I recently found some words but I got confused and don't know whether to say sd or st. I am sure the following are pronounced as st:

  • sister
  • caster
  • ancestor
  • master

But the following are pronounced as sd:

  • mysterious
  • destiny
  • stable
  • staple

Is there a rule for changing the pronunciation of t to d following an s?

  • 3
    This varies between speakers, and is not phonemic. Much of the perceived difference between phonemic /t/ and /d/ in English derived more from losing the aspiration of the stop than from an actual voicing change. You may be miscued here. There should still be some aspiration in the stressed [tʰ] onset of mysterious, though. – tchrist Oct 22 '14 at 2:28
  • 2
    Your premise is faulty. For instance, I speak British English in the accent I grew up using (namely Received Pronunciation), and don't pronounce any of your example words with sd rather than st. This is just one example which shows that the way English words are pronounced varies greatly across the English-speaking world, so you cannot make a generalization like yours that is valid for more than certain subsets of speakers (which you would also have to identify and describe). – Erik Kowal Oct 22 '14 at 4:38
  • Non-scientific, I know, but if I try to pronounce your second set of words (or the first, for that matter) with sd rather than st, I sound as if I have quite a severe cold. – oerkelens Oct 22 '14 at 7:03
  • @ErikKowal This is not one of those areas where you can just reflect on what you do, unlike fore example with grammar. The reason is that your perception of what you hear and do is shaped by factors that is difficult to consciously override. The only way to test this these things is to actually record yourself and then look at the data through a spectrogram or waveform readout. The problem OP has is that their perception of what constitutes a /t/ or /d/ will be slightly different. We as native speakers aren't aware of the differences that /t/, for example, has in different environments. – Araucaria - Not here any more. Mar 7 '15 at 14:15
  • @Araucaria - You've entirely missed the point of my comment. My emphasis was not on how I pronounce these sounds, but on the great variability in how English speakers of all backgrounds say them. FYI, I am in no doubt about the fact that I don't say sd rather than st. It is not necessary for me to resort to recording myself on a spectrogram in order to be certain of this. – Erik Kowal Mar 8 '15 at 2:19

I think what you're hearing is not the difference between [st] and [sd], but whether the 't' is aspirated [stʰ] or not [st]. In Chinese, these are two different phonemes, and Pinyin represents [tʰ] by 't' and [t] by 'd'. These sounds are two different varieties (allophones) of /t/ in English; the difference between /d/ and /t/ is whether it is vocalized or not.

The rule in English (judging from your examples) is that if the syllable starts with 'st', the 't' is not aspirated; if the syllable starts with 't' and the previous one ends with 's', then the 't' is aspirated. Where you break the 'st' between syllables depends on a lot of factors, including which syllable is stressed. If the 'st' occurs at the start of a stressed syllable, it won't be aspirated (unless there's a morpheme boundary, as in mistook).

I would also suspect that it depends on the speaker in some words (I would guess destiny is one of these).

  • My native language is actually Chinese. After reading your answer and comments. I'm sure I understood the "aspirated /t/". I just didn't know there is a particular term for it. However, I am starting to get confuse of "the difference between /d/ and /t/ is whether it is vocalized or not". Can you give an example demonstrating /d/ and /tʰ/? I had "Dan" and "Stan" and they sound the same to me. – Daniel Cheung Oct 23 '14 at 14:03
  • "Dan" and "Stan" is a good example. If they sound the same to you, then you need to train your ears better (although you can probably get by without learning this distinction, because the places where you use [t] rather than [tʰ] are usually positions where it's unlikely to find a [d]). – Peter Shor Oct 23 '14 at 14:38
  • I want to extend my question, if you pronounce "sister" with a /t/, should you pronounced "mister" with a /t/ as well? Because they have very similar structure: xxx+ster. Does it also apply to pronouncing "plastic" and "dance", pronouncing "a" as "ar" or "ae"? Choose one or another but not some "ar" and the other "ae"? – Daniel Cheung Oct 23 '14 at 14:46
  • 1
    @PeterShor Nice answer. However, if, when you say that the difference between /d/ and /t/ is whether it's vocalised or not, you mean the difference is whether there's vocal fold vibration - I think that's how it currently reads -that would be incorrect. In any situation where /d/ is adjacent to a voiceless sound including silence, it will be partially or fully devoiced. So there will be little or no voicing in the /d/ in misdemeanour , for example, because the /d/ follows the voiceless segment, /s/. – Araucaria - Not here any more. Mar 7 '15 at 14:01
  • 1
    @PeterShor I'm not sure there'll be any difference in the length of the vowel there because the /d, t/ there would presumaby be the onset of the following syllable? (in fact must be, I assume, in the case of misdemeanour). – Araucaria - Not here any more. Jan 25 '19 at 0:34

[d] and [t] are allophones of the phoneme /t/. Notionally, all the words are "supposed" to be pronounced with /st/. However, no matter how hard you try, it will come out sounding a bit like /sd/ in a word like staple. What's the difference between master and staple? The syllable division in master lets you enunciate the /t/ sound.

The simple rule is, try to pronounce /st/. Whatever sound happens to come out will be "correct".

I would classify the pronunciation of mysterious and destiny as closer to /st/, though.

  • Given that these are not in complementary distribution (the choice is habit or personal preference; both utterances are correct; use is unpredictable, and do not change the meaning of the words, they are free variants rather than allophones. An allophone is something like the [p] in pill or spill, which is predictably aspirated in pill but not in spill because use is governed by its position - word-initial (or syllable-initial) is aspirated, which means that unaspirated is not used there, so they are in complementary distribution. – Roaring Fish Oct 22 '14 at 4:10
  • How are they in complementary distribution? In all the examples given by the OP my pronunciation is /t/, and I can't think of region of Britain where they would be realised consistently as /d/, while in the USA they likeley would be and both are considered correct. That shows that it is not complementary distribution - both are used in the same context and both are correct. – Roaring Fish Oct 22 '14 at 4:47
  • @Roaring Fish: I don't believe the OP is hearing /t/ and /d/, he's hearing aspirated /t/ and non-aspirated /t/, which do have complementary distribution, and a non-aspirated /t/ sounds like a /d/ to him because that's the distinction between /t/ and /d/ in his native language. We wouldn't pronounce any of these as /d/ in the USA, either. It's /t/s which fall between two vowels, or after an /r/ and before a vowel, like butter and barter, which are pronounced as /d/s. – Peter Shor Oct 22 '14 at 19:43
  • A /t/, aspirated or not, is still a /t/ and not a /d/. Voicing turns a /t/ into a /d/ (which is why the rotic /r/ leads to a /d/ as in your barter, for example), but aspiration doesn't. As an /s/ is also unvoiced, that cannot be the 'cause' so it is stll not complementary distribution. The abscence, in some dialects, of /d/ in the examples given proves it. – Roaring Fish Oct 23 '14 at 4:42
  • I agree on your classification on the 2 words as I am reading them out loud. Would it be correct on judging each pronunciation as American English or British English or other? – Daniel Cheung Oct 23 '14 at 14:12

There is no difference in the 'st' pronunciation in both sets of words there. I think the reason you are confused is because the sound is represented by the letter 't', but is actually pronounced closer to a 'd' sound. This a consistent pattern of pronunciation in 'st' clusters in English. You may also notice this pattern with 'sk' and 'sp' -- the 'k' sounds more like a 'g' and the 'p' sounds more like a 'b'.

  • 2
    This answer is wrong. I am from north Britain, and there st is always pronounced as /st/. Saying they are consistently pronounced as /sd/ is simply not true. – Roaring Fish Oct 22 '14 at 4:16
  • 1
    Well, I cannot say that I'm knowledgeable about all aspects of every single English dialect, as I expect no one is. If you would like more clarification, I can say that I'm talking about the technicality of how it's pronounced in most dialects of English, especially American. Phonetically, acoustically, a /t/ that follows an /s/ is unaspirated, which sounds and is much closer to a /d/ sound. Notice that I did not say that it is a [d] sound. – Nadya Oct 23 '14 at 2:44
  • I am not sure if I can agree to your answer :/ as I really can hear both the pronunciation from different words. – Daniel Cheung Oct 23 '14 at 14:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.