Why English training is no longer a “nice to have” for global teams
A project update goes out to three offices. In London, it lands perfectly. In São Paulo, someone reads it...
Quick answer: Yes, you can improve parts of your English without speaking. Your reading, listening, grammar and vocabulary can all grow through silent study. But the part of your English you use in meetings and conversations (what we call “English 3” in this article) only improves when you actually speak and get feedback.
Someone says your name.
Everyone turns.
And the sentence you had ready two seconds ago is gone.
You say something. It’s fine. It’s just…smaller than what you meant. Shorter. Flatter. About 60% of the answer you actually had.
Then you go back to your desk and think: I know that word. I’ve known that word for ten years.
You do know it. That’s not the problem.
The problem is that knowing a word and finding it in two seconds are two completely different skills. And only one of them gets better while you study alone.
Short answer: Yes.
Reading, listening, grammar, vocabulary: all of it improves in silence. Millions of people have built excellent English this way. Nobody should tell them to stop.
But one part of your English does not improve in silence. It happens to be the part everyone can hear.
Most learners think their English is one number: B1, B2, “not bad”.
It’s actually three.
Think about your wardrobe:
Your English works the same way.
Your three Englishes
| What is it | Where it shows up | |
| English 1 | Words you understand | Reading a report. Watching a film. |
| English 2 | Words you can find help with | Flashcards. Gap-fills. Writing an email slowly. |
| English 3 | Words that arrive on their own | The meeting. The interview. Right now |
English 1 is always the biggest. English 3 is always the smallest.
That’s normal. It happens in every language, including yours. You understand thousands of words in your own language that you would never actually say.
But something happens to English 3 that nobody warns you about.
In 2025, researchers tested 306 university students in Turkey. They measured two things separately:
Results:
Six out of ten. At the level where most people say: “I’m fine, I just freeze sometimes.”
Notice that number. It’s the same one from the meeting at the top of this article. About 60% of what you actually had.
Here’s the strange part.
That number does get better as you improve. Your English 3 is slowly catching up.
But while it catches up, your English 1 is growing too.
Every book you read. Every film you watch. Every meeting you sit through. All of it feeds the big English as well as the small one.
This is why “I’ll just study harder” so often fails to fix the meeting problem. Studying isn’t wrong. It’s just feeding the English you’re already good at.
Think about faces and names.
You recognise hundreds of people instantly:
Easy.
Now imagine standing at a party and having to say all their names.
Completely different feeling. Same people.
When you read a report, the words are already there. Your brain says: yes, I know that one.
When someone asks you a question, there are no words in front of you. Your brain has to:
…while you’re also listening, watching faces, and feeling the silence.
Same words. Completely different job.
In 2023, Pearson asked 5,000 English learners in Japan, Italy, Saudi Arabia, Brazil and the US one simple question: Why isn’t your English where you want it?
The top three answers:
Read those again. Not one of them says “I didn’t work hard enough.”
They all say the same thing in different words: I practised the wrong English.
It’s 2026. Fair question. Let’s take it seriously.
AI speaking partners are genuinely useful:
If speaking to a person makes you freeze, this is an excellent place to start – and the research agrees. AI practice lowers speaking anxiety and makes learners more willing to talk.
But look closely at what that research measured.
In one 2024 study, learners who practised with AI felt calmer and more confident. Good.
Then the researchers tested their actual speaking.
Feeling ready to speak and speaking better are not the same result. The first one is worth having. It just isn’t the second one.
And there’s one thing AI will never do:
Use AI. It belongs in your plan.
Just don’t confuse the practice room with the performance.
Before you change anything, measure it.
Read. One page of something you’d read for work anyway.
Speak. Close it. Open your phone. Record two minutes about what you just read.
Listen back. Count three things:
The distance between step 1 and step 3 is your gap.
It isn’t your level. It isn’t on your certificate. But it is the only number that matters when other people are in the room.
Most people find this uncomfortable. That’s the point. Comfortable practice is what got you here.
Almost all study builds English 1. These four build English 3, and none of them needs another person.
Is English on the front of your flashcards? Then you’re practising the skill you already have.
Turn them around:
Same cards. Opposite direction. Completely different skill.
Writing gives you unlimited time to think. Saying it takes that time away.
That difference is your gap. This attacks it directly.
You don’t meet random English. You meet the same five situations, again and again:
Pick one.
When you play it back, don’t hunt for grammar mistakes. Count the pauses.
Four weeks of this and English 3 will move. That’s more than most study plans ever manage.
And then you’ll hit the ceiling
These exercises have a limit. It’s a very specific shape:
Nobody interrupts you.
Nobody asks the question you weren’t ready for.
And nobody tells you what wasn’t clear.
That last one is everything.
Most of what holds professionals back at work isn’t grammar. The real pattern is clarity, tone and directness – not tenses.
You cannot hear those things in yourself. Somebody has to tell you that:
Your message sounded colder than you meant it, or
Your explanation lost the room in the second sentence.
Less than you fear. More than you’re doing.
You don’t need to move abroad. You don’t need to speak English all day.
Two to three hours a week does it for most people Our small-group online conversation classes are designed exactly for this:
Feedback from someone who knows what to listen for
If your goal is to:
…silent study will take you there. That’s a real result, and it’s worth having.
But if your goal involves another person?
That conversation you keep avoiding.
Then you need English 3. And English 3 only grows when you use it.
Apps and self-study can help you understand English, but confidence comes from speaking. When you practise real conversations, make mistakes, and receive feedback, English becomes something you use – not just something you study.
Yes. You can improve your reading, listening, grammar and vocabulary through silent study. But the English you use in real-time conversations (English 3) only improves when you speak and get feedback.
English 1: Words you understand when reading or listening.
English 2: Words you can use with time and tools (flashcards, writing slowly).
English 3: Words that come automatically in meetings, interviews and conversations.
For most professionals, 2–3 hours per week of real conversation with feedback is enough to see progress in spoken English.
AI speaking partners (for speaking practice) can reduce shyness and anxiety and improve willingness to speak, but studies show they do not significantly improve actual speaking performance on their own. Use AI as part of your speaking practice routine, but not as only speaking experience.
Read one page of text related to your job or work, then record yourself speaking for two minutes about it. Listen back and count long pauses, restarted sentences, and words you couldn’t find. The difference between what you understood and what you produced is your gap.
Live online English classes for continuous learning.
Yılmaz & Kavanoz (2025), Porta Linguarum 44 – receptive and productive vocabulary in 306 Turkish university learners
Pearson (2023), The Impact of English – 5,000 learners across five countries
Wang, Zou, Du & Wang (2024), System 127 – AI chatbots, willingness to communicate, and speaking performance
Laufer (1998), Applied Linguistics 19(2) – the original study separating the three vocabulary sizes
Did you like this article? Share it!
