‘Let’s talk’ isn’t a method. Speaking grows from comprehensible oral input, retrieval, automatisation of high-frequency chunks, explicitly trained interaction and fluency work, sequenced from controlled to spontaneous with low anxiety.
Ten principles
- Flood with comprehensible oral input first. 2. Retrieval over exposure. 3. Recycle orally. 4. Automate high-frequency language. 5. Chunks before rules. 6. Teach speaking as micro-skills. 7. Controlled → spontaneous. 8. Reduce cognitive load. 9. Train interaction explicitly. 10. Low-anxiety environment.
Sequence
Comprehensible input → retrieval → guided practice → structured interaction → semi-spontaneous → spontaneous. Fluency tasks: 4/3/2, Market Place, Messengers.
Before a role-play
Read key sentences aloud → pair oral retrieval (faulty echo, delayed repetition, oral ping-pong) → correct planted errors → sentence-puzzle race.
Sources cited (unverified list)
Levelt (1989); DeKeyser (2007); Karpicke & Roediger (2008); Nation (2007, 2013); Segalowitz (2010); Wray (2002); Skehan (1998); Long (1996); MacIntyre (1999).
Apply in my context (Claude)
For PET/FCE Speaking Part 3, build the micro-skills (suggesting, agreeing, moving on) as retrieval drills before the collaborative task, then repeat the task.
Connections
Creative automatization - five criteria for fluency activities · Twelve conversation strategies worth teaching · Plan KET and PET speaking around the assessment criteria