Consequently, though allophones of the same phoneme pos­sess similar articulatory features they may frequently show con­siderable phonetic differences

It is perfectly obvious that in teaching English pronunciation the difference between the allophones of the same phoneme should be necessarily considered. The starting point is of course the articulation of the principal allophone, e.g. /d-d-d/: door, double, daughter, dark, etc. Special training of the subsidiary allo­phones should be provided too. Not all the subsidiary allo­phones are generally paid equal attention to. In teaching the pronunciation of [d], for instance, it is hardly necessary to con­centrate on an allophone such as [d] before a front vowel as in Russian similar consonants in this position are also palatalized. Neither is it necessary to practise specially the labialized [d] after the labial [w] because in this position [d] cannot be pronounced in any other way.

Carefully made up exercises will exclude the danger of for­eign accent.

• Allophones are arranged into functionally similar groups, that is groups of sounds in which the members of each group are' tytiot opposed to one another, but are opppsable to members of /any other group to distinguish meanings in otherwise similar se-Vfcjucences. Consequently allophones of the same phoneme never $ pccur in similar phonetic contexts, they are entirely predictable (/according to the phonetic environment, and thus carry no useful ' *"(information, that is they cannot differentiate meanings.

But the_phones which are realized in speech do not corre­spond exactly to the allophone predicted by this or that phonetic environmenit. They are modified by phonostylistic, dialectal and individual factors. In fact, no speech sounds are absolutely alike. Phonemes are important for distinguishing meanings, for knowmg^whether, for instance, the message was take it от tape it. But there is more to speaker — listener exchange than just the "message" itself. The listener may pick up a variety of information about the speaker: about the locality he lives in, regional origin, his social status, age and even emotional state (angry, tired, excit­ed), and much other information. Most of this other social infor­mation comes not from ph^qnejnicdistinctions, but from phonet ic ones. Thus, while phonemic evidence is important for lexical and grammatical meaning, most other aspects of a communication are conveyed by more subtle differences of speech sounds, requiring more detailed description at the phonetic level. There is more to a speech act than just the meaning of the words.

Thirdly, allophones of the same phoneme, no matter how dif­ferent their articulation may be, function as the same linguistic unit. The question arises why phonetically naive native speakers seldom observe differences in the actual articulatory qualities be­tween the allophones of the same phonemes. The native speaker is quite readily aware of the phonemes of his language but much less aware of the allophones: it is possi­ble, in fact, that he will not hear the difference between two allo­phones like the alveolar and dental consonants [d] in the words bread and breadth even when a distinction is pointed out; a cer­tain amount of ear-training may be needed. The reason is that the phonemes have an important function in the language: they differentiate words like tie and die from each other, and to be able to hear and produce phonemic differences is part of what it means to be a competent speaker of the language. Allophones, on the other hand, have no such function: they usually occur in different positions in the word (i.e. in different environments) and hence cannot be opposed to each other to make meaningful distinctions.

For example the dark [t] occurs following a vowel as in pill, cold, but it is not found before a vowel, whereas the clear [1] only occurs before a vowel, as in lip, like. These two vowels cannot therefore contrast with each other in the way that [1] contrasts with [r] in lip — rip or lake — rake, there are no pairs of words which differ only in that one has [I] and the other — [1].

So the answer appears to be in the functioning of such sounds in the language concerned. Sounds which have similar functions in the language tend to be considered the" "same" by the community using that language while those which have dif­ferent functions tend to be classed as "different". In linguistics, as it has been mentioned above, function is generally understood as the role of the various elements of the language in distinguishing the meaning. The function of phonemes is to distinguish the meaning of morphemes and words. The native speaker does not notice the difference between the allophones of the same pho­neme because this difference does not distinguish meanings.

In other words, native speakers abstract themselves from the differen£eJ^wjejen^jhe_aUp^hjQines of the same phoneme because L it has no fjjnctionaJLyjlue. The actual difference between the al­lophones of the phoneme [d], for instance, does not affect the meaning. That's why members of the English speech community d6~~n6Trealize that in the word dog [d] is alveolar, in dry it is post-alveolar, in breadth it is dental. Another example. In the Russian word посидит the stressed vowel [aj is more front than it is in the word посадка. It is even more front in the word ся­дет. But Russian-speaking people do not observe this difference

ause the three vowel sounds belong to the same phoneme „«Mxd thus the changes in their quality do not distinguish the gjflneaning. So we have good grounds to sia"te that the phoneme is tpfrin abstract linguistic unit, it is an abstraction from actual speech Ijffsounds, that is allophonic modifications.

As has been said before, native speakers do not observe the difference between the allophones of the same phoneme. At the same time they realize, quite subconsciously of course, that allo­phones of each phoneme possess a bundle of distinctive features, that makes this phoneme functionally different from all other phonemes of the language concerned. This functionally relevant bundle of articulatory features is called the invariant of the* pho­neme. Neither of the articulatpry features thatjprm thej^ariant of the p honeme can be changed without affecting the meaning. All the" allophones of the phoneme [d], for instance, are occlu­sive, forelingual, lenis. If occlusive articulation is changed for.constrictive one [d] will be replaced by [z], cf. breed— breeze, deal — zeal; [a] will be replaced by [g] if the forelingual articula­tion is replaced by the backlingual one, cf. dear — gear, day — gay. The lenis articulation of [d] cannot be substituted by the for-tis one because it will also bring about changes in meaning, cf. dry— try, ladder — latter, bid — bit. That is why it is possible-to state that occlusive, forelingual and lenis characteristics of the phoneme [d] are generalized in the mind of the speaker into!! what is called the invariant of this phoneme.

On the one hand, the phoneme is objective real, because it is realized in speech in the material form of speech sounds, its allo­phones. On the other hand, it is an abstract language unit. That is why we can look upon the phoneme asa di alectic al unity of the materialand abstract aspects. Thus we may state" that it ex­ists in the material lorrn oTspeech sounds, its allophones. Speech sounds are necessarily allophones of one of the phonemes of the language concerned. All the allophones of the same phoneme have sjojme_ajrtic^

possess thejsame jnyariant. Simultaneously each allophone pos­sesses quite particular phonetic features which may not be traced in the articulation of other allophones of the same pho­neme. That is why while teaching pronunciation we cannot ask our pupils to pronounce this or that phoneme. We can only teach them to pronounce one of its allophones. It is interesting to note that Russian word stress may have stylistic distinction and poetic usage, cf. молодец — молодец, девица — девица, шёлковый — шелковый.

The complicated system of the accentual structure of English words makes teacher trainees be very attentive to the subject. The typical mistakes of Russian learners in the sphere of word stress are the mispronunciation of: 1) words with the main and secondary stresses (,conversational); 2) words with two equal stresses in connected speech ("up'stairs, 'reorganize); 3) words with the full vowel in the unstressed syllable ('architect).

The instability of English accentual structure of words presents much difficulty for Russian learners. Students' attention should be attracted to English multisyllabic words the accentual structure of which is regulated by the rhythmical tendency and the use of the secondary stress in those words, as it has no anal­ogy in the Russian language, compare: 'transportation — транс­портировка, de,mocrati 'zation — демократизация.

Another group of words presenting difficulty for Russian learn­ers is large group of compounds which are marked either by two equal stresses (compound adjectives) or by one stress (compound nouns). The semantic factor in defining the accentual structure of compounds should be most decisive, as it has been illustrated above. One more group of words requires learners' attention, the group which forms accentual oppositions of different parts of speech by way of conversion accompanied by the shifting of stress, e.g. 'combine (n) — combine (v), 'insult (n) — in'sult (y).

In case of doubt it Is advisable to consult a pronouncing dictionary.