Task-Based Speaking Activities That Beat Free Discussion
"Discuss in pairs" sounds active. In a class of 28, it usually means four students talk and the rest wait. One strong student gives a two-minute monologue. The weaker partner nods, says "yes" twice, and learns nothing. The problem is not the students. The task has no goal, no information gap, and no reason to actually listen.
If you already use shadowing as a warm-up—repeating a speaker's words aloud with a short delay—this is the companion page to shadowing in the classroom. That page covers the five-minute warm-up. This one covers the half hour after it.
In task-based language teaching (TBLT), language is a tool. The task has a non-language outcome: two matching schedules, a shared ranking, a drawn route. If students do not speak, the task cannot finish. A discussion has no such outcome. That is why a discussion can limp along on gestures and silence.
Task-based does not mean "fun game." It means the student's brain is busy solving a problem while English rides on top. The moment the problem disappears, the language often disappears with it. So every task below has a clear finish line.
Six tasks that run themselves
Spot-the-difference pairs
Give each student a copy of the same picture with eight small changes: a missing hat and a red door. They sit facing each other and cannot look. They describe what they see until they find all eight. The outcome is a numbered list both partners can point to. Nothing forces a quiet student to speak like a picture only the other person can see.
Ranking with forced consensus
Give pairs a list of twelve desert island items: a knife, a mirror, a rope, a fish hook, salt, a plastic sheet. They must agree on the three most useful and put them in order. Set a timer for eight minutes. Because they have to present one shared list, the weaker student cannot just nod.
You can change the list to medical supplies on a spacecraft or items to save from a fire. The point is the forced agreement. One list from two people always creates an argument.
Information-gap timetables
Student A has a half-filled weekly schedule. Student B has the other half. They cannot show the paper. They ask questions like "What time is the meeting on Tuesday?" until both timetables are complete. If one person stops talking, the schedule stays empty. That is the gap doing the teacher's work.
Map directions back-to-back
Give one student a simple map with a marked destination. Give the other the same map with no route marked. They sit back-to-back. The first gives directions, the second draws the route. The task ends when the second student marks the correct building, not when the teacher decides time is up.
Use maps with two or three possible turns. If the map is too simple, the listener guesses and the talk never happens.
Story reconstruction
Cut a six-picture story into individual images. Each student in a group of three holds two pictures. Nobody may show anyone else. They describe what they see and together decide the sequence. The outcome is the correct order of pictures. Showing a picture is cheating because it removes the only reason to speak.
Class survey with report-back
Give the class a question with four fixed options: "Which weekend job would you choose: dog walker, shelf stacker, kitchen helper, or phone assistant?" Each student asks five classmates, records one reason per answer, and reports the most common choice. The form gives weak learners a script they can repeat. The report-back makes them use the same vocabulary a second time.
A survey works because it sends students across the room. They say the same sentence twenty times, but to different ears. That repetition matters.
| Task | Finish line | Grouping |
|---|---|---|
| Spot-the-difference | All eight changes listed | Pairs, facing |
| Forced-consensus ranking | One agreed top three | Pairs, 8 min |
| Information-gap timetables | Both schedules filled | Pairs, papers hidden |
| Map directions | Right building marked | Pairs, back-to-back |
| Story reconstruction | Pictures in correct order | Threes, two cards each |
| Survey with report-back | Most common choice reported | Whole class, then solo |
The TBLT cycle in plain words
Pre-task: model the language and the task. Show a picture difference on the board and say one sentence: "My picture has a red coat, yours has a blue coat." Let students repeat the sentence once or twice. You are not teaching vocabulary. You are giving them a sentence they can steal.
Task: pairs work while you stay out of the way. Do not correct grammar. Do not hover. Write down two or three errors you hear, but save them.
Report: one or two pairs tell the class what they found or decided. This is the bridge between pair talk and public talk.
Language focus after: put the errors on the board and fix them together. That is the moment for language work. Correcting during the task turns it into a test. Students stop solving the problem and start trying to be accurate.
Grade the task, not the student
If the pair found five of the eight differences, the task worked. If the timetable has matching times, the task worked. You do not need to grade every sentence. A student who speaks broken English but completes the task is still practicing under real pressure. That is enough for a Tuesday afternoon.
You can record a quick note: "Pair 4: four differences, good questions." That is grading the outcome. It tells the student that the result matters, not the accent.
Where shadowing fits
I run five minutes of shadowing before a task. It gets the ears and voice in contact with English before the task starts. A few choppy sentences from a YouTube clip are enough. If you want the full setup, see shadowing in the classroom. Then hand out the pictures and leave them alone.
Homework that extends the task
After a class survey or ranking task, have students record a one-minute report-back on their phone. They already did the speaking in class. Recording at home gives them a private version of the same task without a partner to cover the gaps. It also gives you a one-minute audio sample you can listen to without grading every word. More on that at record yourself speaking English.
Do not assign "speak for five minutes about your weekend." That is a discussion prompt with no outcome. Assign the report from the survey. The content already exists. The student only has to say it alone.
Free discussion has a place. It is not a speaking activity.