Yeah, I know the rule that after a voiced consonant, the plural or 3rd person S should be realized as a Z. But I wouldn't trust that rule as far as I could throw it, because I constantly hear, say, 'problemsss' instead of 'problemzzz', or 'comesss' instead of 'comezzz'. What is the right pronunciation?
We do indeed need to pronounce plural, 3rd person and possessive S as described in the Original Posters question. In other words it is realised as an /s/ after unvoiced consonants, and as a /z/ in other situations (after vowels or after voiced consonants).
However! When a /z/ occurs before silence or before an unvoiced consonant (i.e. when it doesn't have a voiced sound afterwards), it will become partially or fully devoiced. In other words, the vocal folds (sometimes called vocal cords) will stop vibrating.
This means that a /z/ which occurs at the end of word, before silence—or a /z/ which occurs before an unvoiced consonant—will sound a lot like an [s]. So if you listen very carefully to the /z/ at the end of the word problems, it will have a very [s]-like quality. Notice though that this does NOT mean that the /z/ has become an /s/!!!
Why not? Well, there are other differences between /z/ and /s/ apart from voicing. Some of these are quite complicated (so, for example, less force is needed for the production of /z/. We see a very high intraoral pressure for /s/, which we don't see for /z/.) The most important difference is prefortis clipping.
Consonants which are usually unvoiced are called fortis consonants. When fortis consonants occur at the end of a syllable, they have a strange effect on the preceding vowels (and any following voiced consonants) in that syllable. So, for example, if you listen to the vowels in the words peas /pi:z/ and peace /pi:s/, you will be able to hear that the vowel in peace is only half as long as the vowel in peas. It is the length of a vowel in a syllable that tells a native English-speaker that the consonant at the end of the syllable is fortis (i.e. normally unvoiced), or lenis (normally voiced).
Native speakers do not hear the actual voicing in consonants when they occur at the ends of words. The preceding vowel tells them whether that sound is a 'voiced' or 'unvoiced' sound. In other words, when a listener hears the long vowel in the word peas their brain tells them that the sound at the end is a /z/. If they hear a short vowel, their brain tells them that the sound at the end is an /s/. In fact this effect is so strong that native speakers will tell you that they can hear the voicing in the /z/ even if we measure it with a machine and show that there is no voicing there. If we just play the sound at the end of the word speakers will confidently tell you that it is an /s/, when it is really a /z/! We can even record someone saying peas and cut out some of the vowel. When we play the same recording, but with some of the vowel missing, listeners will clearly hear the word peace instead of the word peas.
Advice for learners
For 3rd person, plural and possessive S, you need to use a /z/ when the original word ends in a voiced sound. It doesn't matter if the /z/ is fully voiced or not. Don't think about it! Everything will happen naturally if you just put a /z/ at the end of the word. However, you must not use an /s/. If you use an /s/ in the word peas for example, it will be heard as the word peace because you will cause the vowel in the word to be shortened. This will happen without you thinking about it (it happens across different languages). It's nice to have peace with your fish and chips, but peas are much tastier.
The production of phonemes in English can vary considerably between initial, medial and final positions. In general, the final phoneme of a final word tends to be less strong, and the voicing tends to end well before the phoneme has ended.
A native English listener allows for this and considers a final consonant to be voiced even if only the very beginning of the phoneme is voiced.
Here is a spectrograph showing the words docks and dogs.
The voicing is the lowest red line on the spectrum. Note that, in docks, there is no voicing visible in the final /s/, whereas in dogs, it is definitely present at the beginning but fades out gradually and disappears about half way through. When played in isolation, the final /z/ sounds half way between a /s/ and a /z/.
Note also that voicing is not the only discriminator between /s/ and /z/: the /s/ is slightly longer in duration than the /z/.
It is quite possible that you might perceive dogs as ending with a /s/, but to an English listener, the two words are very different and the final phoneme in dogs is definitely a /z/.
This is what problems and violence look like: I chose the latter because it has a nasal consonant before the /s/.
Note that the voicing is only present in the first (transitional) part of the /z/ in problems, and is completely absent in the /s/ of violence. Note also that the /s/ is considerably longer than the /z/. if your are not accustomed to listening to English, you might miss the release of the /m/ which is the boundary between the two phonemes, and mis-hear the final, sibilant part of the /z/ as a /s/.
Part of the process of learning a language is re-programming the ears to recognize each of the variants of each of the phonemes. When you are speaking the language, you should definitely stick to the rule that a final s is pronunced /z/ after a vowel or a voiced consonant.
In many dialects of English, an epenthetic consonant is added between a nasal consonant /m/ or /n/ and a following /s/ (as long as they're in the same syllable).
That means that prince sounds exactly like prints and Amsterdam sounds like Ampsterdam.
So if somebody pronounces sins more like /sɪns/ than /sɪnz/, there's no ambiguity with since (pronounced /sɪnts/). I believe the pronunciation has tended to drift towards /sɪns/ in some dialects, although I suspect it's often somewhere between an /s/ and a /z/. Native speakers hear /sɪnz/ because there's no /t/ in it.
You can't pronounce the plural 's' like /s/ after an /r/ or an /l/, because then you couldn't tell the difference between cores and coarse, or falls and false.
So my answer is
- if people pronounce a plural 's' following an /m/ or /n/ closer to an /s/ than a /z/, they will still be understood.
- Not all native speakers do this, and if you're learning English you should just stick with pronouncing a /z/ after /m/ and /n/ in words like problems, comes, and sins.
As a general rule, you turn the voiceless 's' into a voiced 'z' when it follows another voiced letter. (This is known in linguistics as "progressive assimilation" -- the voicing of the first letter continues on into the second.)
- cats = 't' is an unvoiced consonant, so the 's' remains unvoiced
- dogs = 'g' is a voiced consonant, so the 's' turns into a voiced 'z'
- problems, comes = 'm' is a voiced consonant, so the 's' turns into a voiced 'z'
In case it's unclear, voiced means your vocal cords vibrate when you pronounce that letter (phoneme), whereas as voiceless letters are created with only a release of air. Voiced letters are (usually - there are always exceptions and special cases) b, d, th, v, l, r, z, j and all vowels.
Firstly, you might have to specify if you have troubles with the letter S at the end of words only or also in other positions, e. g. in the middle.
If the problem is about how to pronounce S at the end of the word, then there's no big deal. At the end of a word S can denote either a plural ending or a possessive case of a noun, or it can be the ending of 3rd person singular of Present Simple of a verb. It does not matter very much, because the rules for both noun and verb S-endings are the same. But I want to pay attention at the fact that it mostly depends on the preceding sound and not so much the letter.
So, here are the rules:
S-endings are pronounced as [s] after a voiceless consonant sound. These are: [p], [f], [t], [θ], [k]. Examples: caps, hoofs, hats, paths, rocks.
S-ending is pronounced as [z] in any of two essential cases:
- 1) after a voiced consonant sound. These are: [b], [m], [v], [d], [n], [l], [ɹ], [ð], [g], [ŋ]. Examples: tubs, rooms, sleeves, beds, guns, rails, cars (AmE), mouths, logs, songs.
- 2) after any vowel sound, including all monophthongs and all diphthongs. Examples: trees, stars (BrE), cores (BrE), days, cows etc.
If you have ES-ending, which occurs only after sibilant consonant sounds ([s], [z], [ʃ], [ʒ], [tʃ] and [dʒ]), then you should know that this two-letter ending is pronounced always as [ɪz]. Examples: cases, roses, wishes, peaches, pages etc.
Note also that the final S may be not an ending at all. It can be a part of the root. Examples: bus, walrus etc. In these cases it is usually [s].
For the middle of the word, the general rule is that letter S denotes sound [s] between two vowel letters (vowel letter are: A, E, I, O, U, and Y). Examples include: nose, these, raise, weasel, position, cosy etc. But be aware because there is a bunch of exceptions. E. g. base and research, where we have [s]-sound and not [z].
However, if a middle-word S is not situated between two vowel letters, then, of course, in absolute majority of cases it denotes voiceless [s]-sound. Examples: purse, false (not after a vowel); nasty, biscuit (not before a vowel); holster (between two consonants). Notable exception are S's before voiced plosives ([b], [d] and [g]), where S can denote [z]-sound. For example, we pronounce S as [z] in husband, wisdom or disgust. Also, S can be [z] in the suffix -ism of an abstract noun: communism, capitalism and so on.
Finally, an S at the start of the word almost always is pronounced [s], disregarding of what follows it. E. g. sun, see, sky, slow etc. I think in English we never pronounce initial S as Z, because letter Z serves for that purpose: zeal, zombie, zest etc. One additional note: in a few words (very few), initial S is pronounced as [ʃ] (sh-sound) before a vowel letter. Examples of such words are sugar and sure.