Everyone Talks, Everyone Learns

What four articles taught me about talk in the classroom

Picture a typical lesson. The teacher asks a question. Three hands go up. The same three hands as yesterday. One child answers well. The teacher nods and moves on.

We call that a discussion. It isn’t. One child learned something. The rest of the class watched.

This week I came across four articles on talk and oracy. Two are by Tom Sherrington, one of the best known voices on teaching in the UK. One comes from the Blackpool Research School. One was written for Indian schools in EducationWorld. They come from different places. They say the same thing. If talking helps learning, then everyone needs to talk. [1, 3, 4, 7]

Here is what I took away.

Talking is thinking, out loud

In his latest post, Sherrington explains why talk is so powerful. The ideas in our head are a messy web. To explain them, we must put them in order, one word at a time. That act forces us to pay attention. It makes us recall what we already know. It pushes us to link new ideas to old ones. [1]

We stumble. We pause. We rephrase. Sometimes we realise we don’t actually know the thing we thought we knew. That moment is gold. It shows the gap.

He uses a sports example I love. In a basketball lesson, would you let only five players practise the lay up? Of course not. Everyone practises. So why let only a few students practise explaining? [1]

He also shares a summary of recent Belgian research by Speltincx and colleagues (2025). Getting every student to discuss the content, ideally in pairs, brought five gains: more learning, more engagement, better relationships, a stronger sense of belonging, and less disruptive behaviour. Even more striking, the time students spend talking about the content is linked to their results. [1, 2]

Every voice, literally

His second piece, from March, reflects on a training day with Voice 21, the UK’s oracy charity. Their Oracy Framework looks at talk through four strands: physical, linguistic, cognitive, and social and emotional. [3]

One idea stood out for me. Voice 21 says a good classroom “values every voice.” It has two meanings. First, literally every voice. Not just the volunteers. Not just the confident ones. Second, truly listening to what students say, and treating their ideas as worth hearing. [3]

His biggest takeaway? Teaching oracy is not an extra thing to squeeze in. The chances are already there in every good lesson. We just need the habit of using them. [3]

Behind every “Dunno”

The Blackpool article, by David Middleton, opens with a moment every teacher knows. The teacher asks a student to explain their thinking. A pause. A shrug. “Dunno.” [4]

He makes a sharp point. That shrug is often not a lack of thinking. It is a lack of confidence, words or structure. What students can say often falls behind what they can think. [4]

The evidence is solid. The Education Endowment Foundation finds that oral language approaches can add about six months of extra progress, especially when talk is structured and taught. The UK’s Oracy Education Commission (2024) calls oracy a foundational skill. And children from poorer homes are more likely to have weaker spoken language. So this is about fairness too. [4, 5, 6]

What I liked most was the school’s approach. Start small. A few form groups. A short article, read aloud by the teacher, then structured discussion. Try it, refine it, then take it into subjects. As the author puts it, this is not about adding something new. It is about being more deliberate with something already in every lesson: talk. [4]

A skill AI cannot replace

The fourth article brings it home to India. Georgie Ross of The Oracy Group writes that students can now produce a passable essay with AI in minutes. Speaking and listening, live and with purpose, is different. No tool can do it for them. [7]

Her diagnosis hit me hard. In many Indian classrooms, good speaking is treated as a personality trait. Something confident children are born with. Not something schools build. [7]

We see this every day. The articulate child becomes the default answerer. The quiet child stays quiet for years. In our multilingual classrooms, a child may think clearly in Tamil yet freeze when asked to answer in English. We read that as “doesn’t know.” Often, it is “can’t say it yet.”

Her call is simple. Every child, whatever their language background, gender or confidence, should get better at speaking and listening every year. [7]

Five moves to try tomorrow

None of this needs a new programme. It needs new habits.

1. Turn big questions into pair questions. Ask. Give quiet thinking time first. Then let partners talk for a minute. Now every student speaks, not one. [1, 8]
2. Fix talk partners in advance. When pairs are already set, the switch to talk takes seconds, not minutes. [8]
3. Make students practise explaining. After you teach a concept, each student explains it to a partner. Then they swap. Walk around and listen. You will hear the gaps. [1]
4. Give them sentence starters. “I agree with… because…” “I’d like to add…” “I see it differently because…” Starters take away the fear of the first word. [3]
5. Pick names, not hands. After pair talk, call on students by name. Nobody is caught out, because everyone has just rehearsed an answer. [3]

The real test

Next time you ask a question, count the hands. Then count the students. The gap between those two numbers is the learning we are leaving on the table.

Let’s close it. One pair conversation at a time.

References

1. Sherrington, T. (2026, October 11). Learning through talk: If talking helps learning, that means everyone needs to talk. teacherhead. https://teacherhead.com/2026/10/11/learning-through-talk-if-talking-helps-learning-that-means-everyone-needs-to-talk/
2. Speltincx, G., Vanhoof, S., et al. (2025). Gelijke kansen in de klas [Equal opportunities in the classroom], pp. 82–83. Summarised in Sherrington (2026, October 11), reference 1.
3. Sherrington, T. (2026, March 7). Discover oracy: A brilliant training day with Voice 21. teacherhead. https://teacherhead.com/2026/03/07/discover-oracy-a-brilliant-training-day-with-voice-21/
4. Middleton, D. (2026, March 26). From “Dunno” to dialogue: Why oracy, why now? Blackpool Research School, Research Schools Network. https://researchschool.org.uk/blackpool/news/from-dunno-to-dialogue-why-oracy-why-now
5. Education Endowment Foundation. (n.d.). Oral language interventions. Teaching and Learning Toolkit. https://educationendowmentfoundation.org.uk/education-evidence/teaching-learning-toolkit/oral-language-interventions
6. Oracy Education Commission. (2024). We need to talk: The report of the Oracy Education Commission. https://oracyeducationcommission.co.uk/oec-report/
7. Ross, G. (2026). Oracy: A skill artificial intelligence cannot replace. EducationWorld. https://educationworld.in/oracy-a-skill-artificial-intelligence-cannot-replace/
8. Sherrington, T. (2022, October 24). Think, pair, share forensics. teacherhead. https://teacherhead.com/2022/10/24/think-pair-share-forensics/

Numbers in square brackets in the text point to the references above. References 2, 5 and 6 are the research and reports the main articles draw on.