This guide focuses on one small but demanding OET Optometry Speaking decision: how to move from a fictional client’s metaphor—“the words were swimming,” for example—to a mutually checked description without interpreting it. The skill is conversational calibration. Ask enough to understand the intended wording, then stop before the exchange becomes an improvised clinical assessment.
Every client, object, and exchange below is fictional. The page teaches communication language only. It does not diagnose or interpret symptoms, advise care, recommend treatment, tests, equipment, referral, or discharge, describe clinical procedures, or make professional judgments. OET Speaking and Writing are profession-specific; Listening and Reading are not. For related practice, visit the OET vocabulary practice hub and Exam English hub.
Treat the client’s phrase as a quotation to unpack
A metaphor is evidence of how the person chose to speak, not evidence of a clinical meaning. Keep it attached to the speaker:
“You used the phrase swimming. Could you describe what you noticed on the page?”
This opening has three advantages. It shows that the word belongs to the client, invites a concrete description, and does not offer a replacement. By contrast, “Do you mean blurred?” narrows the answer too early. “Was the page moving?” may also transform metaphor into a literal claim.
Once the client gives a concrete description, reflect it in ordinary language. If they say “I kept losing the beginning of the same line,” use that wording. Do not translate it into a diagnosis, mechanism, or specialist category.
Follow the image–place–time–count sequence
For a compact role play, four language dimensions often produce a clear account:
- Image: What exactly did the person say it looked like?
- Place: Which part of the named object did they mean?
- Time: How long, or at what supplied moment?
- Count: Once, repeatedly, or an approximate number of occasions?
Ask only one dimension per turn. “Which words looked broken, how long did it last, and has it happened before?” sounds efficient but makes it difficult for the listener to know what to answer first. A layered exchange feels more responsive:
“What did broken look like to you?” “Which part of the notice did you mean?” “About how long did that last in this example?” “Was that the single occasion described on the card?”
The final question checks a supplied boundary. It does not seek a real clinical history.
Keep the grammatical distance
Compare four formulations:
- The lower line disappeared.
- The lower line seemed to disappear.
- Nora said the lower line seemed to disappear.
- Nora used “disappeared” to mean she lost her place on the lower line.
The first asserts an external event. The second marks appearance but not source. The third attributes the appearance. The fourth records a negotiated meaning. In a clarification task, the third and fourth are usually the safest because they preserve both speaker and degree.
Small tense choices also matter. She noticed it once yesterday is a bounded event. She notices it may imply an ongoing pattern. She has noticed it connects past and present without giving a count. Use only the time relation supplied by the fictional role card.
Fictional encounter one: clarify without relabelling
Role-card facts: Elena says a line on a concert programme looked “frayed.” She means the final three letters seemed separated from the rest of the line. She noticed this once while reading the programme on Friday, for around ten seconds. She submitted a portal message later that day. The portal marks submission complete but gives no review status.
Candidate: “You described one line as frayed. What did that word mean in the concert programme?”
Elena: “The last three letters looked separated from the rest.”
Candidate: “Was that the top, middle, or bottom line?”
Elena: “The middle line.”
Candidate: “Let me check the wording: on Friday, the final three letters of the middle line seemed separated for about ten seconds, on one occasion. Is that your description?”
Elena: “Yes. Has my message been reviewed?”
Candidate: “The portal marks your submission complete. It does not show a review status, so I can’t describe a later stage from this card.”
The candidate earns precision through Elena’s answers. They do not decide that frayed equals another visual term, and they keep the administrative question separate.
Reflect emotion without confirming an interpretation
A fictional client may be worried, annoyed, or confused. Acknowledging that feeling does not require validating a conclusion about the reported experience.
Useful responses include:
- “I can hear that the uncertainty is frustrating.”
- “You seem concerned that I may misunderstand the word frayed.”
- “Thank you for explaining the wording more precisely.”
Avoid reassurance that predicts an outcome: It is nothing to worry about or Everything will be fine. Also avoid amplifying concern: That sounds serious. Neither judgment is supported by a language-only role card.
A combined response can acknowledge and clarify: “I understand that being represented accurately matters to you. Let me repeat the description in your words.” This keeps the conversation person-centred while staying inside the task.
Understanding is only the first step.
Lyra Practice helps you retrieve and use high-value workplace expressions in realistic situations until they feel natural.
Start a practice session →Use contrastive checks when words are close
Sometimes a client’s clarification leaves two plausible meanings. Offer a contrast only if both options come from what the person has already said.
“Do you mean the letters looked farther apart, or that you lost your place between them?”
This is different from introducing a novel interpretation. The contrast narrows two established descriptions. After the client chooses, reflect the selected version and drop the rejected one.
Do not use loaded contrasts such as mild or severe unless those exact categories appear on the card. They ask the client to accept the candidate’s scale.
Name what remains unknown
A clear clarification may still contain gaps. If the fictional card supplies no duration, say “The duration isn’t given here.” If the client cannot estimate, record that honestly: “Elena could not estimate how long it lasted.” Do not replace uncertainty with briefly unless the person used that word.
Administrative gaps need equally careful wording:
- no review status is shown does not mean not reviewed;
- no appointment time is recorded does not mean no appointment exists;
- a message is logged does not mean a reply is being prepared;
- a request is acknowledged does not mean the request is approved.
The fictional client should leave knowing the limit of the supplied record, not a plausible story about what happens next.
Fictional encounter two: repair a candidate’s overstatement
Facts: Yusuf calls a timetable “striped.” He clarifies that alternating rows seemed lighter while he read the page once on Monday. The card gives no duration. A callback request is logged; no callback time appears.
Poor summary: Yusuf had recurring striped vision for a short period and needs a callback.
Repair it clause by clause:
- had becomes reported to restore source;
- recurring is removed because the card says once;
- striped vision becomes the actual object-based description;
- for a short period is removed because duration is absent;
- needs becomes requested because that is the documented action.
Revised summary:
“Yusuf reported that alternating rows of a timetable seemed lighter while he read the page once on Monday. The duration is not supplied. His callback request is logged, but no callback time is recorded.”
This method—repair source, scope, object, measurement, and status—turns vague self-correction into a repeatable editing skill.
Close the loop with teach-back
After explaining a process distinction, ask the listener to state it in their own words. Keep the prompt specific:
“From this portal entry, what is complete and what status is still missing?”
If the response is “My message was reviewed, but no reply came,” preserve the accurate part and repair one point: “The submission is complete. The portal does not tell us whether review happened.” Then invite a second, shorter teach-back.
Do not make the listener recite the full visual description and process status at once. Separate content checks reduce memory demand and reveal exactly where meaning diverged.
Deterministic dialogue exercise
Use only these invented facts: Aisha calls a theatre seating chart “faded”; she means seat numbers 20–24 on the right looked lighter than the other numbers; she noticed this once for about fifteen seconds; she sent an online enquiry; the enquiry is marked received; response field blank.
1. Choose the best first clarification.
A. “Do you mean a serious blur?” B. “What did faded look like on the seating chart?” C. “Why did the numbers become faint?”
Answer: B. It invites Aisha’s meaning without severity or causation.
2. Supply a faithful reflection.
Answer: You said seat numbers 20 to 24 on the right of the chart looked lighter than the other numbers, once for about fifteen seconds. This retains comparison, location, count, and approximation.
3. Which administrative sentence is accurate?
A. “Your enquiry has been assessed.” B. “Your enquiry is marked received, and the response field is blank.” C. “A reply will arrive shortly.”
Answer: B. The other choices invent review or timing.
4. Aisha says, “So blank means they rejected it?” Give a one-point repair.
Answer: The blank field does not give a response status; it does not say the enquiry was rejected. This corrects only the unsupported inference.
5. Which detail answers “where” and which answers “how long”?
Answer: seat numbers 20–24 on the right answers where; about fifteen seconds answers how long. Keeping dimensions separate helps prevent omission.
6. Correct the unsafe summary: Aisha’s symptoms show that the chart problem requires testing.
Answer: Aisha reported that five seat numbers on the right of the chart looked lighter than the others. The revision returns to her description and removes interpretation and recommendation.
A rehearsal scorecard for language only
After practising, award one point for each communication achievement:
- the metaphor remains attributed to the fictional client;
- the concrete description uses the client’s clarified meaning;
- location, time, and count are not confused;
- missing details are named rather than guessed;
- emotion is acknowledged without reassurance or alarm;
- process status stops exactly where the card stops;
- the closing check targets one distinction.
This is not an OET score prediction or examiner rubric. It is a private editing device for ensuring that a rehearsal remains faithful and safe.
The narrow success condition is simple: the fictional client can recognize their own words in the summary, and neither participant has turned those words into an explanation. Precision comes from collaborative clarification, not from technical interpretation.
Sources
Related
- OET vocabulary for optometrists
- OET checking-understanding language
- OET Speaking empathy vocabulary
- OET vocabulary by subtest
To practise careful clarification language, try Lyra Practice.
Lyra Practice is not affiliated with or endorsed by OET or Cambridge Boxhill Language Assessment Unit Trust.