Listening: extracting the claim from spoken English
The Listening domain measures how well you pull meaning out of spoken English under time pressure. Items ask for the main idea, a supporting detail, an inference, the speaker's purpose or attitude, and the meaning of a phrase as used in the talk. You usually hear the passage once, so your notes - not your memory - carry the answer.
Core rules
- Read the question stem before the audio begins and predict what kind of information you need: a number, a reason, a contrast, a recommendation.
- Take shorthand notes: nouns, verbs, numbers, arrows. Write
pilot 6wk -> 3wk overinstead of a sentence. - Speakers rarely state the point first. Listen for the pivot: "but", "the real issue", "what this means is", "next time".
- Separate claim from evidence. A budget figure is evidence; the claim is what the figure proves.
- An inference must be forced by what was said. If you needed outside knowledge to get there, the option is wrong.
Worked example
Audio: "We budgeted six weeks for the pilot. Three weeks in, we had burned the whole training allotment - not because the software was hard, but because we kept changing the workflow. Next time we freeze the process before we touch the tool."
Stem: What is the speaker's main point?
Reasoning: The overspend is the setup. "Not because X but because Y" names the cause. "Next time" marks the takeaway: settle the process decision before the technology decision. An option saying "the pilot went over budget" is true, but it reports the evidence, not the point.
Traps that cost marks
- Choosing a true detail over the main claim.
- Missing negation and contrast, especially "not A but B".
- Over-reading emotion into a neutral speaker.
- Answering from general knowledge rather than from the recording.