It's not so much a case of 'mixing', Helen, but of following the phonological rules of one's own language, often without realising it. For example, many Germans speak excellent English, but even those who are near-native sometimes pronounce 't' instead of 'd', and 's' instead of 'z', at the ends of words. This is because the German language tends to end words with unvoiced rather than voiced consonants. Many Spanish speakers don't actually have a /z/ phoneme in their native pronunciation, so will pronounce all /z/ sounds as /s/ - for example saying 'iss' instead of 'iz' for the word 'is'.

And yes, native speakers do notice when these words are pronounced 'wrong', but in most cases it doesn't matter. It's very rare for these variations to hinder comprehension. The human ear/brain connection has an amazing capacity to filter and interpret input. For example, it's strange at first to hear an Arabic speaker pronounce all English /p/ sounds as /b/, but the listener soon adapts to this and interprets correctly what the speaker is saying.

As for native speakers, it's not that we don't think about the rules, it's more that most of us don't even know that these rules exist. And not only that - the vast majority of native speakers are actually completely <em>unaware</em> of the sounds that they are making. Most don't actually realise that the final consonant of 'passed' is different from 'raised', for example. When asked, the average English speaker would probably swear to you that all 'ed' past tenses are pronounced the same. They'll say 'It's got a 'd' sound at the end, so we say a 'd'' sound.' This is because we learn about spelling, but not about pronunciation. Spelling is taught, pronunciation is acquired instinctively.

They don't. But they can tell when we mispronunce a word.

Yes, Phil's made a good point about the length of the vowel before a voiced or unvoiced consonant. For example, if you compare the words 'write' and 'ride', the vowel sounds are not exactly the same. The /ai/ sound in 'ride' is fractionally longer, because of the voiced vowel that follows.

@Su.Ki.

Thank you, Su.Ki.! You're always patient and wise! I think teaching makes me pay too much attention to some trivialities. Then I forget learning language is to communictae with others.

@Phil

You're right, Phil! When I'm thinking about how I should pronounce the end of the word, I get it worse..But, I don't really understand this 'an unvoiced consonant at the end of a syllable will shorten the preceding vowel.'. What do you mean by that? So pronouncing /t/ is shorter than /d/?

And for your Chinese questions. 姐姐 should be read as jiě jie. Because when we pronounce the two same words, the second word does not have a tone actually. For example, 哥哥 as gē ge instead of gē gē. Hope this helps. :)

The theoretical pronunciation of “-ed” is /d/, and you should think of it as such.

Verbs ending in the D or T sound: Since an English word isn’t aloud to end in a double consonant (I’m referring to a consonant pronounced as double — spelling is another matter), a vowel (schwa or barred-i) is inserted to separate them. The same goes for verbs ending in T, since a T is just an invoiced D. Make sure you follow this “rule,” or native speakers will not realize you’re using the past tense.

For the rest of the regular verbs, native speakers don’t really care if you get it “wrong.” In any event, if you focus on pronouncing the final sound of the root correctly, you can simply imagine making a /d/ and it will automatically sound like a /t/ where appropriate — this is especially true if you’re a Chinese speaker. If this doesn’t work, you really need to work on your pronunciation of syllable-final voiced and invoiced final consonants. Here’s a tip: Other things being equal, an unvoiced consonant at the end of a syllable will shorten the preceding vowel. Native speakers are not consciously aware of this, but it’s at least as important as getting the consonant right. Pay attention when you’re listening, and trust your ears.