First published on 16 March 2026 at https://sergiopantojaelt.substack.com/p/students-are-talking-now-what
You’ve set up the task. Students are engaged, excited, and speaking in English. You step back, smile, and think: “Nice. They’re talking.”
But then a familiar doubt creeps in.
Are they actually learning anything? Or are they just talking?
This is a question that haunted me more than once, especially in those moments when a lively speaking task ended, and there was nothing to show for it beyond “good energy.”
That’s when I came across Merrill Swain’s work on the role of output in second language learning. Her ideas fundamentally changed how I saw speaking activities, and what they were really for.
Swain challenged the popular view (Krashen’s, mainly) that learners just need enough comprehensible input and the rest will take care of itself. According to her, producing language plays a much bigger role in learning than we give it credit for.
And no, output isn’t just about fluency. It’s not just “practice makes perfect.” It’s about noticing, testing, reflecting.

Let’s break that down.
When students try to say something and get stuck, they often become aware of a gap in their knowledge. They realize:
I don’t know how to say this.
That’s not failure. Not at all. That’s progress.
It’s that internal “Hmm… is that right?” moment that pushes them to seek answers or try again. Speaking reveals what’s missing, and that awareness is where learning begins.
Try this in class: After any speaking task, ask:
Was there anything you wanted to say but didn’t know how?
Did you simplify or avoid anything because you weren’t sure how to say it?
When learners speak, they make guesses about how the language works. Sometimes they get it wrong. Sometimes a peer corrects them. Or they hear a better version. That’s not a problem. it’s a process.
Think of output as experimentation. Speaking isn’t just expressing; it’s testing mental models of grammar, vocabulary, and usage.
Support this in class:
Give students chances to revise. Run the task once, then again after feedback. Show them that refining output is part of learning, not a sign they failed the first time.
Speaking also pushes learners to talk about language itself, especially when they collaborate. Ever heard students say, “Wait, is it ‘didn’t went’ or ‘didn’t go’?” That’s gold!
That’s metalinguistic thinking in action. And it’s a sign they’re not just using language; they’re learning it.
Build this in:
Encourage paired work where students co-construct sentences or write together. Let them disagree, hesitate, ask questions. That’s where the real thinking happens.
Create space to reflect: not just on what they said, but on how they said it.
Revisit the task: let them try again, with better tools.
Prompt awareness: use questions that trigger noticing and reflection.
Highlight learner language: bring useful (or problematic) examples to the board and discuss them.
Make output visible: turn spoken output into written text for analysis and editing.
Fluency is important. But it’s not the endgame.
If we want speaking tasks to lead to development, not just expression, we need to design them with space for noticing, testing, and reflecting.
When learners talk, we listen: not just to what they say, but to what it reveals.
Our role is to turn their output into insight. That’s how talk becomes learning.