Tradi­tionally three lip positions are distinguished, that is spread, neu­tral and rounded. For the purpose of classification it is sufficient to distinguish between two lip positions: rounded and unround­ed, or neutral. In English lip rounding is not relevant phonologically since no two words can be differentiated on its basis. Lip rounding takes place rather due to physiological reasons than to any other. The fact is that any back vowel in English is produced with rounded lips, the degree of rounding is different and de­pends on the height of the raised part of the tongue; the higher it is raised the more rounded the lips are. So lip rounding is a pho­neme constitutive indispensable feature, because no back vowel can exist without it.

Our next point should be made about another property of English vowel sounds that is traditionally termed checkness. This quality depends on the character of the articulatory transi­tion from a vowel to a consonant. This kind of transition (VC) is very close in English unlike Russian. As a result all English short vowels are checked when stressed. The degree of checkness may vary and depends on the following consonant. Before fortis voiceless consonant it is more perceptible than before a lenis voiced consonant or sonorant. All long vowels are free.

It may be well to mention that though this characteristic has no phonological value it is of primary importance for Russian learners of English. It should be remembered that since all Rus­sian vowels are free special attention should be drawn to making English short vowels checked. It is not the length of vowels that should be the point of attention but the character of the transi­tion of a vowel into a consonant. Such words as body, seven, bet­ter, matter should be divided into syllables in such a way that the vowels should remain checked unlike Russian Боря , Сева , бита , мята .

At this point we are ready to consider another articulatory characteristic of English vowels, that is their length or quantity.

The English monophthongs are traditionally divided into two varieties according to their length:

83

a) short vowels: [ı], [e], [æ], [ɒ], [u], [ʌ], [ə];

b) long vowels: [ı:], [a:], [ɔ:], [з:], [u:].

We should point out that vowel length or quantity has for a long time been the point of disagreement among phoneticians.

It is common knowledge that a vowel like any sound has physical duration — time which is required for its production (ar­ticulation). When sounds are used in connected speech they cannot help being influenced by one another. Duration is one of the characteristics of a vowel which is modified by and depends on the following factors:

1) its own length,

2) the accent of the syllable in which it occurs,

3) phonetic context,

4) the position of the sound in a syllable,

5) the position in a rhythmic structure,

6) the position in a tone group,

7) the position in a phrase,

8) the position in an utterance,

9) the tempo of the whole utterance,

10) the type of pronunciation,

11) the style of pronunciation.

The problem the analysts are concerned with is whether va­riations in quantity or length are meaningful (relevant), that is whether vowel length can be treated as a relevant feature of English vowel system.

Different scholars attach varying significance to vowel quan­tity.

The approach of D. Jones, an outstanding British phonetician, extends the principle, underlying phonological relevance of vow­el quantity (64). That means that words in such pairs as [bid] — [bi:d], [sıt] — [si:t], [ful] — [fu:l], ['fɔ:wз:d] (foreword) — ['fɔ:wəd] (forward) are distinguished from one another by the opposition of different length, which D. Jones calls chronemes. The difference in quantity is considered to be decisive and the difference in quality (the position of the active organ of speech) is considered to be subordinate to the difference in quantity. According to the point of view of the outstanding Soviet phonetician V. A. Vassilyev, English is not a language in which chronemes as separate prosodic phonological units can exist (79, p. 204).

If a phonetician wants to approach this aspect from phono­logical point of view he should base his theoretical conclusion on the two laws characterizing any system:

84

1. A relevant feature must characterize a number of units. Let us take a sample of palatalization in Russian. Compare: ел — ель , рад — ряд , нов — новь and so on. These oppositions form a correlation system. Any correlation should have a number of op­positions. A sign of correlation (palatalization in the above-men­tioned example) is a distinctive feature of a number of pho­nemes. The analysis of English vowels shows that they can hardly form quantitative correlation. For the sake of economy the following correlation is often brought about.

Let us analyze each of these pairs.

In actual speech the sounds [i:] and [u:] are normally realized in RP as diphthongized vowels. So [ı] and [u] are opposed to diphthongoids but not to long monophthongs.

The opposition [з:] — [ə] is a fairly specific one because the [ə] phoneme never occurs in a stressed syllable and forms the core of unstressed vocalism in English. The phoneme [з:] seldom occurs in an unstressed syllable.

The opposition [a:] — [ʌ] is arbitrary. As a result there is only one pair of opposed phonemes remaining, e.g. [ɔ:] — [ɒ]. That means that quantitative correlation exists only in one opposi­tion, so on this ground it cannot be treated as a phonologically relevant feature.

2. A feature can be systemic if it does not depend on the con­text. As to the absolute length of English historically long and historically short vowels it varies and depends on a lot of factors, the first being phonetic context. A. C. Gimson, for example, points out that [i:] in beat is only half about as long as the [i:] of bee and may approximately have the same duration as the [ı] vowel of bid because it is generally known that a voiced conso­nant following a vowel increases its length (57). But still the words bid and bead are perceived as different words because the vowels are different in quality, [ı] being front retracted, a pure monophthong, and [i:] being front close (narrow) and a diph­thongized vowel. The conclusion that follows is that vowel quantity cannot be considered a minimal distinctive feature since it varies under the influence of different phonetic context. So it is an incidental feature that characterizes sounds of a certain quali­ty. It is worth noting here that an element accompanying anoth­er element cannot be a sign itself and therefore cannot be classed as part of a system. This is one of the basic laws of any system. Summarizing we may say that this is the approach to quantity

85

of English vowels from phonological point of view. It is share by all Soviet specialists in English phonetics as well as by most modern British phoneticians.

It may be well to mention that the [æ] vowel being classed historically short tends to be lengthened in Modern English, especially before lenis consonants [b], [d], [g], [ʤ], [m], [n], [z]. In this^: position [æ] has the same quantity as long vowels [i:], [a:], [ɔ:]J [u:], [з:]. This extra length, as A. C. Gimson points out, serves as additional distinctive feature and the qualitative — quantitative relation of [æ] — [e] tends to become of the same type as [i:] — [i] (57). From this point of view [æ] can possibly belong to the sub class of long vowels, and consequently the twelve English Iong vowel phonemes may be divided into six phonemic pairs which| members differ both in quality and in quantity, and of the two factors it is likely that the quality carries the greater contrastive weight.

There is one more articulatory characteristic that needs our, attention. That is tenseness. It characterizes the state of the or­gans of speech at the moment of production of a vowel. Special instrumental analysis shows that historically long vowels are-tense while historically short vowels are lax. This characteristic is of extraphonological type so tenseness may be considered as indispensable concomitant feature of English long vowels. On this ground it may be included into classificatory description of vowels because it might be helpful in teaching the students of English since there are no tense vowels in Russian.

Summarizing we could say that phonological analysis of ar­ticulatory features of English vowels allows to consider function­ally relevant the following two characteristics:

a) stability of articulation,

b) tongue position.

The rest of the features mentioned above, that is lip position, character of vowel end, length, and tenseness are indispensable constituents of vowel quality. Though they have no phonologi­cal value they are considerably important in teaching English phonetics.

So far we have given a lot of attention to the problems of classifying English vowels. Vowels as items of vocalic system were analyzed as if pronounced in a stressed position in a word. At this point we should discuss various properties displayed by vowels in unstressed positions and consider them from articula­

86

tory and phonological point of view. It is well-known that a vowel in an unstressed syllable is perceived as very short, weak and indistinct. The unstressed syllables are usually associated with vowels of central or centralized quality [ə], [ı], sometimes [u] and the diphthongs [зu], [aı] (or a syllabic consonant), e.g. among [ə'mʌŋ], before [bı'fɔ:], useful ['ju:sful], tomato [tə'ma:tзu], exercise ['eksəsaız], sudden ['sʌdn].

Also vowels of full quality sometimes occur in unstressed po­sitions, often in borrowed words of Latin and Greek origin, e.g. architect ['a:kıtekt], paragraph ['pærəgra:f], canteen [kæn'ti:n].

These nonreduced vowels in unstressed syllables are typical of all styles of pronunciation.

It is important to mention here that in the Russian language a vowel never preserves its full quality in unstressed position, it is usually reduced. So the English vowels of full quality in unstressed syllables require the teachers' special attention. Cf.: transport ('trænspɔ:t] — транспорт ['транспърт].

Then again partially reduced sounds are found in unstressed positions. They appear in more formal and careful style of pro­nunciation instead of the neutral sound used in informal casual speech. Cf.: phonetics [fзu'netıks — fз'netıks — fə'netıks].

Our next point should be made in connection with the phone­mic status of the neutral sound [ə]. The phonological analysis marks the opposition of the neutral sound to other unstressed vowels, the most common among them being [1]. In the minimal pairs: officers ['ɒfısəz] — offices [ɒfısız]; accept [ək'sept] — except [ık'sept], armour ['a:mə] — army ['amı] the neutral sound is phonologically opposed to the phoneme [ı] with its own distinctive fea­tures capable of differentiating the meaning of lexical units. So the neutral sound [ə] in officers, accept, armour is an independent pho­neme opposed to the [1] phoneme of the minimal pairs given above.

On the other hand, the problem of the phonemic status of the neutral sound has a morphological aspect. In English as well as in Russian there are numerous alternations of vowels in stressed and unstressed syllables between the derivatives of the same root or different grammatical forms of the same word. Cf.:

[æ] — [ə] man — sportsman

[ʌ] — [ə] some — wholesome

[ɒ] — [ə] combine n — combine v

[eı] — [ə] operation — operative

[зu] — [ə] post — postpone

87

The alternated sounds are allophones of one and the same phoneme as they are derivatives of the same lexical units, the same morphemes. Thus the neutral sounds in the examples above are the neutralized allophones of the nonreduced vowels of full formation; so [ə] in sportsman is an allophone of the [æ] phoneme as in man; [ə] in photography is an allophone of the [зu] phoneme as in photograph.

We might conclude by saying that we have tried to look at the consonantal and vocalic systems of the English language from phonological point of view. Applying this sort of analysis enables us to define what properties displayed by English sounds are significant in making them items of a system. It is worth saying that phonologically relevant articulatory features should attract direct attention of a would-be teacher of English because they form the basis of the pronunciation system of the language. Non-relevant but indispensable features should also be acquired being both phonetically correct and necessary for teaching pur­poses. If we want to speak a foreign language in an objectively correct way it is natural we should pay attention to the quality of our sounds which is constituted by articulatory features of both kinds.

 

⇐ Предыдущая12345678910Следующая ⇒

Поделиться: