The question paper says cheap; the audio says it won't stretch your budget. Nearly every answer line in the listening test is rewritten between the page and the recording, and candidates who wait to hear the words on the page wait forever. This lesson catalogues the eight rewriting devices the test uses, and retrains you to listen for meaning instead of matching sounds.
The core deception
Here is the mechanism behind most meaning misses (lesson 1, type M): your eye reads the question, loads its exact words into your ear as a search pattern, and your ear then hunts for those sounds. The test writers know this — so they guarantee the sounds never come. Worse, they often do play the exact words from a wrong option, so that sound-matching candidates catch the bait.
Rule one of IELTS listening: a word from the page heard verbatim in the audio is a warning, not a confirmation. The real answer usually arrives in disguise.
(The exception: concrete anchors — names, numbers, technical nouns — are rarely paraphrased. That is why lessons 4 and 6 have you anchor on them. Everything around the anchors gets rewritten.)
The eight devices
Every disguise the test uses is one of these. Learn them as a checklist, because after every practice test you will name which device beat you.
| # | Device | Page says | Audio says |
|---|---|---|---|
| 1 | Synonym | cheap | affordable / won't cost you much |
| 2 | Word-class shift | analysis | analytically / analyse |
| 3 | Hypernym ↔ hyponym | equipment | a helmet and gloves |
| 4 | Explanation | pedestrians | people walking in the street |
| 5 | Negation flip | not unusual | fairly common |
| 6 | Pronoun burial | marine animals | ...dolphins. They... |
| 7 | Structure flip | a talk by a researcher | a researcher giving a talk |
| 8 | Context match | tend not to | less likely to |
Three of these deserve a closer look, because they beat even strong candidates:
The negation flip (5). Not unusual = common. Didn't complete = dropped out. No charge = free. Your brain must do a polarity calculation mid-audio. When an option contains a negative (not, un-, -less, few), expect the audio to state the same idea with reversed polarity — and slow down before answering.
The pronoun burial (6). The keyword sentence plays early: "Most dolphins in the bay..." — but the answer arrives two sentences later, attached to they. Candidates stop listening after the keyword sentence and miss the answer that follows. The keyword opens the answer zone; it does not end it.
The explanation (4). Single hard word on the page, whole plain phrase in the audio (or the reverse: plain phrase on the page, hard word in the audio — understaffed for staff shortages). This device is why growing your reading vocabulary directly raises your listening score.
Retraining the ear: predict the disguise
The fix happens in the preview, not during the audio. For each question, take the two or three content words and spend one second asking: how else would a person say this?
Question: "The museum tour is free for ___." Your one-second forecast: free → "no charge", "won't pay anything", "at no cost", "complimentary".
You will not predict the exact phrasing — that is not the point. The act of forecasting unlocks your ear from the printed form, so that when "there's no charge for..." plays, it lands in the prepared slot instead of sliding past. Listening for a meaning is robust; listening for a sound is brittle.
The transcript autopsy: where the skill is actually built
Prediction is the exam-day skill; the training is what you do after each test. This is the highest-value fifteen minutes in listening study:
- Mark your test. Take every wrong answer and every answer you got right but hesitated on.
- In the transcript, find the answer sentence. Write a two-column line: page words → audio words.
- Name the device (1–8 from the table).
- Say the audio version aloud once. You are filing the equivalence in your ear, not just your notebook.
Ten tests of autopsies produces a personal paraphrase bank of 100+ pairs — and because the test recycles its devices (and even many of its specific pairs), your bank starts predicting future tests. This is also why our library structure matters here: every test's audio is recreated from real exam recalls with an exact transcript, so the page-words → audio-words comparison is always possible, line by line.
Cross-check with the glossary when an autopsy surfaces a word family you half-know — half-known words are exactly the ones that fail under audio speed.
Paraphrase and question types
- Gap-fill: the paraphrase lives around the gap; the gap word itself is spoken verbatim (lesson 6, rule 2). So paraphrase-listening gets you to the sentence; verbatim-catching fills the gap.
- Multiple choice: every option is a paraphrase battlefield — the wrong options quote the audio, the right option rewrites it. Full treatment in lesson 9.
- Matching: short options (booking required) against long audio explanations (you'll need to reserve a place in advance). Forecast all options in the preview.
Your drill (30 minutes)
- Open Listening 2026-05 Test 7 and sit Sections 2 and 3 under exam conditions.
- Mark them. Run the transcript autopsy on every miss and every hesitation: page words → audio words → device number.
- Count which devices caught you. Most candidates find two devices account for over half their meaning misses — those two go on the front page of your error log (lesson 12).
- Before your next test, do the one-second disguise forecast on every question in the preview. Then sit Section 3 of Listening 2026-01 Test 2 and see how many rewrites you caught in flight.
- Later this week, autopsy one more full test — Listening 2025-11 Test 1 — and add the pairs to your bank. One hundred pairs is the working threshold: past it, the test starts sounding like it is quoting your notebook.