34. Phonological rules for English plurals and more
There are three rules for forming regular plurals in English, exemplified by: book/books ([-s]), dog/dogs ([-z]), and bus/buses [-əz]). (Note that in General American English [-əz] is probably more common than the [-ɪz] you likely learned in English class. [-ɪz] is more common in British English than in American.) You probably learned these rules fairly early on in your study of English. These are phonological rules, because the phonological structure of a word determines which type of plural ending is added.
The same rules also apply to adding the /-s/ marker for possessives, e.g. Mark/Mark's, Bob/Bob's and Trish/Trish's, as well as for the third person singular form of verbs, as in take/takes, drag/drags, and push/pushes. While anticipatory assimilation is more common in English, as exemplified by the nasalization of vowels that occur before nasal consonants, the rules for English plurals, possessives, and third person singular verb forms offer examples of perseverative assimilation, in which a sound(s) is (are) influenced by sounds that occur before it. In this case, the factors involved are voicing and sibilant consonants.
For an introduction to and description of these rules in Chinese, please read this article in Cave's English Teaching (CET 師德) magazine, Hello! ET:
大師開講 — "-s" 和 "-ed" 詞尾 怎麼唸？ in No. 76, March/April 2013, p. 12-14 (pdf)
Here is a very succinct summary of the plurals rules by a teacher of English in the Philippines. Apparently Filipino students have the same tendency as Taiwan students do, that of pronouncing all final s's in spelling as [s]. (The rules are copied below in case the source page is removed, as often happens with the Internet.)
of nouns: Pronunciation.
In American English, the sound of "s" depends on which sound comes before it.
1. If the noun ends in an unvoiced consonant sound: /f/, /k/, /p/, /t/, /th/-(thin), pronounce "s" as /s/.
2. When it ends in a voiced consonant sound, /b/, /d/, /g/, /l/, /m/, /n/, /ng/, /r/ or with a vowel sound, /a/, /e/, /i/, /o/, /u/, pronounce "s" as /z/.
3. If it ends with /s/, /z/, /sh/, /ch/-chair, /zh/-the second "g" in garage, /dz/-(j), pronounce "s" or "-es" as /iz/.
In other words, if the noun ends with a sound other than the 5 unvoiced consonants, pronounce "s" with a /z/ (or with an /iz/ as the case may be).
It is not only assimilation that is in operation here. Its opposite, dissimilation, is also at work. In assimilation, two sounds become more like each other when they are in proximity. In dissimilation, just the opposite happens: some kind of phonological change occurs in order to make two sounds more distinct. In the case of words ending in a sibilant, an epenthetic ("extra") schwa [ə] is inserted in order to separate sounds that are phonetically close. What sibilants have in common is that they are all produced with a hissing sound. If similar hissing sounds are pronounced in succession, e.g. bus + s [bʌss], the plural ending cannot be clearly perceived by the listener. By assigning it a separate syllable in which it is preceded by the neutral vowel, we can hear the plural, possessive, or third person singular verb marking clearly. This is an example of dissimilation at work.
The pronunciation rules for regular verbs marked for past tense and past participle (we'll just say "past tense" here for simplicity) also involve both perseverative assimilation and dissimilation. Using spelling as a clue, we will assume the underlying plural marker to be /s/. In the case of the regular past tense verb marker, we will assume the underlying form to be /d/. As with the plurals rules, /-d/ is devoiced when preceded by a voiceless sound, and maintains its voicing when preceded by a voiced sound. But to obtain the dissimilation rule, we must examine other features of the past tense marker besides voicing. /d/ has no hissing sound, so the sibilant rule will not apply here. Sounds classified as sibilants are based on both manner of articulation they are all fricatives or affricates and place of articulation all are either alveolar or palato-alveolar sounds. If the same pattern applies to the past tense marker, what kinds of sounds can we expect will require a dissimilation rule? You may know the answer already, but try to derive it anew by yourself on the basis of manner and place of articulation. Please think about this carefully before reading on.
/d/ is a stop, produced at the alveolar ridge. What are the alveolar stops in English? /d/ and /t/. And indeed, if a verb ends in /d/ or /t/, we also add an epenthetic [ə] schwa before adding the final /d/ marking. But do we pronounce it [t] as in picked or [d] as in canned? According to the rule of perseverative assimilation which applies here, the final consonant should have the same voicing value as the preceding sound. That sound in this case is a schwa, which is a vowel, and vowels are voiced. So the past tense of regular verbs ending in [d] or [t] is [d], just as the plural marking after sibilant consonants is [z] rather than [s].
The rules are explaned in this video:
Here is another audio explanation:
Here is a British version from the BBC:
Contraction (page 32) and schwa elision (page 33) both involve phonetic reduction. Another important kind of phonetic reduction is the neutral vowel or schwa. We have thus far, including on this page, referred to schwas as a given, without discussing in detail what a schwa actually is, where the word schwa and concept of the schwa originally came from, and the role of the schwa in the rhythm of spoken English. We'll do this on the next page.
Next: Neutral Vowels
to next page back