The Science of Speaking 03 / 12 Interaction · Michael Long 
"Sorry, What?" Why Misunderstandings Teach You More Than Lessons
Michael Long spent the early 1980s doing something unglamorous: recording native speakers in conversation with learners, then counting what happened.
The assumption at the time was that native speakers help by making their language easier — shorter sentences, simpler words, slower delivery. The thing everyone does instinctively when they realise the person opposite is struggling. Long found plenty of that. He also found it was not the interesting part.
What stood out was the repair work. "Sorry?" "You mean the red one?" "The what?" Conversations between native speakers and learners turned out to be full of small breakdowns and small fixes. Long called this negotiation for meaning, and argued that it was where acquisition actually took place.
What a breakdown does to your brain
When a conversation stalls, several useful things happen at once. You find out exactly which part of your sentence failed — not in general, but that part. Your partner reformulates it, handing you the correct version at the precise moment you want it. And then you have to produce it again yourself.
A misunderstanding is a grammar lesson delivered on demand.
No amount of listening can supply this. A podcast never fails to understand you. A textbook never asks what you meant. Long's 1996 revision of the hypothesis made the claim sharper: interaction works because it puts input, attention and output in the same moment, and because the learner's own need to be understood is what drives it.
Simplified, or negotiated?
The prediction was testable, and it was tested. If what helps is easier language, then pre-simplified input should win. If what helps is the negotiation, then learners given normal language plus permission to interrupt should do better — even though their input is harder.
That result reframes what a conversation partner is for. Not a broadcaster of graded material — a person who notices when you are lost and lets you say so.
What to take from it
- Choose conversations with something at stake — a decision to reach, a problem to solve. Breakdowns only get repaired when being understood matters.
- Say "sorry, what?" instead of nodding. Nodding protects your pride and costs you the only part of the exchange that teaches you anything.
- Favour small groups over lectures and over large classes. The useful unit is turns per person, not minutes in the room.
Three articles, one argument. Input gets you understanding. Output shows you the gaps. Interaction is where the two meet — and it is the only one of the three you cannot do alone with a screen.
Stop reading about speaking.
Small groups, real conversations, a moderator who corrects you. Spanish, French, English or Ukrainian — online or in Barcelona.
Try a Conversation Lab →Sources
- Long, M. H. (1981). Input, interaction, and second language acquisition. Annals of the New York Academy of Sciences, 379, 259–278.
- Long, M. H. (1983). Native speaker/non-native speaker conversation and the negotiation of comprehensible input. Applied Linguistics, 4(2), 126–141. doi.org/10.1093/applin/4.2.126
- Long, M. H. (1996). The role of the linguistic environment in second language acquisition. In W. Ritchie & T. Bhatia (Eds.), Handbook of Second Language Acquisition (pp. 413–468). San Diego: Academic Press.
- Pica, T., Young, R., & Doughty, C. (1987). The impact of interaction on comprehension. TESOL Quarterly, 21(4), 737–758. doi.org/10.2307/3586992
- Mackey, A., & Goo, J. (2007). Interaction research in SLA: A meta-analysis and research synthesis. In A. Mackey (Ed.), Conversational Interaction in Second Language Acquisition (pp. 407–452). Oxford: Oxford University Press.