下载掌阅APP，畅读海量书库
立即打开
To cook a dish, we start by collecting the ingredients. There are so many materials available in the market, but what are necessary for my dish? If we compare the study of phonology to cooking, we will start by identifying the ingredients of spoken language—the sounds included in the sound system of the language under study.
Divide any English utterance into its smallest elements, you will have some sound segments. No matter how many utterances you try, the sound segments will not go beyond the 20 vowels and 24 consonants as listed in Table 3-1 (see 3.2.6). It is surprising, isn't it? On the one hand, human voice is capable of various sounds, then how does it come that only 44 of them are functional in English? On the other hand, though only 44 sounds are functioning, they can work out any meaning an English speaker needs to convey.
Man: Fish!
Lady: Dish! ...... Dine?
Man: Fine.
Look at the caricature. What colorful meanings are realized through the limited sound combinations in the dialogue between the man and the lady! Seeing a fish biting, the man shouts "fish" [fiʃ], and the lady responds "dish" [diʃ] to express that the fish reminds her of delicious dish; the lady goes on proposing to "dine" [dain] together, and the man thinks the idea is "fine" [fain].
In [fiʃ] / [diʃ] and [dain] / [fain], the two words in each pair are different in meaning, but quite similar in the sound patterns.
Words in the relationship such as "fish" to "dish" and "fine" to "dine" are called MINIMAL PAIRs. They are made up of similar sound sequence except for the difference of one sound in the corresponding position—[fiʃ] starts with a [f] sound and [diʃ] starts with a [d] sound.
In minimal pairs, we can always identify two "key" sound segments that semantically distinguish one word from the other. The two key sounds are therefore "contrastive" in meaning. They are in CONTRASTIVE DISTRIBUTION. The two sounds are functioning in language by distinguishing meaning, because the replacement of one with the other will result in meaning change. Since they have the function of expressing distinctive meanings in a language, we call them the distinctive PHONEMEs of the language. Usually, phonemes are enclosed in slashes. So the minimal pairs [fiʃ] / [diʃ] and [dain] / [fain] tell us that /f/ and /d/ are the distinctive phonemes of English.
Task 3-1: In this way, the minimal pair test helps us to discover the phonemes of a language. But look at the following pairs of words, do you count on them to identify the phonemes in English?
seed [siːd] / deem [diːm] soup [suːp] / seem [siːm]
wood [wud] / foot [fut] sky [skai] / skip [skip]
The above pairs of words are not minimal pairs because they differ from each other in more than one sound, such as
and
or they are not of the similar sequence, such as
(consonant ＋ consonant ＋ vowel) and
(consonant ＋ consonant ＋ vowel ＋ consonant). Remember that all minimal pairs must sound alike in the same place of the words, and their difference in pronunciation is only caused by one sound in the same place of the two words. Otherwise, the two words do not form a minimal pair.
As we have seen from the discussion in the last chapter, speech sounds differ from one another in a number of properties. For example, what distinguishes /p, t, k/ from /b, d, g/ is that the former set of sounds are voiceless and the latter set are voiced. We call the property of voicing a DISTINCTIVE FEATURE because it helps us to distinguish one phoneme from another, e. g. /p/ from /b/; or distinguish one set of sounds from the others, such as /p, t, k/ from /b, d, g/.
The distinctive features of a language work like a set of binary "switches" —each sound is a unique configuration of these "switches". For one sound, some of the switches may be turned on, while others are turned off. Which switches are on or off determines what kind of sound is produced. To represent these switches, every switch is given a name, such as [voiced], or [continuant], etc. To represent whether the switch is turned on or off, either a "＋" or "－" is placed before the name. For example, if the sound is voiced, the switch [voiced] is on, it is represented as [＋ voiced]. If it is voiceless, it is represented as [－ voiced].
Distinctive features include the phonetic features we discussed in the previous chapter. But we still need new features so that the distinction could be clear and complete. One such feature is the feature [continuant]. A [＋ continuant] sound is produced with continuous airflow through the oral cavity. Fricatives, such as /f, v, θ, ð, s, z, ʃ, ʒ/, and some approximants, such as /l, r, h/, have the feature of [＋ continuant]. But stops, such as /p, b, t, d, k, g/, and nasals, such as /m, n, g/ shall be described as [－ continuant]. Thus the distinctive feature of [continuant] divides speech sounds into two sets.
There are more than 20 distinctive features that we use to tell the difference between phonemes or classes of sounds in English. For example, the presence or absence of nasality can be shown as [＋ nasal] or [－ nasal], to be a labial sound or not can be [＋ labial] or [－ labial], etc. With the help of such features, we can explicitly define the difference between /p/, /b/ and /m/ in the following way:
So the three sounds share the feature [－ continuant] and [＋ labial], and what distinguishes /p/ from /b/ is [voiced], what distinguishes /p/ from /m/ is [voiced] and [nasal], what distinguishes /b/ and /m/ is the distinctive feature of [nasal].
Look at the following transcriptions. You will notice that /p/ is transcribed in three different ways.
pat [p h æt] pit [p h it] pot [p h ɔt]
spat [spæt] spit [spit] spot [spɔt]
tap [tæp ┐ ] tip [tip ┐ ] top [tɔp ┐ ]
The above transcription includes symbols to mark the minor features of sounds. These features may not be important in distinguishing meanings, but still they are noticeable. We call it NARROW TRANSCRIPTION of sounds, as opposed to the BROAD TRANSCRIPTION we find in the dictionary, which normally does not describe the minor features of sounds.
You can demonstrate how these sounds are different on your own. Place your hand in front of your mouth. Now say "pat" and "spat" alternatively. You feel more air against your hand in the production of "pat" than "spat". In phonetic terms, that extra puff of air in "pat" is known as ASPIRATION, and is represented in the transcription by the right superscript" h ". Moreover, when you say "tap", "tip" and "top", there is hardly any release of air for the /p/ sound. This is known as an UNASPIRATED sound, represented by the right superscript " ┐ " . So the sound that English speakers hear as /p/ can be pronounced in three different ways: REGULAR, ASPIRATED, and UNASPIRATED.
Look at the distribution of the different pronunciations across the above data. The distribution is not random. That is, there are rules by which we can predict where each minor feature can occur. Such rules can be shown as the following:
a. /p/ is pronounced as [p h ] when it occurs at the beginning of a word, such as [p h æt], [p h it] and [p h ɔt];
b. /p/ is pronounced as [p ┐ ] when it occurs at the end of the word, such as [tæp ┐ ], [tip ┐ ] and [tɔp ┐ ];
c. /p/ is pronounced as [p] everywhere else, such as [spæt], [spit] and [spɔt].
The above rules show:
(1) In English, /p/ is realized in three different ways.
(2) This difference in pronunciation is not random, but predictable, depending upon where the /p/ sound occurs in the word: beginning of the word, end of the word or any other position.
(3) To describe the regular variation of the phoneme /p/ in various phonetic contexts (as we do above), we are in fact formulating a PHONOLOGICAL RULE.
It is important to note that the places in which the different pronunciations occur are unique, and do not overlap. That is, you never find [p h ] at the end of the word, or [p ┐ ] at the beginning of the word, or [p] in either place. Since [p h ], [p ┐ ] and [p] never appear at the same place in words, they are no longer in contrastive distribution. They are said to be in COMPLEMENTARY DISTRIBUTION.
Sounds in complementary distributions do not appear in the same context. If we pronounce the /p/ in "pat" as [p], that will sound abnormal, but it does not change the meaning of the word. So the difference between them does not help with the distinction of meaning. They are not the distinctive phonemes of a language, but the ALLOPHONEs of a phoneme.
To write in English, we will have to use the 26 English letters. In actual writing, each of the 26 letters may take on different types, different colors or different sizes. So, in our mind, any English letter represents a range of varieties. An "a" letter can be "a", "a" or "A". It can be as large as a building or as small as an ant. Whatever appearance it takes, we recognize it as a variant of the English "a" and associate it with the "a" of "bat" in our mind. However, this range has a boundary. If we come across a symbol too much different from what we know about "a", e. g. "u" or "o", we will not regard it as the first English letter any more.
Similarly, each phoneme of a language represents a range of actual sounds. They are basically abstract in the mind of a language user, but can take on different forms when they are used. The actual realization of the phonemes of a language differs from speaker to speaker, or situation to situation. So a phoneme is a concept in our mind, and it can be realized by a range of sounds we hear or say. Such phonetic sounds (the PHONEs) that represent a single phoneme are the allophones of the phoneme.
In the case of the English stops, the sounds [p h ], [p ┐ ] and [p] are all predictable variants of /p/. They are in complementary distribution, but their difference in phonetic features won't result in different meanings.
Now are you confused by the relationship between phoneme and allophones? To make a better understanding, let's turn to the Superman analogy.
We know that Clark Kent and Superman are from a single person, even though they look and act differently. They usually turn up in different contexts. We may find Clark Kent when he is on the journey, in the supermarket or dating with Lois, while the Superman will only turn up in the context of emergency, such as when the world is to explode.
The fact that Clark Kent and Superman never appear in the same context means that they are in complementary distribution. Since they are in complementary distribution, if they were sounds, they would be allophones of the same phoneme.
Which is the phoneme, [Clark Kent] or [Superman]? Superman only appears in the context of being a hero; Clark Kent is the man living in different occasions. And in our mind there is a man who combines the two. Though the "mental" man only exists in our mental world, he has two identities in different contexts. So a phoneme is the man in our brains.
If we add Batman into our analogy, the explanation may turn out to be closer to our phonological case. Batman only appears in the context of being a hero, while Bruce Wayne appears everywhere else (i. e. they are in complementary distribution). Hence [Bruce Wayne] and [Batman] are allophones.
But /Batman/ and /Superman/ are not allophones to a single phoneme. They can appear at the same time in the same context while playing different roles. That is to say they are in contrastive distribution. Furthermore, they are physically different persons, each with his own distinctive identity. So if they are sounds, they stand for different phonemes.
Now, we may represent this complicated relationship in a more visible way. Let's use [X 1 ] to stand for Superman, [X 2 ] to stand for Clark Kent, and /X/ to stand for the combination of the two in our mind. Similarly, [Y 1 ] stands for Batman, [Y 2 ] for Bruce Wayne and /Y/ for the combination of the two in our mind. Then their relationship is shown in Fig. 3-1:
Fig. 3-1 The Superman analogy of phoneme and allophones
Now, we can regard [X 1 ] and [X 2 ] the allophones to a phoneme /X/, and [Y 1 ] and [Y 2 ] the allophones to a phoneme /Y/, what is the relationship among them now?
If we put all the distinctive phonemes of a language together, that will form the phonemic system of a language. So when we are talking about the 24 consonants and 20 vowels of English (see Table 3-1), we mean there are 44 distinctive phonemes of that language
. But that doesn't mean there are only 44 possible sounds in English, because in different contexts, the phoneme may be realized differently.
Table 3-1 The English phonemes
Different languages have different phonemic systems. Two distinctive phonemes in one language may be just allophones of one phoneme in another language. For example, in English, "bit" and "beat" are minimal pairs and we can conclude that /i/ and /iː/ are two phonemes in English. But Chinese only has one vowel (i) which may vary from /i/ to /iː/. That explains why, when we Chinese are learning English, we have difficulty in distinguishing "bit" from "beat", or "ship" from "sheep". The following is a dialogue between an American (Amy) who insists on the different meaning of /i/ and /iː/ and a Chinese student (Wang) who is not aware of the difference.
Amy: Hey! How was your trip? Did you fly or travel by train?
Wang: No, I came by /ʃiːp/.
Amy: /ʃiːp/? You must mean /ʃip/.
Wang: Yes, that's what I said—/ʃiːp/.
There are also examples that allophones of one phoneme in English are treated as distinctive phonemes in Chinese. [p h ] and [p] are allophones in English. We may feel it funny to hear somebody "[sp h iːk]", but we still understand him. But in Chinese, [p h ] and [p] are two distinctive phonemes. For example, words in each of the following pairs are semantically different, but the only difference of pronunciation between them is the aspiration feature between [p] and [p h ]:
胖［p h aːŋ］／棒［paːg］ 扑［p h u］／布［pu］ 皮［p h i］／比［pi］
Can you find more examples to prove that different languages may have different phonemic systems?