How to Practice English Speaking with a Language Partner Who Is Also Learning

Is it a bad idea to practice English speaking with a language partner who is also learning?

The most common objection to practising with a fellow learner is that you will copy each other's mistakes. Two people at B1 level reinforcing each other's wrong prepositions sounds like a recipe for fossilised errors. The concern is real, but it misunderstands what speaking practice is for.

Speaking practice builds automaticity: the ability to retrieve words and assemble sentences without pausing to translate. That skill develops through volume of output, not through perfect input. A learner who speaks for five hours a week with an imperfect partner will outpace a learner who speaks for thirty minutes with a perfect tutor, because fluency is a motor skill and motor skills respond to repetition.

The research on output supports this. Merrill Swain's output hypothesis argues that producing language, not just hearing it, forces learners to notice the gaps in what they can say. When you fumble for a word mid-sentence, your brain flags that gap, and you are far more likely to remember the word when you next meet it. Swain's output hypothesis explained shows that the act of struggling to produce language is itself the learning mechanism.

So the risk is not that you copy errors. The risk is that you drift into unstructured chat where nobody notices anything is wrong. The fix is structure, not a native speaker.

Why two learners beat one native speaker

Native speakers are, on average, terrible conversation partners for intermediate learners. They correct nothing, because correcting feels rude. They speak at full speed, because slowing down feels patronising. And they fill every silence, so the learner never has to push through the hard part of finding the next word. This is why a language partner who is also learning can actually be the better choice for speaking practice: you both share the same goal of maximum output, not a polished social exchange. Tutoro's conversation tutors run on the same principle, keeping the thread moving so you stay in output mode rather than waiting for a native speaker to carry the exchange.

A fellow learner has none of those problems. You share the same fear of silence, so you both understand why the other needs ten seconds to finish a sentence. You are both willing to repeat, rephrase, and ask "what do you mean?" without embarrassment. And crucially, you both want the same thing: maximum speaking time, not a polished social interaction.

There is also a motivational effect that native speakers cannot replicate. Watching a peer make progress is one of the strongest predictors of your own persistence. You are not comparing yourself to an unreachable native level; you are comparing yourself to someone one step ahead, which is exactly the distance that feels closable.

One thing a peer partner cannot reliably provide is accurate pronunciation feedback. Two learners can hear that a word sounded off, but neither can always say why or model the fix. For that specific gap, an AI English tutor that scores each word as good, fair or poor fills in what a peer cannot hear. But that is a supplement to your partner sessions, not a replacement for them.

Set one goal before you speak a single word

The single biggest reason learner-to-learner sessions fail is that they have no objective. Two people open the call, say "so, how was your week?", and thirty minutes later they have produced pleasant small talk and learned nothing. Small talk is the easiest thing you already know how to do, which makes it the worst possible practice.

Before every session, agree on one target. It should be specific enough that you can tell at the end whether you hit it. Good examples: "use the past perfect at least five times", "describe three rooms in detail using prepositions of place", "tell a story using ten new words from this week's list". Bad examples: "improve my fluency", "talk about travel".

Write the goal in the chat at the start of the session. Then, at the end, spend two minutes checking: did you hit it? If not, what blocked you? This single habit turns a social call into a training session, and it costs nothing.

This is also where an app with a structured curriculum helps. When you practice English with a partner around a goal you set together, you are improvising the lesson plan yourself. A tutor that runs the lesson inside the conversation removes that planning load, leaving you to focus entirely on speaking.

Use prompts, not free chat

Free conversation is the enemy of deliberate practice. It drifts to the topics you are already comfortable with, and it lets both of you hide from the structures you find hard. Prompts solve this by forcing you into specific language territory.

Build a shared list of prompts before you meet. Each prompt should push a particular grammar point or vocabulary set. For example: "Describe a time you changed your mind about something important" forces past tenses and opinion language. "Explain how to cook a dish from your country" forces sequencing words and imperatives. "Argue for or against remote work" forces conditionals and hedging.

Keep a bank of twenty prompts in a shared document. Each session, pick two. Give each person three minutes to speak on their prompt while the other listens. The timer matters: three minutes is long enough to force you past your comfortable phrases and into the ragged edge where real growth happens.

Record, replay, and spot the errors together

You cannot hear your own mistakes while you are speaking. Your brain is fully occupied with producing the next word, and there is no spare capacity for monitoring grammar. This is why recording is the highest-leverage tool in a learner-to-learner setup.

Record each three-minute turn on a phone. Then replay it together and hunt for errors, like two detectives reviewing footage. You will each catch things in the other that you completely missed in yourself. Write down every error you find, and at the end of the session, pick the three that matter most to fix next time.

The replay step does something subtle but powerful: it turns your own speech into input. Listening back to yourself is uncomfortable, and that discomfort is exactly what makes the errors memorable. Research on self-assessment through recordings shows that learners who listen to their own recorded speech develop sharper awareness of their recurring errors than those who only receive external corrections.

Do not skip the replay because it feels awkward. The awkwardness fades after two sessions, and the payoff is permanent.

Correct like a coach, not a teacher

Most learner pairs either never correct each other, or they correct everything, which is worse. Interrupting every sentence to fix an article kills fluency and confidence. The goal is a correction style that keeps the conversation flowing while still flagging the errors that matter.

Agree on a rule before you start: no interrupting. Instead, the listener writes down errors as they happen, then shares them after the speaker finishes their turn. This keeps the speaker in flow and gives the listener a real job to do, which is itself a form of practice.

When you do share corrections, focus on one category at a time. If your goal this session was past tenses, only correct past tense errors. Let the articles and prepositions go. Correcting one thing well beats correcting ten things badly, because a single focused correction is something the brain can actually retain.

Rotate who asks the questions

There is a hidden asymmetry in most language exchanges: one person asks questions and the other answers. The questioner gets easy practice with simple structures; the answerer gets the hard practice of producing extended speech. Both are valuable, but they build different skills, and if you never swap, you each get stuck in a groove.

Make question-asking an explicit role that rotates every session. The person asking questions is practising interrogative forms, follow-up questions, and active listening. The person answering is practising extended output. Swap halfway through, or alternate sessions entirely.

Follow-up questions deserve special attention, because they are the skill that makes a conversation feel alive. "Why do you think that?", "What happened next?", "How did you feel about it?" These are simple structures that most learners never practise, because in real conversations the native speaker always asks them. Practising them deliberately with a partner closes a gap that otherwise never closes.

What to do when you both go silent

Every learner-to-learner session hits a wall: one of you cannot find a word, the other cannot help, and the conversation stalls. What happens next determines whether the session is productive or a washout.

The wrong move is to switch to your shared native language. The moment you do, the English session is over and restarting it takes real effort. Instead, agree on a rescue strategy in advance. Options: describe the word you cannot find ("it's the thing you use to open a bottle"), say it in your native language and have your partner guess the English, or simply say "skip it" and move on, then look it up together at the end.

Silence itself is not the enemy. The pause where you search for a word is where retrieval happens, and retrieval is what makes vocabulary stick. The retrieval practice effect on memory is one of the most replicated findings in learning science: the effort of pulling a word from memory strengthens it far more than re-reading it. So let the silence breathe for a few seconds before you reach for the rescue phrase.

A 20-minute session you can run tomorrow

Here is a complete session template you can run with a partner tomorrow, no preparation beyond reading it once. It works at any level from A2 to B2.

Minutes 0-2: state one shared goal and write it down. Minutes 2-5: person A speaks for three minutes on a prompt while person B records and notes errors. Minutes 5-8: person B speaks, person A records. Minutes 8-12: replay both recordings together and list every error you find. Minutes 12-15: pick three errors and take turns producing corrected versions aloud, five times each. Minutes 15-20: two minutes of free conversation on any topic, where you both try to use the corrected forms naturally.

The final free-conversation block is important. It lets you rehearse the corrected language in a low-pressure setting, which is the step that moves it from "I know this" to "I can say this without thinking".

If you want to practice English speaking with a language partner who is also learning but cannot always find one, Tutoro gives you a conversation partner that is always available. Its tutors run the whole lesson inside the chat, introducing vocabulary with photos and audio, then reading your replies and adapting to your weak spots. Weak words and grammar resurface on their own days later, so what you practice in a partner session gets reinforced without you managing a review deck.

It runs on Android, and it is not for learners who want a human teacher's cultural nuance in every exchange.

Frequently asked questions

No, but you need structure. Two learners chatting freely will drift into comfortable topics and miss each other's errors. With one shared goal per session, timed prompts, recording and replay, and a no-interruption correction rule, a learner partner gives you something native speakers rarely do: unlimited low-stakes speaking time.

Aim for two or three sessions a week of twenty to thirty minutes. Consistency beats length. Three short sessions produce more fluency growth than one long one, because the retrieval and error-noticing happen each time you restart, not just when you speak for an hour straight.

Do not correct each other mid-sentence. The listener notes errors while the speaker talks, then shares them after the turn ends. Focus on one category at a time, such as past tenses, and let everything else go. Correcting one thing well is more useful than correcting ten things badly.

Set a rescue rule before you start: describe the word, say it in your native language and guess together, or skip it and look it up later. Most importantly, agree never to switch to your shared native language mid-session. The pause where you search for a word is where the learning actually happens.

They are different tools. A partner gives you volume, low stakes, and shared motivation, but cannot reliably correct your pronunciation or model the right sound. An AI tutor that scores each word as good, fair or poor fills that gap. Many learners use both: partner sessions for output, and an app for the feedback a peer cannot give.