Phoneme and allophones

Dynamic models

Comparing the sound systems of English and Russian

The vowel system of English

The consonant system of English

Phoneme and allophones

PHONOLOGY

LECTURE 2

How many vowels and consonants are there in the system of English? Or in Russian? Or in any other language of the world?

While establishing the set of relevant consonants and vowels we can find numerous phonetic differences in the sounds. What we have to decide is:

Is it "the same sound" or "a different sound"? An innocent native speaker of English will possess the necessary linguistic knowledge, quite unconsciously, saying that /k/ in both Kate and skate, /p/ in both pie and spy, as well as /t/ in take and stake are "the same sounds", although, as we know from phonetic evidence, they are not quite the same: in each pair of words the first alternative is aspirated and the second one is not:

[k^heit - skeit], fp^hai - spai], [t^heik - steik].

The reason for the "sameness" is that in English aspirated and non-aspirated realizations or allophones appear in different phonetic contexts and never compete with each other: aspirated consonants appear before stressed vowels unless they are preceded by /s/; non-aspirated realizations appear when /p/, /t/, /k/ are preceded by /s/. The two allophones are said to be in complementary distribution which means that they complement each other, and belong to the set of all the (phonetically different) allophones of one phoneme, /k/, for example: Kate, skate, key, book, cushion, tobacco.

However, the native speakers of Hindi or Korean may have a notion that aspirated and non-aspirated [k^h] and [k] are, in fact, different sounds because in their native languages aspirated and non-aspirated plosives may appear in the same phonetic context and contrast the meanings of words; more than that, in Korean they distinguish non-aspirated, weakly aspirated and strongly aspirated voiceless consonants. Thus in Hindi and Korean [k^h] and [k] are different phonemes because they may occur in the same context and cause a change in the meaning of words, i.e. a semantic contrast:

Hindi: [kal] span of time, [k^hal] skin [pal] nurture, [p^hal] knife blade [Ы] beat, [Ы] platter.

Conversely, in English and in Russian /r/ and /1/ are different phonemes as they may appear in the same phonetic context and cause semantic difference:

English: rip, lip — [rip], [lip], right, light- [rait], [lait], Paris, palace — f'paens], ['paelis]. Russian: ром — лом — сер — сел — кора — кола.

In Korean /r/ and /1/ are allophones of the same phoneme: /1/ appears at the beginning and at the end of the syllable, while /r/ is realized between vowels:

Korean: [mal] horse, [mare] at the horse [mul] water, [mure] at the water.

Allophonic variation is predictable as it depends on the rules which operate in certain phonetic contexts. This is the way Korean girls sang the once popular song in Russian: Мирион, мирион арихлоз...

Another example: in English the dental voiced fricative /9/ and the alveolar voiced plosive stop /d/ are different phonemes: they may appear in the same phonetic context and by replacing one for the other we get a dif­ferent word, i.e. bring about a semantic contrast:

English: [dei] day, [3ei] they

From the point of view of distribution the English dental (or American interdental) sound /9/, as well as the voiced plosive /d/ have parallel distribution: both may occur at the beginning, at the end of a word or a syllable and between vowels: this, mother, booth, day, today, Ted. In Castillian Spanish the dental plosive [d] and the dental fricative [3] are allophones of the phoneme /d/, they appear in different phonetic con­texts (complementary distribution): the dental fricative is used only between vowels: [nafla] nothing, [predo] meadow, [роЭег] to be able to. In all other contexts the plosive /d/ is used: [dias] days, [andar] to go.

Thus we can conclude that two or more sounds are realizations of the same phoneme if they are in complementary distribution and if they are phonetically similar. Phonetic similarity is a relevant condition for the case of/h/ and /n/ which have what may be called complementary distribution because /h/ occurs only at the beginning of a word, in syllable onset; /л/ can occur only at the end of a word, in syllable coda, but because of their phonetic dissimilarity they cannot be allophones of one phoneme.

Two or more sounds are realizations of different phonemes if they are in parallel (overlapping) distribution and they serve to differentiate words, i.e. a semantic contrast is caused by substitution of one for the other.

Дата добавления: 2014-01-11