Why you understand textbook English but not real speech
You understand your textbook audio. You follow slow podcasts made for learners. Then two native speakers start talking, and you catch one word in five. That doesn't mean your vocabulary is weak: you probably know every word they said. You just don't recognise those words in the shape they take in fast speech.
Yevhenii, founder of YouTongue
Nov 4, 2026 · 4 min read
Textbook audio is recorded slowly, word by word, by actors who pronounce every sound. Real speech changes words in four ways. Once you know them, you can train your ear for each one.
1. Words run together
In fast speech there are no pauses between words. When one word ends with a consonant and the next starts with a vowel, the consonant moves over: turn it off sounds like "tur-ni-toff", an apple like "a-napple". Your ear is looking for it and off, but they never arrive as separate words.
What helps: read short phrases out loud the linked way, slowly: "tur-ni-toff", "pi-ki-tup" (pick it up). When you have said it yourself, you will hear it.
2. Small words go weak
Grammar words — to, and, of, for, you, can — are almost never stressed, so they shrink. To becomes "tuh", and becomes "n", of becomes "uh": a cup of tea sounds like "a cuppa tea", rock and roll like "rock n roll". Learners often hear only the stressed words and lose the grammar between them.
What helps: expect the weak forms. When you hear "I want tuh go", your brain should fill in to without stopping.
3. Sounds change or disappear
Some sounds don't survive full speed at all:
- T becomes a quick tap in American English: whatever sounds like "whadever". (We wrote a whole post about the American t in water.)
- T disappears after n: twenty can sound like "twenny".
- Whole phrases melt into one word: I don't know becomes "dunno", going to becomes "gonna" (more in gonna, wanna, gotta).

Uh, you know, whatever. Whatever that song, you know.
- Twenty-eight. - Twenty-three.
Jubilee · “twenty” on YouTongue
Compare a British speaker saying party:
>> was my birthday party.
Would I Lie To You? · “party” on YouTongue
I dunno, taking something.
The Royal Institution · “dunno” on YouTongue
4. The rhythm is uneven
English gives a strong beat to the important words and rushes the rest. In I'm going to the shop, the beats fall on go and shop; everything in between is squeezed into the gaps. That's why a sentence of eight words can take almost as long as a sentence of three.
What helps: listen for the beats first. Mark the stressed words in a sentence you hear, then fill in the weak words between them.
A short daily routine
You don't fix this with more grammar. You fix it by hearing real speech often, at real speed, and checking yourself. Ten to fifteen minutes a day is enough to start:
- Do one dictation. Listen to a real clip with the captions hidden and type what you hear. The words you miss show you which of the four changes you don't hear yet. On YouTongue the Daily dictation is free, and the words you miss come back until you get them.
- Hear your missed words again. Search each one and play it in five or six clips by different speakers. Slow it to 0.75× if you need to, then go back to 1×.
- Shadow one line. Repeat a sentence right after the speaker, with the same rhythm. Saying the linked, weak, fast version yourself is the quickest way to start hearing it.
- Watch something you like with subtitles. When a word surprises you, stop and hear it in other clips.
The goal is not to understand everything at once. It is to make the four changes familiar, one by one, until fast speech sounds like the words you already know.