How to Design Language Role-Plays That Transfer

Design a language role-play around a real outcome, private information, role constraints, one manageable complication, and a transfer round with changed details. Give learners the goal and facts, not a complete dialogue. A role-play can provide evidence of flexible interaction only when the other person's next turn is not fully known.

“Pretend you are in a restaurant” is a theme. “Agree on a meal for two people with different allergies and a fixed budget” is a communicative scenario.

Build from function and context

The Council of Europe provides illustrative CEFR descriptors for communicative activities through its official Companion Volume publication record. The Companion Volume frames descriptors as reference material rather than a prescribed curriculum. That makes “what must the speakers accomplish?” a better starting point than a vocabulary theme.

Write a one-sentence outcome:

• choose one supplier and document the reason;
• obtain a correct replacement or refund;
• find an appointment both people can attend;
• explain a symptom and confirm the next action;
• resolve which task each teammate owns.

Then identify whether the task is mainly spoken interaction. If one person delivers a prepared speech while the other waits, the scenario is not testing much turn-by-turn negotiation.

Use the TRANSFER scenario canvas

T — Target outcome

State a result both parties can verify. “Practise conditionals” is a language target, not an outcome. “Agree on a backup plan for bad weather” creates a reason to use conditions.

R — Roles with legitimate goals

Give each person a reason to participate. Avoid making one role a machine that reads questions. A customer wants a fast solution; an agent must follow a policy and protect limited stock.

A — Asymmetric information

Each role knows something the other needs. Separate calendars, budgets, policies, symptoms, preferences, or route options. This prevents memorising both sides.

N — Non-negotiable constraints

Add one or two boundaries: maximum price, unavailable date, allergy, approval rule, or delivery window. Too many constraints turn language practice into a logic puzzle.

S — Surprise or trouble

Introduce one complication after the exchange begins. It should force clarification, adaptation, or conversation repair, not merely punish the learner.

F — Feedback rule

Decide in advance what stops the task and what waits. A dangerous misunderstanding may require immediate correction; lower-priority forms can be saved. Use the STOP–SAVE–SKIP rule.

E — Evidence

Track outcome, information exchanged, clarification, repair, and level of support. Do not score acting talent.

R — Reconfigured round

Change the facts, roles, or constraint while preserving the communicative function. This transfer round tests whether language resources survive beyond the rehearsed case.

Example: return a faulty product

Element	Speaker A: customer	Speaker B: store agent
Goal	Leave with a working device before a trip	Apply policy and resolve the case
Private fact	Device failed after two days; receipt is digital	Refunds require a receipt; one replacement remains
Constraint	Cannot wait for repair	Must verify purchase date
Complication	Replacement is a different colour	Customer must choose replacement or later refund
Evidence	Explains fault, supplies proof, makes decision	Clarifies timeline, explains options, confirms outcome

Do not give either speaker model sentences before round one. Provide a small language bank only if the task is beyond current reach: “The issue is…,” “Could you clarify…?,” “The available options are…,” and “Just to confirm…”

After feedback, transfer to a damaged rental bicycle. The return policy, urgency, and desired outcome change. The function—report a problem, understand options, negotiate a resolution—remains.

Calibrate difficulty one lever at a time

Role-plays become harder through:

• less preparation;
• more unfamiliar content;
• weaker visual support;
• a new relationship or formality level;
• less predictable follow-up;
• one extra constraint;
• noisier or time-limited conditions.

Change one lever between rounds. If topic, speed, vocabulary, role, and outcome all change, failure reveals little.

For beginners, make the outcome concrete, offer visible choices, and keep turns short. For more experienced learners, require reasons, polite disagreement, and contingency planning. Do not equate advanced practice with obscure topics.

Review transfer honestly

After both rounds, ask:

1. Did the participants reach a valid outcome?
2. Did each discover the other's private information through language?
3. Which misunderstanding was repaired?
4. Which support was used?
5. Did the target function appear in the reconfigured round?
6. What failed because of language rather than scenario confusion?

A learner who completes the original restaurant task from memory but cannot negotiate a different budget has shown rehearsal, not strong transfer evidence. A learner who uses simpler language yet resolves both scenarios has demonstrated something more useful.

Before reusing a scenario, run a fairness check. Does one role require cultural, financial, medical, or professional knowledge the learner was never given? Could the same language function be tested with a less personal context? Give both roles enough information to reason, and separate a language breakdown from a domain-knowledge gap. If the “correct” outcome depends on an unstated local custom, revise the card or disclose that assumption. Transfer practice should change communicative conditions, not quietly test whether learners share the designer's background.

You can create role prompts with the LinguaLive tools. Keep private role cards separate and tell any AI partner exactly when to introduce the complication. Preserve the task design so a later attempt is comparable.

Limitations and disclosure

Role-play is a simulation. It cannot reproduce the social consequences, emotions, domain knowledge, power relations, or speed of every real encounter. Successful transfer to a second practice scenario does not guarantee real-world transfer. TRANSFER is an editorial scenario canvas, not a CEFR assessment instrument. Avoid simulations involving trauma, discrimination, medical danger, or legal jeopardy without appropriate expertise and consent.

Sources and editorial review

This guide was checked against its primary official or academic reference on 29 July 2026. Language usage can vary by region, relationship, and situation. Review the primary source.