English Grammar: Memorized on Paper, Lost When Natives Speak

On this page
- The grammar you need to know, condensed
- Grammar you know but can't hear
- Three grammar mistakes Vietnamese speakers commonly make because Vietnamese doesn't have them
- Tense lives on the verb, not on the time word
- Plural lives on the noun, not just in your head
- Articles don't exist in Vietnamese
- How to train grammar into a reflex
- Further reading
- Quick lookup for the past tense of each verb
- Frequently asked questions
You've studied English grammar for years at school, done exercises conjugating verbs, distinguishing present perfect from simple past, and your grammar test scores were never bad. But on a work call, or watching a video without subtitles, you hear a sentence and you're not sure what tense it was just said in, singular or plural, whether there was an article or not. You know the rule, you just can't hear it being used.
That's not because you've forgotten your grammar. Most of the grammar that carries meaning in spoken English lives in the smallest sounds in a sentence: a final consonant -s, a -ed swallowed into the next word, an article reduced down to a single schwa. Grammar books teach you to write those parts correctly, but none of them teach you to hear them when a native speaker talks at normal speed.
This article won't list out the twelve tenses from A to Z again, plenty of other pages have already done that. It focuses on the part almost no one talks about: grammar you know on paper sounds completely different when it rings out in real speech, three grammar mistakes Vietnamese learners commonly make because Vietnamese has no corresponding structure, and how to train so that grammar becomes part of your listening reflex.
- The grammar you need to know, condensed
- Grammar you know but can't hear
- Three grammar mistakes Vietnamese learners commonly make because Vietnamese has no equivalent
- How to train grammar into a reflex
- Frequently asked questions
- Further reading
- Quick lookup: the past form of each verb
The grammar you need to know, condensed
Before getting into the main part, here's the framework section: the English tenses that actually show up in everyday speech, condensed into a single table instead of spread across all twelve tenses. If you've already studied these tenses, skim through to check yourself, the important part of the article is in the next table.
| Tense | Structure | When to use it | What to listen for in speech |
|---|---|---|---|
| Simple present | S + V(s/es) | Habits, facts, fixed schedules | the -s/-es at the end of the verb is a light consonant, nearly gone when spoken fast |
| Present continuous | S + am/is/are + V-ing | Happening right at the moment of speaking, a plan already arranged for the near future | am/is/are contract to 'm/'s/'re, all the stress falls on -ing |
| Present perfect | S + have/has + V3 | Something just finished, an experience, something that started in the past and still affects the present | have/has contract to 've/'s, sometimes reduced to just a short breath before the verb |
| Simple past | S + V-ed / V2 | Something already finished, with a clear time marker in the past | -ed is usually swallowed into the final consonant, except for the group that adds the /ɪd/ sound |
| Past continuous | S + was/were + V-ing | An action in progress at a point in the past, usually setting the scene for another action that interrupts it | was/were contract lightly, sounding like a filler sound before -ing |
| Future with going to | S + am/is/are + going to + V | An intention already planned in advance | going to is almost always pronounced gonna |
| Future with will | S + will + V | A decision made right at the moment of speaking, a prediction, a promise | will contracts to 'll, sticking right onto the subject before it |
| Second and third conditional | If + simple past, S + would + V / If + past perfect, S + would have + V3 | A hypothesis contrary to fact in the present or the past | would contracts to 'd, would have contracts to 'd've, both easily confused with had |
Some tenses you studied carefully in books but rarely need when listening to everyday conversation. The past perfect, had + V3, only really stands out when someone is telling a story with a specific before-and-after order of events. The future perfect, will have + V3, is almost absent from daily communication. Don't spend as much time practicing listening for these two tenses as you do for the eight lines above.
Grammar you know but can't hear
This is the reason this article exists. Spoken English doesn't stress every sound in a sentence evenly. Nouns, main verbs, adjectives are said clearly and carry stress. The grammar part, meaning whatever connects those words correctly in tense and number, falls into the group of unstressed sounds, spoken quickly, spoken lightly, and often nearly disappears.
The result is you can know the third-person singular verb conjugation rule by heart, but when you hear "He work here" and "He works here" at normal speed, the two sentences sound almost identical. The final -s consonant is the shortest, lightest sound in the whole sentence, and a Vietnamese ear, used to a language that doesn't change verb form according to the subject, finds it even harder to catch.
The table below lists grammar points you've almost certainly studied, how they're written, how they actually sound in speech, and the specific reason they tend to slip past your ear.

| Grammar point | How it's written | How it actually sounds when spoken | Why it's easy to miss |
|---|---|---|---|
| Third person singular (-s) | He works here. | Sounds almost like "He work here" — the -s is just a very short /s/ or /z/ consonant, carrying no stress. | It's the last sound in the sentence and also the weakest one, the easiest to drop while you're busy trying to catch the meaning of the whole sentence. |
| Past tense -ed, /t/ group | asked, walked, finished | Sounds like "askt", "walkt", "finisht" — the /t/ sticks tightly to the final consonant of the base verb. | -ed in this group doesn't add an extra syllable, it just adds a /t/ consonant that gets swallowed right into the consonant cluster before it. |
| Past tense -ed, /d/ group | played, called, loved | Sounds like "playd", "calld", "lovd" — the /d/ blends into a final sound that's already voiced. | The only difference between the present and past forms comes down to one tiny /d/ sound at the end, with no accompanying vowel for you to catch. |
| Past tense -ed, /ɪd/ group | wanted, needed, decided | Clearly heard as a separate syllable, like "want-ɪd", "need-ɪd". | This is the only group you can actually hear, because -ed adds a whole extra syllable and isn't swallowed the way it is in the two groups above. |
| Plural nouns -s/-es | three books, two matches | Sounds like "three book", "two match" if you're not used to catching final consonants. | Same mechanism as the third-person -s — a light consonant at the end of the word, usually run together with the next word. |
| Articles a/an | a friend, an idea | Reduced down to just the sound /ə/, almost fused to the word before or after it. | Articles carry no lexical meaning, so native speakers have no reason to stress them, and you don't have the reflex to catch a sound that carries no meaning either. |
| Article the | the book, the idea | Reduced to /ðə/, changing to /ði/ before a vowel — both are extremely short. | Like the article a, the is just a grammatical signal, not a meaning-bearing word, so it gets glossed over in speech. |
| have/has in the present perfect | I've seen it. She's already left. | 've and 's are reduced to a single consonant stuck onto the pronoun, sometimes down to just a short puff of breath. | This is the entire marker of the perfect tense in the sentence, and it gets compressed into a sound smaller than a single syllable. |
| 's carrying three different meanings | He's tired. He's finished. Anna's car. | All three sound the same — a short /z/ sound after a name or pronoun. | 's can be is, has, or the possessive; you have to rely on the word that follows to guess the meaning, not on the sound itself. |
| Contracted be ('m/'re/'s) | I'm going. They're leaving. It's done. | Almost completely fused to the preceding pronoun, sounding like part of the pronoun itself. | In the continuous and passive tenses, be carries important grammatical information but is the weakest sound in the verb phrase. |
| be going to | I'm going to check it. | Sounds like "I'm gonna check it". | going to is almost always linked and shortened in speech, even in fairly formal speech. |
| have to / has to | You have to submit it. | Sounds like "You hafta submit it". | The to after have changes how it's pronounced, and have to gets run together into a chunk that doesn't sound like what you learned on paper. |
| want to | I want to go. | Sounds like "I wanna go". | want to is almost never pronounced as separate words in everyday conversation. |
| would in second conditionals | If I had more time, I would help. | would contracts to 'd, sounding like "I'd help". | 'd is also the contracted form of had, so you have to hear the whole sentence to know whether the 'd here is would or had. |
| would have in third conditionals | If I had known, I would have called. | Sounds like "I'd've called", sometimes reduced all the way down to "I'da called". | This is two layers of contraction stacked on top of each other — would becomes 'd and have becomes 've — leaving almost no sound left to hold on to. |
| modal + have (should/could/must have) | You should have told me. | Sounds like "You shoulda told me". | have after a modal loses the /h/ consonant and the vowel entirely, leaving just /ə/ stuck onto the modal before it. |
| do/did in questions | Do you like it? Did you see it? | Sounds like "D'you like it?", "Didja see it?". | do and did carry no meaning of their own, so they get run together with the you that follows — to the point that many beginners mistake it for a statement. |
| will | I will call you later. | Sounds like "I'll call you later". | will contracts down to /l/, a consonant that attaches directly onto the pronoun, hard to pick out by ear if you're not used to hearing it. |
Looking back at the table above, there's one thing in common: everything that gets swallowed is a grammatical element, while the core content words are almost always kept intact and clearly stressed. Native speakers don't need to hear -s or 've clearly, because they automatically fill in the gap based on context and the subject already mentioned earlier. You don't have that reflex yet, so you need to hear the actual sound, not just guess the meaning.
Three grammar mistakes Vietnamese speakers commonly make because Vietnamese doesn't have them
The three mistakes below aren't because you haven't studied carefully enough. They recur in most Vietnamese learners of English because Vietnamese handles these three concepts in a completely different way.
Tense lives on the verb, not on the time word
In Vietnamese, tense is expressed with separate words: đã, đang, sẽ, từng. The verb itself doesn't change form — "Hôm qua tôi đi làm" and "Ngày mai tôi đi làm" use the same word "đi". English works the opposite way: tense sits right on the verb, through -ed, through V3, through the auxiliary have/has/had. Because your brain is used to treating a time word as enough to signal tense, you easily say "Yesterday I go to work" instead of "went", since "yesterday" has already done that job. The same happens when listening: if a sentence has no clear time marker, you easily miss the verb ending altogether, because you're not used to relying on it.
Plural lives on the noun, not just in your head
Vietnamese doesn't require marking plurals. "Ba quyển sách" is already clear enough thanks to the word "ba" — the word "sách" itself doesn't change whether you mean one copy or ten. English requires marking the plural right on the noun with -s/-es, even when a number already precedes it: "three books", not "three book". Because Vietnamese has no equivalent structure, you often forget the -s when writing and fail to hear the -s when listening, because to native-language intuition, it's a sentence-final sound that's both light and redundant.
Articles don't exist in Vietnamese
Vietnamese has no articles. "Tôi mua sách" doesn't need to distinguish between buying one specific book or books in general. Learning English, you have to add an entirely new layer of distinction: definite or indefinite, countable or uncountable, singular or plural. Because this layer doesn't exist in Vietnamese, you struggle both to pick the right article when writing and to catch it when someone else is speaking, because your ear isn't used to listening for a sound that carries almost no meaning within a sentence.
All three mistakes are a natural consequence of moving between languages, not a sign that you're a poor learner. Knowing why you make these mistakes helps you correct them with direction, instead of re-memorizing the rules without understanding why you keep forgetting them.
How to train grammar into a reflex
Reading back through the list of rules above doesn't solve the problem, because you've already read these rules many times before. The problem isn't knowing the rule — it's that your ear has never had to process the actual sound of that rule at natural speaking speed.
The way to turn a grammar rule into a listening reflex is to force yourself to hear it correctly and then write it down — in other words, listening dictation. When you listen to a sentence and transcribe it word for word, you can't skim past the grammar the way you do when reading for general meaning — you're forced to decide whether this sentence has an -s or not, a 've or not, an article here or not. When you transcribe something wrong, you listen again, compare it against the original, and see exactly where you missed it.
For example, say you hear a sentence in a podcast: "She's already finished the report." At natural speaking speed, this sentence can easily come out sounding like "She already finish the report", because the 's contraction almost disappears and the -ed of finished sticks tightly to the consonant right after it. If you're only listening for the general idea, you still catch the content: she has already finished the report. But transcribing it word for word forces you to listen to that segment again, notice a very faint /z/ sound before "already" and a /t/ sound stuck onto the end of "finish", and only then fill in the 's and -ed correctly. Next time you hear a similar sentence, your ear will already be used to listening for those sounds.

The WELE podcast library comes with transcripts so you can check your work right after transcribing, instead of guessing or skipping over the parts you misheard. If you want to understand more deeply why natural speech swallows so many sounds, the article connected speech, the secret to understanding native speakers explains the mechanics of linking and dropped sounds in more detail than this section. If you've never done dictation before, the article 5 steps to effective listening dictation walks you through the specific process starting from the very first sentence.
If you're not sure where your current listening level stands, the free listening level test gives you a concrete starting point instead of a guess. And if the three transcription errors above made you curious about the deeper reasons Vietnamese learners tend to have weak listening skills, the article why Vietnamese learners have weak listening skills, three causes and how to fix them goes further into this topic.
Further reading
Five more articles in the same series, each one digging into a different area:
- Vocabulary: how many words you need and how to learn them so they stick
- Pronunciation: the IPA chart and the sounds Vietnamese speakers usually get wrong
- Communication: sentences organized by function, and what to say when you can't keep up
- Self-study: why self-study often fails and how to fix each mistake
- Starting from zero: starting in the right place instead of starting over from scratch
Quick lookup for the past tense of each verb
Each verb below has its own article: a table of forms, how the two past forms sound in real speech, common phrases, and exercises. They're grouped by pattern of change so you can see which verbs share the same mold.
Irregular verbs and auxiliaries: be, go, do, have.
The -ought group: buy, bring, think, catch, teach.
No change in form, or only a change in pronunciation: put, read.
-nd ending becomes -nt: build, send, spend.
One form for both V2 and V3: meet, hold, find, stand, sell, tell, hear, lose, make, feel, leave, get.
Vowel change, V3 takes -en or changes further: take, see, eat, fall, choose, drive, ride, fly, wear, break, rise, drink.
Regular verbs that still get misheard: help, talk.
Frequently asked questions
How many tenses does English grammar have?
Textbooks usually list twelve tenses. In everyday speech, most conversation revolves around only about eight: simple present, present continuous, present perfect, simple past, past continuous, the two future forms, and would in conditional sentences. The rest rarely show up outside formal writing.
How long does it take to get good at English grammar?
There's no universal timeline, because your speed depends on how you practice, not just on how long you've been studying. Grasping a structure on paper is usually much faster than being able to hear it and use it correctly on reflex.
Do you need to learn all of grammar before you can speak English?
No. Many people communicate well with incomplete grammar, as long as they understand what they hear and express themselves clearly enough for the listener to get the point. Grammar makes you more precise, but it isn't a prerequisite for starting to speak.
Why do people who know grammar well still fail to understand what they hear?
Because most of the grammar that carries meaning in natural speech sits in unstressed sounds, like -s, -ed, articles, and contracted auxiliaries. Books teach you to write those parts correctly, but that doesn't automatically teach you to hear them at real speaking speed.
Should you learn grammar from books or from listening?
You need both, and neither can replace the other. Books tell you that a rule exists and how it works. Listening and then transcribing gives you the actual chance to catch that rule in the sound, which is the point where most learners stop at reading without ever practicing further.
What's the hardest part of English grammar for Vietnamese speakers?
The three hardest points are usually tenses and their accompanying auxiliaries, because Vietnamese verbs don't change form, the articles a/an/the, because Vietnamese has no equivalent concept, and fixed prepositions that go with particular verbs or adjectives, because English preposition choice doesn't follow direct-translation logic.
English grammar isn't a list you learn once and then you're done. The hardest part isn't memorizing the rules, it's training your ear to catch that rule when it's compressed into one small sound in the middle of natural speech, and the most substantive way to train that is listening dictation done regularly, rather than reading through the grammar book one more time.