4-3-2

A speaker delivers the same talk to three successive partners in four, then three, then two minutes — the shrinking time limit forces progressively faster, more automatic delivery.

Procedure

1. Give students a familiar topic on which they have ideas and some vocabulary (a topic from earlier in the lesson works well). Allow 2 minutes of silent preparation — thinking only, no notes.
2. Arrange students in two rows facing each other, or two concentric circles, so each student has a partner.
3. Speaker A talks for exactly 4 minutes; Speaker B listens. When time is called, B gives one piece of brief feedback (one thing they found interesting).
4. Speaker A moves to a new partner. Delivers the same talk for 3 minutes.
5. Speaker A moves to a third partner. Delivers the same talk for 2 minutes.
6. Switch roles: Speaker B completes rounds of 4/3/2 while A listens.

Tips

• Nation's research shows the third delivery is measurably more fluent than the first — faster, fewer hesitations, more complex lexis — without explicit instruction. The mechanism is automatisation through forced repetition.
• Topics must use language learners already know — this is fluency practice, not new learning.
• Unlike 3-3-3 Repeated Speaking (which varies partners at constant 3-minute turns), 4-3-2 specifically targets speed: students must convey the same information in progressively less time.
• Do not correct during the activity — the fluency benefit depends on uninterrupted, sustained output.
• Topic selection matters: students need enough content knowledge to fill 4 minutes. The brief preparation period is non-negotiable; without it, students stall in round one and the mechanism collapses.
• Use a visible timer and give 30-second warnings before transitions. Number students 1–2; 1s move clockwise each round to avoid repeat partners.
• Works well as a weekly routine rather than a one-off: fluency gains compound across sessions.