Task-Based Language Teaching: 7 Activities That Work

Quick Answer: Task-based language teaching (TBLT) structures lessons around real-world tasks — ordering food, planning a trip, solving a problem — rather than grammar drills. Students complete the task first, then get language feedback. The seven activities below are the most adaptable for mixed-level ESL classes.

N.S. Prabhu’s 1987 Communicational Teaching Project in Bangalore produced one of the most cited results in applied linguistics: students who learned English by completing tasks outperformed students who studied grammar explicitly, even on grammar tests. That finding has held up in replication studies for forty years, yet most ESL classrooms still open with a grammar rule on the board. This article explains how to run task-based language teaching activities that change that pattern — without throwing out your existing curriculum.

What Separates a Real Task from a Grammar Exercise

The distinction matters because the wrong kind of “task” produces the same shallow output as a worksheet. A real TBLT task has three features: it has a communicative goal (not a grammar point), a non-linguistic outcome (a decision, a product, a list), and information gaps that force students to actually communicate to complete it.

A role-play where students are told “practise the modal ‘should'” is not a task — it’s a controlled practice drill with costumes. A role-play where two students plan a weekend trip with conflicting schedules and a fixed budget, and must reach a final itinerary, is a task. The difference is whether language is the subject of the lesson or the tool students use to do something that matters to them.

Researchers Ellis (2003) and Skehan (1998) both identify task authenticity as the key predictor of output quality. When students believe the task outcome matters — even fictionally — their language production is more complex and more fluent than when they know it’s just practice.

The Willis Framework: Three Phases You Need to Know

Jane Willis’s 1996 framework remains the clearest practical guide to TBLT lesson structure. It has three phases, and knowing them prevents the most common implementation mistake: skipping straight to the task and wondering why students revert to their L1.

Pre-Task: The teacher introduces the topic and task, activates relevant vocabulary, and models the outcome. This is not a grammar lesson — it’s a preview. Keep it under 10 minutes. Let students brainstorm words they already know, not words you want to teach them.

Task Cycle: Students complete the task in pairs or small groups (typically 10–15 minutes), then plan a short report on their outcome, then present or report to the class. The teacher monitors but does not interrupt to correct — feedback comes later.

Language Focus: After the task, students and teacher analyze language that came up during the task. This is where grammar instruction happens — but it’s tied to language students actually produced or needed, not pre-selected by the textbook. This phase is the most commonly skipped and the one that makes TBLT pedagogically defensible.

7 Task-Based Language Teaching Activities That Actually Work

The activities below are ranked by adaptability — the easiest to run with minimal prep first, more complex ones later. Each can be scaled for different proficiency levels.

1. Information Gap Tasks

Give Student A a train timetable with some times missing. Give Student B the completed version. Neither can show their sheet to the other. The only way to complete the task is to speak. Information gap tasks are the backbone of communicative teaching for a reason — they create genuine communicative need.

Variations: map directions (one student has the route, the other draws it), picture differences (spot what’s different between two versions), or schedule coordination (find a meeting time that works for both students’ busy weeks).

2. Jigsaw Reading and Sharing

Divide a longer text into four sections. Each group reads one section, discusses it, then recombines into mixed groups where each person shares what their section contained. The outcome: every student must understand their section well enough to explain it clearly, and every student needs all four pieces to complete the final task — answering comprehension questions or building a summary timeline.

Students who know they will need to explain what they read process text significantly more deeply than students reading for a test. See also the cooperative learning techniques that pair well with jigsaw at different proficiency levels.

3. Problem-Solving Tasks

Present a realistic problem with a defined set of constraints. Classic example: “You have $200 and 3 days to plan a trip for four people — here are four destination options, each with costs and pros and cons. Agree on a destination and present your reasoning.” Students must negotiate, compare, and justify — all using the target language — to reach a group decision.

The constraints are the engine. Remove them and the task collapses into opinion sharing. With them, students face real communicative pressure: they cannot simply agree with the loudest person because the numbers won’t add up.

4. Decision-Making Scenarios

Decision-making tasks are problem-solving tasks with higher ambiguity — there is no objectively correct answer, only supported opinions. Example: “Your school has funding to add one new class. Options are: coding, cooking, public speaking, or environmental science. Each group has two minutes to argue for their choice, then the class votes.” The voting mechanic makes the report stage feel real, not performative.

For lower-level classes, provide opinion language frames (“I think… because…”, “The best option is… because…”) in the pre-task phase. Provide them as tools for the task ahead, not as a grammar lesson.

5. Role-Play with Real Stakes

Most role-plays fail because students know nothing is real. Fix this by adding a decision that must be recorded: a written agreement, a class vote, or a product the students keep. A job interview role-play where other students fill out an actual evaluation form — and the “interviewer” must write one sentence of feedback — produces far more authentic language than “practise asking interview questions.”

The research on student motivation consistently shows that students produce their best language when they believe the outcome matters to someone other than the teacher. Role-plays with evaluative audiences — classmates, not just the teacher — create that condition without leaving the classroom.

6. Opinion Exchange Activities

Surveys and interviews that students design, conduct, and report back on work particularly well with adult learners. Students write five questions on a topic they’ve been studying — health habits, environmental behaviors, technology use — interview three classmates, tally results, and present a short findings report. The writing, interviewing, tallying, and presenting are all task stages, and each generates a different type of language production.

A key implementation note: the interview must happen with students they don’t usually talk to. If students always partner with their closest friend, they self-limit vocabulary and never stretch. Assign interview partners.

7. Project-Based Tasks

Longer projects — a class podcast episode, a travel guide to a local neighborhood, a short documentary script — span multiple lessons and create the most sustained language output of any TBLT activity. They also require the most scaffolding. The scaffolding techniques that work best for project tasks are chunked deadlines, peer review at each stage, and teacher conferences (not corrections) at the midpoint.

Projects suit intermediate and advanced learners best. For beginners, the cognitive load of managing project logistics on top of L2 production is usually too high — stick to single-session tasks until students have the confidence to sustain multi-day work.

How to Handle Grammar in a TBLT Lesson

This is where teachers get nervous, and the nervousness is understandable: if students never study grammar explicitly, won’t they fossilize errors? The short answer is yes — but TBLT is not a no-grammar methodology. The language focus phase after the task is where grammar instruction lives. The difference is sequence, not absence.

During the task, note errors that impede communication or that multiple students make. Write them on the board anonymously after the task cycle. Ask students to spot and correct the errors collaboratively before you explain the underlying rule. This process — noticing, hypothesis-forming, confirmation — produces deeper retention than a grammar presentation followed by gap fills, according to Schmidt’s 1990 Noticing Hypothesis, one of the most replicated findings in second language acquisition research.

Three Mistakes That Undermine TBLT in Practice

The first is interrupting the task to correct errors. Every interruption breaks the communicative flow and signals to students that accuracy matters more than meaning — the exact opposite of what TBLT is designed to teach. Take notes. Correct after.

The second is skipping the report stage. When students know they will present their outcome, they plan more carefully, use more varied vocabulary, and produce more structurally complex sentences during the task itself. The report stage is not optional — it is part of what generates the language complexity that makes TBLT more effective than drilling.

The third is choosing tasks that are too difficult for the proficiency level. A task should be challenging enough to require language the students are still acquiring, but not so difficult that students have nothing to say. If most students switch to their L1 within two minutes of starting, the task is too hard or the pre-task preparation was too short.

TBLT is not a magic methodology. It is a sequencing principle: meaning before form, communication before correction, outcomes before rules. Once you internalize that sequence, you can adapt almost any activity in your existing toolkit into a task.