CELPIP Listening Part 6 (Viewpoints): Three Minutes, One Voice, and the Trap of Partial Agreement

What Part 6 actually is

The longest clip in the whole Listening section is delivered by a single speaker — and that speaker changes position as they go.

Part 6 is "Listening to Viewpoints": one speaker, one audio clip, about three minutes, six questions. Nothing else in the section runs that long. (Mentioned once so it doesn't trip you up: celpip.ca's test-format page calls it "Listening to Viewpoints", while the Study Pack, the Participant Package and the July 2026 Guidebook call it "Listening for Viewpoints". Same task.)

Because there is only one person talking, people listen for "their opinion" as though it were one fixed thing. It is not. A viewpoints monologue typically concedes something, then qualifies it, then lands somewhere narrower than where it started — and the six questions are built precisely on those turns.

The failure: partial agreement

Here is the failure mode nobody names. Call it partial agreement.

The speaker agrees with the premise but rejects the proposal. Or they supported something last year and do not now. Or they back the idea everywhere except here, except at this price. Each of those is a sentence you genuinely hear — and an option restating it word for word can still be the wrong answer, because it was said in the concession rather than in the conclusion.

Accuracy is not the test. Placement is. An option is correct when it reflects what the speaker believes by the end, not merely something true that came out of their mouth on the way there.

The trap also runs the other way: an option that takes the real objection and strips the qualifier off it. They said "not on this site." The option says "opposed to the project." That is the speaker's view with the volume turned up, and turned-up is wrong.

A worked example

(original AlmiWorld practice item)

Three consecutive extracts from a monologue on a proposal to relocate a town library.

Concession. "Let me start by giving the proposal its due, because I think the people behind it are right about the problem. The Carleton Street building is too small. The children's section has been in a corridor for six years. I have made that argument myself, in print, more than once, and I have not changed my mind about it."

Qualification. "Where I part company with them is the leap from 'this building is inadequate' to 'therefore we move to the highway site'. That parcel sits outside the bus loop. For a family without a car — and roughly a third of our regular borrowers are in that position — a fifteen-minute walk becomes a transfer and a fare."

Landing. "So my position, in the end, is narrower than people assume when they hear I have concerns. I am not against a new library. I would vote for one tomorrow. What I am against is this parcel of land. Put the same money into the empty block behind the community centre and I would be first in the queue to support it."

Question: What is the speaker's overall position on the proposal?

• A. The Carleton Street building is too small for the community it serves.
• B. He is opposed to constructing a new library.
• C. He supports building a new library but objects to the site that has been selected.
• D. He has reversed a position he previously argued for publicly.

C is correct. It carries both halves — the support and the limit — and it is written from the landing, not from anywhere earlier.

A is wrong in the most uncomfortable way: it is accurate. He says it, almost in those words. But it is the concession — it reports what he agrees with the proposal's authors about, and the question asked for his position on the proposal. This option catches strong listeners, because recognising a line you definitely heard feels like confirmation.

B is wrong because it is stronger than what he held. He states the opposite outright — "I am not against a new library" — and offers to vote for one. B is the shape of his objection with the qualifier removed.

D is wrong because it inverts a detail he was careful about. He mentions arguing in print, then says he has not changed his mind. D takes a true fact — his public record — and attaches a reversal that never happened.

The shape that works

1. Assume movement. Go in expecting the position to shift at least once. If it does not, you have lost nothing.
2. Split the page before the audio starts. One narrow column for what he grants, one wide column for what he argues. Every claim goes in one column or the other.
3. Listen for the pivot, not for keywords. The most valuable second in three minutes is the one where the speaker turns. Mark it hard — an arrow, a line across the page.
4. Write the last thirty seconds most carefully. The landing is where the position answers come from, and it is the part your memory has held for the shortest time.
5. Keep the qualifiers attached. "Not yet", "not here", "not at that price" — write them next to the claim, not underneath it. They separate the right option from the too-strong one.
6. Read every option to the end before choosing. The options are on screen, not in your ears — the format page confirms you are required to read answer choices in the Listening Test. A correct-sounding first clause with a wrong second clause is the standard build.

Signal phrases for a moving position

Group them by what they do to the answer, not by what they mean alone.

Concession markers — true, but not the answer. Admittedly… It is true that… I will grant that… No one disputes… There is a real case for… To be fair to them… I have argued this myself. Whatever follows one of these is genuine, and it will very likely appear in a wrong option.

Reversal markers — the position starts here. That said… Where I part company… But here is my difficulty… And yet… Even so… What troubles me is… The problem comes when… This is your pivot. Everything before it was scaffolding.

Hedges and scope narrowers — the qualifier is load-bearing. In principle… up to a point… to some extent… in the short term… for certain neighbourhoods… not on this timeline… as things stand. If an option states the view without the narrowing, the option is too strong.

Landing phrases — write these down verbatim if you can. So where does that leave me… My position, in the end… If I had to choose… What I would actually support is… Which is why I would… The correct answers to the position questions are written from these sentences.

Attributed views — heard correctly, owned wrongly. Supporters say… The argument goes… You will hear people claim… I used to think… My neighbour's view is… These are other people's opinions inside the speaker's mouth. Options love them.

The practical reality nobody mentions

Three minutes of one voice, then six questions. Your notes have to survive far longer here than anywhere else in the section — the audio plays one time only, the test moves forward on its own, and you cannot return to a previous part.

That changes what you write. Near-verbatim notes collapse here, because by question five you are hunting through a wall of your own handwriting. Paper and a pen are provided — use the space. Two columns, wide gaps, one arrow at the pivot, the last claim boxed. Sparse notes you can navigate beat full notes you cannot.

What actually costs you the mark

• Choosing the option you recognise instead of the option that answers the question.
• Treating the concession as the conclusion because it came first and you wrote it down most clearly.
• Taking a qualified objection and reading it as total opposition.
• Attributing to the speaker something they explicitly assigned to supporters, critics, or their younger self.
• Writing so much during the opening minute that you have no hand or attention left for the landing.
• Committing to an answer from the first clause of an option without reading the second.

Try one

[Try Listening Part 6 →] — one full task, marked, no card required.