Speak first. Fix one thing. Speak again.
Words and grammar are everywhere online, and they're free. What most learners are missing is a way of practising that turns them into speech. This is ours. It comes from the lessons that taught Muddassir to speak English, and we checked each step against language-learning research.
Six steps, every lesson, 7–10 minutes
Zayaan's rule for every lesson was simple: speak first → reflect → correct selectively → speak again. We added the two steps around it that research says matter most: easy listening at the start and recall later.
Hear a little
Start with a short, real conversation you can mostly understand. Not a grammar table.
Krashen; Nation — meaning-focused input
Speak first
Try to say something before you study any rule. Messy is the point.
Swain — output makes you notice what you can't say yet
Notice the gap
Compare what you wanted to say with what you managed to say. That gap is today's lesson.
Swain; Long — noticing during real interaction
Fix one thing
Correct only a few points, after you finish speaking — and try to fix them yourself first.
Lyster & Ranta — learners repair more when prompted than when simply corrected
Speak again, faster
Say the same thing again, in less time. Same ideas, better English.
Maurice 1983; Nation — the 4/3/2 fluency activity
Recall tomorrow
Test yourself on the new phrases a day later, then a few days later. Don't just reread.
Dunlosky et al. 2013 — practice testing and spaced practice
Why method beats more material
Most learners already own more English than they use: hundreds of words they recognise, grammar rules they could explain. What they can't do is find those words while someone is waiting for an answer. That only improves by speaking, noticing what went wrong and speaking again. Reading another word list won't do it.
So every Learn Angrezi lesson is the same loop with a different situation. Once the loop feels normal, you can run it on anything: a YouTube clip, a customer call, a chat with a friend.
The one thing we don't skip: enjoying it
Muddassir and Zayaan's lessons worked because they were fun. Even the boring topics felt easy. Research agrees that this matters: Horwitz, Horwitz and Cope (1986) showed that many learners feel real anxiety in language classes, mostly fear of being judged. That fear makes people go quiet, and a quiet learner doesn't practise. So we keep correction small, delayed and kind, and we always point out one thing you said well.
Basics still matter — just in the right order
Paul Nation recommends learning a few things to real fluency early: numbers, polite phrases, and phrases that control a conversation, like “Could you say that again?”. That's why lesson 1 is rescue phrases, not the alphabet. After that we cover the small things that make the biggest difference for Urdu speakers: contractions, a/an/the and tense control. Then come longer speaking tasks, like opinions, customer calls and stories.
A balanced week
Nation's “four strands” suggests spending roughly equal time on four kinds of practice. Here's how they look on Learn Angrezi:
- Meaning-focused input: the dialogue at the start of each lesson, plus podcasts and stories you enjoy (see free tools).
- Meaning-focused output: the “speak first” task. You say something real.
- Language-focused learning: the “notice” and “fix one thing” steps, a small amount of grammar and phrases.
- Fluency development: “speak again, faster”, using what you already know, with a bit of time pressure.
What we stopped recommending
Rereading notes and highlighting feel productive, but in Dunlosky and colleagues' 2013 review they rated low for learning. Testing yourself and spreading practice over days rated high. So each lesson ends with phrases you save and recall later, in Recall practice.
Our earlier 5-step path (listen, capture words, speak, write, review) is still good advice. The loop is the same idea, with speaking moved to the front and correction kept small.
The research, and the people behind it
We used work by named researchers and their universities, not anonymous “ultimate guides”. Where a summary page is linked, it cites the original paper.
- Paul Nation (Victoria University of Wellington): Best Practice in Vocabulary Teaching and Learning — the four strands
- Merrill Swain (University of Toronto): The output hypothesis — summary and references
- Michael H. Long (University of Maryland): The interaction hypothesis (1996) — summary and references
- Roy Lyster & Leila Ranta (McGill / Concordia): Corrective feedback and learner uptake (1997)
- John Dunlosky, Katherine Rawson et al. (Association for Psychological Science): Improving Students' Learning With Effective Learning Techniques (2013)
- Elaine K. Horwitz, Michael B. Horwitz & Joann Cope: Foreign Language Classroom Anxiety, The Modern Language Journal (1986)
- Stephen Krashen (University of Southern California): The input hypothesis — summary, references and criticism
- Te Kete Ipurangi (NZ Ministry of Education): 4/3/2 (Maurice, 1983) — how the activity works
Honest note: Krashen's ideas are influential and also debated, and Nation's 25% split is a guideline, not a law. No method gives fluency by a fixed date. What the research agrees on is that you need input, real speaking, a little focused correction and repeated recall.
Your first loop takes 8 minutes
Lesson 1 teaches the phrases that keep you talking when you forget a word. No sign-up.
Start lesson 1