Standard English Pronunciation: The IPA Chart and Sounds Vietnamese Speakers Get Wrong

You've watched enough mouth-shape tutorial videos. You can read the IPA chart, you know which symbol maps to which sound on paper. But out in the real world, someone still asks "Sorry, what?" when you speak, and you don't know what you just got wrong, because your ear can't hear the difference between what you just said and what you should have said either.

That's the real problem for most people who self-teach English pronunciation: training the mouth before training the ear. You try to shape your mouth to match the illustration, but there's no way to check for yourself whether the sound that came out is correct, because your ear has never distinguished "ship" from "sheep" when hearing someone else say those two words. If you can't hear the difference, there's nothing to fix, even if you've memorized the whole IPA chart.

This article still has an IPA chart, the core part anyone learning pronunciation needs, but compact and organized rather than dumping all 44 symbols on you at once. The rest goes straight into the specific sounds Vietnamese speakers tend to get wrong, how mis-stressed words make listeners misunderstand the meaning, and why listening practice has to come before speaking practice if you want your pronunciation to actually improve, not just be correct in theory.

• The IPA chart, the part you actually need to know
• Sounds Vietnamese speakers commonly get wrong
• Stress matters more than you think
• You have to be able to hear it before you can say it
• Common mistakes when self-training pronunciation
• Frequently asked questions
• Further reading

The IPA chart, the part you actually need to know

English has 44 sounds: 20 vowels and 24 consonants. This number shows up on almost every pronunciation-teaching page, but few pages group it in a way that's easy to look up. The three tables below separate monophthongs, diphthongs, and consonants, so you can look up exactly the part you need instead of reading through one long, sprawling IPA chart from start to finish.

Monophthongs

Diphthongs

Consonants

24 consonants, grouped by voiceless/voiced pairs (vocal cords not vibrating vs. vibrating) to make them easier to compare.

Sounds Vietnamese speakers commonly get wrong

The sounds below aren't hard because you don't know the symbol. They're hard because Vietnamese doesn't have that sound, or has it but in a different position within the syllable, so the pronunciation reflexes you already have don't help at all.

Final consonants get swallowed

Vietnamese almost never ends a syllable with a clearly released consonant like /k/, /t/, /p/, /d/, and even less often ends with a cluster of several consonants in a row. So when speaking English, Vietnamese speakers tend to drop the final consonant entirely or turn it into something so soft it barely sounds at all. "Like" /laɪk/ ends up sounding like "lai", the final /k/ disappears. "Cats" /kæts/ loses the whole /ts/ cluster, leaving just "cat".

The longer the final consonant cluster, the more likely it is to disappear, because Vietnamese has no syllable that ends in three or four consonants in a row. "Strengths" packs a whole chain of consonants at the end, so does "sixths", and "asked" and "texts" add a final consonant onto a cluster that's already there. Native speakers say these words very fast and don't separate out each consonant clearly, but the whole cluster is still present in the airflow. If you're used to dropping the final consonant entirely, you'll hear these words as far more clipped than the original, and when you say them yourself they'll come out just as clipped.

/θ/ and /ð/ turn into /t/ and /d/

Vietnamese doesn't have these two sounds. The tongue has to sit between the upper and lower teeth, pushing air through the gap, rather than touching the gum ridge the way it does for /t/ or /d/. Since that reflex isn't already there, Vietnamese speakers automatically substitute the closest sound they do have: "think" /θɪŋk/ becomes "tink", "this" /ðɪs/ becomes "dis". Heard on its own, each word can still be guessed from context, but in fast natural speech, this mix-up makes native listeners pause a beat to figure out what was meant.

/s/ and /ʃ/, /z/ and /s/ are easy to mix up

/s/ and /ʃ/ differ in tongue position: for /ʃ/ the tongue pulls back further and the lips round slightly more than for /s/. /z/ differs from /s/ in whether the vocal cords vibrate, not in tongue position, so for many Vietnamese speakers this is a difference that's hard to feel with your hand or see in a mirror — you can only pick it up by ear, after hearing it enough.

Short and long vowels are easy to confuse

Vietnamese distinguishes vowels mainly through tone, not length. English is the opposite: the same tongue position but a different length produces two completely different words. Getting the length wrong doesn't just sound "a bit off" — it changes the word's meaning entirely.

Stress matters more than you think

Get a vowel wrong, and the listener can usually still guess the word from context. Get the stress wrong, and the word can end up sounding like a completely different word, even if you pronounced every individual sound correctly. Stress isn't a decorative detail — it's part of the word's structure, and a native speaker's ear relies on exactly that stress placement to recognize the word instantly.

The classic example is "record". Say REcord, with the stress on the first syllable, and it's a noun, meaning a recording or a file. Say reCORD, with the stress on the second syllable, and it's a verb, meaning to record or write something down. Say "I want to REcord this" instead of "I want to reCORD this", and a native listener can usually still guess from the sentence grammar, but with less familiar words, a stress mistake genuinely makes a sentence much harder to understand than getting a vowel wrong.

Stress doesn't just live inside a single word — it also lives across a whole sentence. Natural spoken English doesn't stress every word equally; it only stresses the words that carry the main information — nouns, main verbs, adjectives — while function words like articles, prepositions, and auxiliary verbs get reduced to a weak form, spoken very quickly and faintly.

"Can" and "can't" are the clearest example. These two words aren't mainly told apart by the final /t/ consonant, because in fast speech that /t/ is usually almost swallowed before the consonant that follows it. Native speakers tell the two words apart by the vowel: the weak form of "can" is /kən/, a short vowel that nearly disappears, while "can't" keeps a clear, longer vowel. If you hear the sentence "I can help you" and try to listen for a "t" sound at the end to tell it apart from "can't", you'll always be a step behind, because the distinguishing information is in the vowel, not the final consonant.

You have to be able to hear it before you can say it

Most pronunciation-teaching content focuses on the production side: where to place the tongue, how far to open the lips, how to push the air out. Very little content talks about the step that has to come before that: you have to hear the difference before you can check whether your mouth just produced it correctly or not.

The reason is simple. When you practice pronunciation alone, you hear the sound you just produced through the very ear that's already used to the Vietnamese sound system. If that ear has never distinguished the long /iː/ from the short /ɪ/ when hearing someone else speak, it won't distinguish them when you speak either, because it's the same listening ability, just applied to a different speaker. You might place your tongue exactly where the textbook says to, but if you can't hear the difference, you don't know whether you just got it right or wrong, and there's nothing to fix next time.

Here's a concrete example. A podcast says "I saw three sheep in the field" at natural speed. If your ear isn't yet used to telling /iː/ and /ɪ/ apart, you can easily hear it as "three ship in the field," write it down exactly like that, and then notice the sentence doesn't make sense because "ship" doesn't fit the grammar or the meaning here. The moment you catch that mistake, by comparing against the original, is when your ear actually learns the difference — not the moment you memorize the rule "sheep is long, ship is short" as a piece of theory.

This is why dictation practice helps pronunciation even though it isn't a pronunciation exercise in the strict sense. You listen to a sentence, transcribe it word for word, then compare it against the original. Wherever you transcribed something wrong is exactly where your ear couldn't yet tell the sounds apart, and that's information you simply don't get if you only practice speaking alone in front of a mirror. The podcast library on WELE has content at every difficulty level, so you can practice with real material, not short sample sentences trimmed specifically for pronunciation lessons. If you want to know where your ear currently stands before you start, you can try the free listening level test.

The hardest part isn't individual words on their own — it's when words link together in natural speech: the end of one word blends into the start of the next, stress falls on the words that carry meaning while function words get reduced almost to nothing. If you want to go deeper into this, connected speech is a topic worth reading about on its own. And if you want to understand the root reason Vietnamese speakers generally struggle to listen to English, there's a separate article on the three causes of poor listening among Vietnamese learners that goes deeper into that.

Common mistakes when self-training pronunciation

A few mistakes repeat across most people who teach themselves pronunciation alone, not because they're lazy, but because the way they're practicing is missing a step.

• Only watching mouth-shape videos, never recording yourself and listening back. Without listening back, you don't know whether the sound you produced actually matches the original — you only know whether your mouth looked like the illustration, and those two things don't always go together.
• Memorizing the IPA chart like an alphabet, detached from real sound. Symbols on paper don't produce sound by themselves. They're only useful when tied to listening to real words over and over, many times, at different speeds.
• Only practicing isolated words, never inside a sentence. Holding onto /θ/ correctly when reading the word "think" alone is completely different from holding onto that sound when it sits inside a fast sentence, pulled along by the words before and after it.
• Ignoring stress, focusing only on individual sounds. A word can have every individual sound correct but the wrong stress, and the listener still won't recognize it as a word they already know.
• Practicing pronunciation separately from practicing listening. This is the root mistake in most self-taught pronunciation approaches today, and it's also why many people who've practiced pronunciation for years still aren't confident in real conversation.

Further reading

Five more articles in the same series, each one going deep on a single area:

• Vocabulary: how many words you need and how to learn them so they stick
• Grammar: grammar points you know on paper but can't hear
• Speaking: sentences organized by function, and what to say when you can't keep up
• Self-study: why self-study usually fails and how to fix each mistake
• Starting from zero: starting from the right point instead of starting all over again

Frequently asked questions

Is it really necessary to learn the IPA chart?

Yes, but you don't need to memorize all 44 symbols at once. Knowing how to read IPA lets you look words up in the dictionary accurately, and know for certain that "record" has two different pronunciations with two different meanings instead of guessing from the spelling. But the IPA chart on its own doesn't teach you to hear the difference — that's the job of listening practice.

Should I learn British or American pronunciation?

Pick one accent and stick with it consistently, don't mix the two halfway through. Cambridge Dictionary and Oxford Learner's Dictionaries both let you listen to both accents for every word, so you can compare and pick whichever sounds more familiar against the content sources you listen to every day.

Is using a pronunciation-scoring app like ELSA enough to get pronunciation right?

Apps like ELSA measure the sound you produce and score it — that's the production side of practice. It can't replace listening practice, meaning the ability to tell two sounds apart yourself when someone else says them, which is the foundation for knowing what you need to fix before you can actually fix it.

How long does it take for pronunciation to visibly improve?

There's no single timeline for this, because it depends on which sounds you're getting wrong and how often you practice. Some sound pairs like /s/ and /ʃ/ can improve fairly quickly once your ear can tell them apart, while stress and sentence intonation need much longer listening exposure to become reflexive.

Do I have to fix every pronunciation mistake before I can communicate?

Not necessarily. Plenty of people speak fluent English with a strong regional accent and are still understood just fine. Problems only arise when a pronunciation mistake happens to land exactly on a sound or stress pattern that changes the meaning of a word — those are the spots worth prioritizing.

Can I teach myself pronunciation alone, or do I need a teacher?

Yes, as long as you have a way to check yourself. Recording your own voice and listening back against the original, or practicing dictation to find out exactly which sounds you're hearing wrong — both are ways to check yourself without a teacher sitting next to you.

Correct pronunciation doesn't come automatically from memorizing the IPA chart or watching enough mouth-shape videos. It comes from your ear gradually learning to tell apart sounds that used to sound identical to you. If you want to start there, try a short dictation exercise and see exactly where you transcribe something wrong.