Phonetic system of the language, its segmental and prosodic (suprasegmental) subsystems, their units (elements and structures) and the relationship between them. Functions of phonetic units. Aspects of speech sounds: articulatory, acoustic, auditory, functional (phonological, linguistic).
Language exists in two main speech forms oral and written. Speech is a manifestation of' language, it is a process of communication by means of language. Both the oral and written speech forms have a material substance. In oral speech the substance is phonic, it is the sound substance or the sound matter. In written speech the substance is graphic.
The phonetic system of a language contains two systems (or levels) — segmental and suprasegmental, or prosodic, each of which is a specially organized language system with a certain number of its units. Segmental units are elementary sounds, vowels and consonants, which form the vocalic and the consonantal subsystems. Prosodic units are syllables, accentual (rhythmic) units, intonation groups, utterances, which form the subsystems of pitch, stress, rhythm, tempo, pauses.
The sound substance is a medium in which the whole system of language is
embodied. Segmental and prosodic units serve to form and differentiate units
of other subsystems of language, the lexical and grammatical units. The grammatical form of a verb or a noun can be changed only by changing the sounds which compose them. By changing the prosodic structure (intonation) of an utterance one changes the meaning of the utterance. It is clearly seen from the utterances of identical lexical and grammatical structures. For example, "Well done?" pronounced with the rising tone, is a question, "Well done!" (with the falling tone) is an exclamation.
Aspects of speech.Sound is a product of human activity. It indicates the speaker's personality (sex, age, individual features) and reveals his physiological and emotional state, geographical origin, education, social status and so on.
Every act of speech presupposes the presence of a person who speaks and a person who listens. The speaker produces sounds; the sounds travel through the air to the listener in the form of complex combinations of sound waves, the listener hears and interprets them. Communication is possible only because the speaker and the listener interpret the sounds as units of the same language. Let us consider the speech chain, which may be diagrammed in simplified form like this:
Consequently, sound phenomena have different aspects, which are closely interconnected: the articulatory aspect, the acoustic, the auditory and the linguistic aspect.
Articulation comprises all the movements and positions of the speech organs necessary to pronounce a speech sound. According to their main sound-producing functions, the speech organs can be divided into the following four groups: the power mechanism; the vibration mechanism; the resonator mechanism; the obstruction mechanism.
The functions of the power mechanism consist in the supply of the energy in the form of the air pressure and in regulating the force of the air stream. The power mechanism includes: the diaphragm, the lungs, the bronchi, the windpipe, or trachea. The glottis and the supra-glottal cavities enter into the power mechanism as parts of the respiratory tract. The vibration mechanism consists of the larynx, or voice box, containing the vocal cords. The most important function of the vocal cords is their role in the production of voice. The pharynx, the mouth, and the nasal cavity function as the principal resonators thus constituting the resonator mechanism. The obstruction mechanism (the tongue, the lips, the teeth, and the palate) forms the different types of obstructions.
The acoustic aspect studies sound waves. The basic vibrations of the vocal cords over their whole length produce the fundamental tone of voice. The simultaneous vibrations of each part of the vocal cords produce partial tones (overtones and harmonics). The number of vibrations per second is called frequency. Frequency of basic vibrations of the vocal cords is the fundamental frequency. Fundamental frequency determines the pitch of the voice and forms an acoustic basis of speech melody. Intensity of speech sounds depends on the amplitude of vibration.
The auditory (sound-perception) aspect, on the one hand, is a physiological mechanism. We can perceive sound waves within a range of 16 Hz-20.000 Hz with adifference in 3 Hz. The human ear transforms mechanical vibrations of the air into nervous and transmits them to brain. The listener hears the acoustic features of the fundamental frequency, formant frequency, intensity and duration in terms of perceptible categories of pitch, quality, loudness and length. On the other hand, it is also apsychological mechanism. The point is that repetitions of what might be heard as the same utterance are only coincidentally, if ever, acoustically identical. Phonetic identity is a. theoretical ideal. Phonetic similarity, not phonetic identity, is the criterion with which we operate in the linguistic analysis.
Functional aspect. Phonemes, syllables, stress, and intonation are linguistic phenomena. They constitute meaningful units (morphemes, words, word-forms, utterances). Sounds of speech perform different linguistic functions
Word stress. The nature of English word stress. Linguistically relevant degrees of word stress and accentuation patterns in English. Accentuation tendencies in English (the recessive, the rhythmic and retentive tendencies, the semantic factor). The functional aspect of word stress.
A word, as a meaningful language unit, has a definite phonetic structure. The phonetic structure of a word comprises not only the sounds that the word is composed of and not only the syllabic structure that these sounds form, it also has a definite stress pattern.
Monosyllabic words have no stress pattern, because there can be established no correlation of prominence within it. Yet as lexical units monosyllables are regarded as stressed. Word stress should not be confused with utterance stress. The placement of utterance stress is primarily conditioned by the situational and linguistic context. It is also conditioned by subjective factors. The stress pattern of a word is conditioned only by objective factors: pronunciation tendencies and the orthoepic norm.
As stated above, the auditory impression of stress, is that of prominence. . The effect of prominence may be produced by a greater degree of loudness, greater length of the stressed syllable, some modifications in its pitch and quality. To produce the effect of prominence the following acoustic parameters interact: intensity, duration, frequency, formant structure.
In different languages stress may be achieved by various combinations of these parameters. Depending upon which parameter is the principal one in producing the effect of stress, word stress in languages may be of different types. There are languages with dynamic word stress. The stressed syllables are louder than the unstressed ones. Chinese, Japanese, Vietnamese are languages with musical word stress (or tonic word stress). The meaning of the words in those languages depends on the pitch levels of their syllables. In languages with quantitative word stress the effect of stress is mainly based on the quantity of the sound, i.e. its duration. In such languages vowels in the stressed syllables are always longer than" vowels in unstressed syllables (Russian). Until recently, English word stress was considered to be dynamic, as stress was generally correlated with loudness. But numerous investigations have made it clear that that English word stress is of a complex nature. Thus, as far as English word stress, is. concerned, relative prominence in the listener's mind is created by an interaction of four acoustic parameters: intensity, fundamental frequency, duration and formant structure. There may also be a combination of any of these parameters. What complicates the matter is that in English a vowel in an unstressed syllable may be non-reduced and longer than in a stressed syllable (as in "pillow", "compound"). Vowels differ in their intensity as well. Besides that, the quantity of English vowels also differs: in identical phonetic environment an open vowel is longer than a close vowel. Linguistically relevant degrees of word stress.One of the main questions for the linguist is to determine the number of contrastive degrees of word stress in a language. A polysyllabic word has as many degrees of prominence as there are syllables in it. But not all these degrees of prominence are linguistically relevant. The problem is to determine which of these degrees are linguistically relevant. There are two views of the matter. Some (e.g. D.Jones, V.Vassilyev ) consider that there are three degrees of word stress in English: primary (or strong stress), secondary (or partial stress) and weak (the so-called "unstressed" syllables have weak stress). Secondary stress is chiefly needed to define the stress pattern of words containing four or more syllables, and compound words. E.g. "e,xami'nation", ",qualifi'cation". But there are certain positions in the stress patterns of English words where the vowel duration is considerable, though the syllable it occurs in does not actually bear primary or secondary stress. (e.g. "elevate", "recognize", "occupy"). On this account some American linguists (G.Trager, A.Hill) distinguish four degrees of word stress: primary stress / ' / (as in "c'upboard"), secondary stress / ^/ (as in "discr^imination"), tertiary stress / `/ (as in "´anal`yse"), weak stress / ˇ/ (as in "c´upboˇard", but very often the weakly stressed syllable is left unmarked). Secondary stress generally occurs before the primary stress (as in "examination"), while tertiary stress occurs after the primary stress (as in "handbook", "specialize"). The distinction between secondary and tertiary degrees of stress is too subtle to be noticed by an untrained ear and there are no words in English the meanings of which depend on whether their stress pattern is characterized by either secondary or tertiary stress. That is why the stress pattern of English words may be defined as a correlation of three degrees of stress. There are languages in which stress always falls on the first syllable (as in Czech and Finnish), or on the last syllable (as in French and Turkish). Word stress in such languages is said to be fixed. English word stress is said to be .free. It is free in the sense that stress is not fixed to any particular syllable in all the words of the language. G.Torsuyev distinguishes more than 100 stress patterns, which he groups into 11 main types. The most common among them are: words with one primary stress, as in "after", words with two primary stresses, as in "week-end", words with one primary and one secondary stress, as in "hairdresser", "magazine"). They are the most productive types of stress patterns. Though word stress in English is called free, there are certain tendencies in English which regulate the accentuation of words. According to the recessive tendency, stress falls on the first syllable which is generally the root syllable (e.g. "mother", "father”) or on the second syllable in words which have a prefix of no special meaning (e.g., "become", indeed") . In the English language a considerable part of the vocabulary consists of monosyllabic words, some of which are stressed, others not. This created the rhythmic tendency to alternate stressed and unstressed syllables. According to the rhythmic tendency stress is on "the 3rd syllable from the end (intensity, possibility). It is the usual way of stressing four— syllabled words (e.g. "political, de'mocracy). In words with more than four syllables we very often find the influence of both the rhythmic and the recessive tendencies (e.g. "i'ndi'visibie, phondlogical"). It has also been noticed that the stress of the parent word is often retained in the derivatives ('personal —personality). This regularity is sometimes called theretentive tendency.There is a tendency to stress the more important elements in words(semantic factor), such meaningful prominence is given to negative prefixes "un—", "in—" ( "unknown"), such prefixes as "ex—", "vice-", "sub-", "under—", ("ex—president"), suffix "—teen" ("fourteen"), semantically important elements in compound words (''well-known", "bad-tempered"). These are the tendencies that to some extent regulate the placement of stress in English words and condition their stress patterns.
The functional aspect of word stress.
Word stress has a constitutive function, as it moulds syllables into a word by forming its stress pattern. Without a definite stress pattern a word ceases to be a word and becomes a sequence of syllables. Word stress has a distinctive function in English, because there exist different words in English with analogous sound structure which are differentiated in speech only by their stress patterns (‘abstract, n.- ab’stract, v.)
Word stress has an identificatoryfunction as well, because the stress patterns of words enable people to identify definite combinations of sounds as meaningful linguistic units.
4. Standard pronunciation and its variants. The orthoepic norm. English-based pronunciation standards of English. RP (Received pronunciation) as the norm of pronunciation in Great Britain. Current tendencies in RP. Regional Non-RP accents of English. Stylistic variation in pronunciation. American-based pronunciation standards of English.There exist numerous varieties of pronunciation in any language, the English language as well. The pronunciation of almost every locality in the British Isles has peculiar features that distinguish it from the pronunciation of other localities. Besides, pronunciation is socially influenced. It reflects class distinctions, education, upbringing. The varieties that are spoken by a socially limited number of people and used only in certain localities are called dialects. There are therefore local dialects and social dialects. Moreover, there are innumerable individual differences, called idiolectal differences. At the same time all these varieties have much, more in common than what differentiates them. They are varieties of one and the same language. Dialects have some peculiarities in pronunciation, vocabulary and grammatical structure. Dialect speakers are as a rule the less educated part of the population. Dialects enrich the language and make it more lively and fresh. Among the most well-known dialects one should mention Cockney (spoken by the less educated part of the Londoners). In British English phoneticians generally distinguish three main regional types of pronunciation: Southern, Northern and Scottish regional types of English. Besides there may a non—regional type of pronunciation that is not native to any particular locality in the country. One of the types of pronunciation, generally the one that is spoken by the educated people in the capital, is recognized as the orthoepic norm (the standard pronunciation adopted by native speakers as the right and proper way of speaking. It is used by the most educated part of the population). The orthoepic norm is based on the variants of pronunciation that are widely used in actual speech. New pronunciations which are in common use gradually become 'acceptable’ and are included into the norm. But some of the pronunciations, which had been acceptable, fall out of use are labeled as old fashioned and are excluded from the norm (chemist [kemist] - was [kimist]) Variation of the orthoepic norm is a natural objective phenomenon, which reflects the development of language. The orthoepic norm also includes a number of stylistic norms, none of which can be considered "neutral" and acceptable in all circumstances. The pronouncing dictionaries record the well-established pronunciations as first variants. The less frequent variants of pronunciation are generally recorded as secondary variants. In connected speech the sound structures of words are greatly modified under the influence of rhythm, tempo and utterance stress. The dictionaries register the model pronunciations and only partly reflect the actual pronunciation in general use. But the prosodic norm is an important component of the orthoepic norm. The intonation of a speaker may also be acceptable or unacceptable. Therefore, the main factors that condition variation of the orthoepic norm are social, territorial and stylistic factors. RP (Received pronunciation) as the norm of pronunciation in Great Britain.Received Pronunciation (RP) was accepted as the phonetic norm of English about a century ago. It is mainly based on the Southern English regional type of pronunciation, but it has developed its own features which have given it a non-regional character RP is spoken all over Britain by a comparatively small number of Englishmen who have had the most privileged education in the country — public school education. RP is not taught at these schools, "it is absorbed automatically by the pupils. It is often referred to as the 'prestige accent'.
Though RP is carefully preserved by the public schools and the privileged class in England, the RP of today differs in some respects from the former refined RP used half a century ago. A.Gimson distinguishes three varieties of RP today: (1) the conservative RP used mainly by the o older RP speakers, (2) the general RP heard on radio and TV, that is less conservative and has received a number of changes, (3) the advanced RP mainly used by the younger RP speakers (many more changes). Nowadays PR=Southern Eng. RP has been investigated and described more thoroughly than any other type of English pronunciation. That is why RP pronunciation is often accepted as the teaching standard in many countries where English is taught as a foreign language. But there are many educated people in Britain who do not speak RP, though their English is good and correct. They speak Standard English with a regional type of pronunciation. D. Abercrombie divides English people by the way they talk into three groups: (1) RP speakers of Standard English (those who speak without any local accent); (2) non—RP speakers of Standard English (speak Standard English with a regional accent); (3) dialect speakers. Regional Non-RP accents of English. Scholars often note that it is wrong to assume that only one type of pronunciation can be correct. This primarily concerns the Northern and the Scottish types of pronunciation, which are used by many educated people in Britain. The main distinctions of the Northern type of English pronunciation, as opposed to RP, are: [æ]is more open as in ‘bad’, [υ]is used instead of [^] could=cud and so on. Scottish type of pronunciation from RP, are as follows: instead of RP /3: / they use the sequences /ir/, /er/ or /Ar/, all vowels are short no distinction in the length of the vowels "cot" and "caught”. There are certain peculiarities jn the intonation of the Scottish type of English pronunciation. Special Questions may end with a high level tone after a fall on the interrogative word. One of the best examples of a local dialect is Cockney. The final /ŋ /sounds like /n/ (e.g. "something"), /ei/ is an almost open vowel, so that it reminds of /ai/ (e.g. "take"). American-based pronunciation standards of English. English is the national language in the USA, Australia, New Zealand, and of a great part of the population in Canada. Each of those nations has its own orthoepic norm which exists alongside of regional types and numerous dialects. American English (AE), which is a variant of the English language, has developed its own peculiarities in vocabulary, grammatical structure and pronunciation. The most widely used regional types of AE pronunciation are the Eastern, the Southern, and the GA (General American) types, the latter spoken mainly in the Middle Atlantic States Region. The GA pronunciation is usually referred to as the standard pronunciation of AE. The peculiarities of GA lie in 1) the pronunciation of sounds and sound combinations; 2) differences in the stress patterns of words, and 3) differences in intonation. Peculiarities of pronunciation of GA sounds and sound combinations as compared to those of RP. [∫]is voiced in words like "excursion" [ჳn], "version" [ჳn], [ј] is omitted before /u/ (e.g. "duty" /du:ti/), /k/ is omitted before /t/ (e.g. "asked" /æst/) and so on. American speakers make much greater use of secondary stress in polysyllabic words than British speakers do. In words which end in "—ary", "—pry","—ery", "—mony", "—ative" the first syllable in the suffix bears tertiary stress. The most frequent intonation contour for statements and requests in GA is the tune, beginning low, rising to a high level, and then steadily falling. "Rising" tunes that rise from a low pitch level and end on a high pitch level occur with some General Questions, especially in situations where a very polite form is desirable. The RP Special Question pronounced with a rising tone (polite questions) are perceived by the Americans as questions implying curiosity. Stylistic variation in pronunciation. Language functions in two main forms: the spoken language and the written language. The main concern of phonetics is to investigate the varieties of the spoken language. Scholars distinguish a number of functional styles of the written language such as belles-lettres, publicist, newspaper, style of official documents and that of scientific prose. The styles of the spoken language are not as yet unanimously defined. But each native speaker uses several varieties of the language, which differ only phonetically, nevertheless they are easily identified by all the native speakers. The main circumstances of reality that cause phonetic modifications in speech are as follows (a) the aim of speech, (b) the extent of spontaneity of speech (unprepared speech, prepared speech, etc.); (c) the use of a form of speech which may either suggest only listening, or both listening and an exchange of remarks (a lecture, a discussion), (d) social and psychological factors (a friendly conversation, a quarrel). These factors are termed extra linguistic factors. Different ways of pronunciation, caused by extra linguistic factors and characterized by definite phonetic features are called phonetic styles, or styles of pronunciation. There is no generally accepted classification of pronunciation styles. All the classifications differ not only in the, number of styles which are singled out. The main distinction between them is that they are based on different principles: the degree of carefulness (the full style and the colloquial style L.Shcherba) the extent (степень) of formality (familiar colloquial, formal colloquial, public—speaking style and public-reading style), the rate of speech (the rapid familiar style, the slower colloquial style, the natural style used in addressing an audience, the acquired style of the stage, and the acquired styles used in singing), the social situations. Elision, reduction and assimilation may signal stylistic differences. Besides theses segmental features there are prosodic features which enable people to distinguish between different phonetic styles (loudness, speech tempo, pauses).
2. The functional aspect of speech sounds.When we talk about the sounds of lang., the term "sound" can be interpreted in 2 different ways: we can say that [t] & [d], for ex., are 2 different sounds in Eng. (ten-den) & we know that [t] in let us & [t] in let them are not the same. It is clear that the sense of "sound" in these 2 cases is different. To avoid this ambiguity, linguists use 2 separate terms: phoneme & allophone. The phoneme is a minimal abstract ling. unit realized in speech in the form of speech sounds opposable to other phonemes of the same lang. to distinguish the meaning of morphemes & words. The ling. role of the phoneme is clearly seen from its ling. functions: the distinctive function (distinguishes 1 morpheme from another, 1 word from another (bath-path) / 1 utterance from another (He was heard badly - He was hurt badly). The phoneme is realized in speech in the form of speech sounds, its allophones −>the phonemes constitute the material form of morphemes (constitutive f.). The phoneme performs the recognitive / identificatory function, because the use of the right allophones & other phonetic units facilitates normal recognition. The use of the right phoneme is not the only significant factor, the use of the right allophone is not much less important. So the phoneme is a material & objective unit as well as an abstract & generalized one at the same time. Thus both phonemes & sounds are simply 2 sides of the same phenomenon — the sound substance of lang., which can be analyzed on either the phonemic (functional) level or the allophonic (variational) level.
Every phoneme displays a vast range of variation in connected speech. We distinguish Idiolectal variation − the individual peculiarities of articulating sounds, which are caused by the shape & form of the speaker's speech organs & by his articulatory habits (a speaker may say "thish ish" for "this is). Diaphonic var. affects the quality & quantity of particular phonemes. It is caused by concrete historical tendencies active in certain localities (in some dialects / æ / is much longer than the standard sound). Diaphonic variants do not affect intelligibility of speech they inform the listener about the speaker's origin & his social standing. The listener easily notices both idiolectal & diaphonic variants & understand it. The less noticeable is allophonic var., which is conditioned by phonetic position & phonetic environment.
In connected speech the sounds undergo various modifications, under the influence of neighboring sounds & the intonation patterns they occur in. The modifications are observed within both words & word boundaries. Types of modification: OF CONSONANTS1. Assimilation -the adaptive modification of a consonant by a neighboring consonant in a speech chain 1.1. Place of articulation • t, d > dental before [ð, θ]: eighth, at the, • t, d > affricates before [j]: graduate, 1.2. Manner of articulation • loss of plosion: glad to see you, • nasal plosion: sudden, at night, • lateral plosion: settle, at last.1.3. Work of the vocal cords • voiced > voiceless: newspaper, gooseberry Notice: In Eng. typical assimilation is voiced > voiceless; voiceless > voiced is not typical. 1.4. Degree of noise • sonorants > are partially devoiced after [p, t, k, s]. 2. Accommodation is used to denote the interchanges of VC or CV types. 2.1. Lip position • consonant + back vowel: pool, rude, • consonant + front vowel: tea, sit, keep (spread) 3. Elision - a complete loss of sounds, both vowels & consonants. 3.1. Loss of [h] in personal & possessive pronouns & the forms of the auxiliary v.have. 3.2. [l] tends to be lost when preceded by [o:]:always, already, 3.3. In cluster of consonants: next day, just one. 4. Inserting of sounds - a process of sound addition 4.1. Linking [r] (potential pronunciation of [r]): car owner, 4.2. Intrusive [r]: [r] is pronounced where no r is seen in the spelling china and glass: it is not recommended to foreign learners. OF VOWELS 1.Reduction- qualitative / quantitative weakening of v. in unstressed positions. 2. Accommodation 2.2 Positional length of v.: knee - need – neat. 2.3. Nasalization: preceded or followed by [n, m]: never, The phonemic inventory of Eng.The 1stproblem of phonological analysis is to establish the phonemes in a definite lang. This can be carried out only by phonological analysis based on phonological rules. There are 2 methods to do that: 1.The distributional method is based on the phonological rule that different phonemes can freely occur in the same position, while allophones of the same phoneme occur in different positions & cannot be phonologically opposed to each other (as /p/ & /b/ can freely occur in the same phonetic context ("pea" — "bee"), they are consequently different phonemes." But one cannot find [p] aspirated & [p] non-aspirated in the same phonetic position in Eng. In Eng. they are allophones of the same phoneme). 2.The semantic method is based on the phonological rule that a phoneme can distinguish words when opposed to another phoneme or zero in an identical phonetic position. The opposition /z/ − Itl is called a phonological opposition. The opposition/z/−/—/is called a zero (phonological) opposition. The semantic method of identification of the phonemes in a lang. attaches great significance to meaning. The investigator studies the function of sounds by collecting minimal pairs of words in the lang. If 2 speech sounds distinguish words with different meanings, they form a phonological opposition - realizations of 2 different phonemes. If not, they are allophones of the same phoneme ([s] & [t] are realizations of 2 different phonemes ("sea"—"tea",), while [t] aspirated & [t] none—aspirated are allophones of the same phoneme − they cannot distinguish words). Difficulties as to the phonemic status of certain sounds. The problem is whether there is a schwa vowel I ə I phoneme in Eng. It can form phonological oppositions with a number of other phonemes & can distinguish words (/ə/ vs. / ı / accept — except).
If the Eng. the sounds /t∫/ /dჳ/, /tr/, /dr/, /ts/, /dz/ are monophonemic & should be included into the phonemic inventory? N.Trubetzkoy −rules which help to determine whether a sound of a complex nature is monophonemic: a phoneme is indivisible; a ph. is produced by one articulatory effort; the duration of a ph. should not exceed that of other ph. in the lang. −> It∫I & /dჳ/ in words like "cheese, each" are monophonemic. /ts/, /dzl − biphonemic combinations. As for /tr/, /dr/ ("tree, dream") their phonemic status will remain undecided. D.Jones calls them affricates. Most phoneticians regard /tr/ & /dr/ as biphonemic clusters. If Eng. diphthongs & "triphthongs" are monophonemic / biphonemic clusters in Eng.? The syllabic & articulatory indivisibility of Eng. diphthongs & their duration clearly determine their monophonemic character in Eng. /aiə /, /aυə/ −>biphonemic clusters of a diphthong with the schwa vowel.In such a way it has been established that in RP there are 12 vowel phonemes: [i:,1, e, ae, a , a:, υ, u:, ə:, ^, 3:, ə], 8 diphthongs: 3glides to /I/- /ei , ai, ai/, 2glides to /υ/ - /oυ, aυ/; 3glides to/ə/-/iə, εə, υə/. 24 consonant phonemes: [n, t, d, s, l, ð, v, m, k, w, z, r, b, f, p, h, ŋ, g, ∫, j, dჳ, t∫, θ, ჳ].V.A. Vassilyev includes the diphthong Iσə / & the labiovelar fricative /ო/ ("pour- paw", "which - witch") into the phonemic inventory and states that in RP there are 21 vowel phonemes and 25 consonant phonemes.
The 2nd problem of phonological analysis is the identification of the inventory of distinctive features on which all the phonological oppositions in the lang. are based. The problem is to decide which of the features of a group of common sounds in a certain lang. are phonologically relevant ( its constant distinctive features-distinguish the phoneme from all the other phonemes of the lang.) & which of them are irrelevant, or incidental (distinguish the allophone from all the other allophones of the phoneme. The substitution of one irrelevant feature does not affect communication.