Nobody is marking where you are from
Paragon says it in plain words: raters do not consider your accent. Two sentences on its myth-busting page, both verbatim:
"Accents are simply a result of a variety of circumstances, and have no bearing on how your CELPIP Speaking component is evaluated."
"Our raters specifically do not consider your accent when rating your responses."
But that sentence gets repeated online in a way that quietly becomes false — "so pronunciation doesn't matter." It does. Pronunciation sits inside the rubric, under a dimension called Listenability, alongside rhythm and intonation.
The distinction is not a technicality, and it is the most useful thing anyone can tell a nervous test-taker: nobody is marking where you are from — somebody is marking whether a listener has to work.
The failure: hearing "accent is not rated" as "speech sounds are not rated"
The two claims sit one word apart and they are not the same claim. One is about identity. The other is about whether the person listening can follow you the first time, at speed, with no chance to ask you to repeat yourself.
Read the two quotes above, then read the rubric below, and you can see the difference yourself. This page reports what is published. It will not tell you your accent is fine and it will not tell you it is a problem, because that is not what the rubric measures.
What Speaking is actually rated on
Four named dimensions, from Paragon's Performance Standards and restated on the test results page:
| Speaking dimension | Writing's equivalent |
|---|---|
| Content/Coherence | Content/Coherence |
| Vocabulary | Vocabulary |
| Listenability | Readability |
| Task Fulfillment | Task Fulfillment |
Three of the four are identical across Speaking and Writing. Only the fourth differs — and that fourth one is the whole of what your voice contributes. Everything else on the list you could be graded on from a transcript.
Listenability is officially defined as:
"How easy is it to listen to and understand the response?"
Its sub-points include:
"Rhythm, pronunciation, and intonation"
There it is, in the rubric, in writing. Pronunciation is named. Accent is not — and no official source names "intelligibility" or "comprehensibility" as a criterion either. Listenability is the dimension that exists.
The threshold is "impedes communication", not "correct"
This is the part that changes how the whole thing feels. The level descriptors do not ask whether your vowels match anyone's. They ask what your speech sounds cost the listener. At level 8:
"Grammar and pronunciation errors rarely impede communication"
Lower down the scale:
"pronunciation difficulties may significantly impede communication"
Source for both: Paragon's Score Guidance PDF.
Read the verb. Not "errors are present" — errors that impede. A pronunciation feature that a listener absorbs without slowing down is not what those descriptors are describing. A feature that makes them stop and reconstruct the word is. That is a question about the listener's effort, and it is measurable in a way "correctness" never was.
What actually makes a speaker hard to follow
Mostly not the accent. Mostly these — and every one of them is a habit you can change inside a week, because none of them requires you to sound like a different person.
| Habit | What it does to the listener | Practise this |
|---|---|---|
| Speeding up when nervous | Words run together and the rhythm flattens | Record a task, play it back, halve the pace |
| Trailing off at the end of sentences | Your conclusion is the part that goes quiet | Keep volume up to the final word |
| Long unfilled pauses mid-clause | The listener loses the thread of the sentence | Pause between ideas, not inside them |
| Swallowing final consonants | Word endings vanish, tense and plurals with them | Read aloud, exaggerating last sounds |
| Flat intonation | Nothing marks where a sentence ends | Let your pitch drop at full stops |
Notice what these have in common. Every one is about rhythm and intonation — the two words that sit beside "pronunciation" in the same rubric line. And every one is fixable with a phone recording and ten minutes, without a coach and without changing your voice.
Speaking is 15 minutes, 8 tasks — under two minutes each including preparation. Speed is the one that bites hardest under that clock, because nerves and a short timer push the same direction. Slower is almost always more listenable than faster, and it is the cheapest change on this list.
What not to do about it
Do not try to imitate a Canadian accent for the test.
The rubric does not ask for one. It asks whether the response is easy to listen to and understand. Nothing published by Paragon rewards a target accent, and the two quotes at the top of this page say the opposite as directly as a test-maker can say anything.
There is also a practical cost. Accent modification is slow and difficult, and under exam pressure a half-learned accent tends to arrive as hesitation, self-monitoring and a broken rhythm — Listenability problems you did not have before. The effort is better spent on pace and word endings, which are days of work, not months.
What nobody can tell you
No official source publishes how the four dimensions are weighted. Not Paragon, not the Performance Standards, not the Score Guidance PDF.
So when someone tells you Listenability is worth a quarter of your Speaking score, or a tenth, or "less than content" — they are guessing. There is no raw-score or points breakdown for the four dimensions anywhere in the published material. What is known is that Listenability is one of four named dimensions and that pronunciation is inside it. How much it moves your level is not published, and this page will not invent a number for it.
One thing does outrank the rubric, though:
A response giving insufficient information to assess scores 0 or 1.
Levels 0 and 1 map to no Canadian Language Benchmark level at all. Slowing down so much that you say almost nothing is not a Listenability win — it trades one problem for a worse one.
What actually costs you
- Speeding up because you are nervous, then losing the last third of every sentence to it.
- Believing "accent is not rated" means speech sounds are not rated, and never recording yourself.
- Spending three months on accent modification the rubric never asked for.
- Dropping final consonants, so a rater hears "work" where you said "worked".
- Pausing in the middle of a clause instead of between ideas.
- Reading someone's confident claim about how much Listenability is "worth" and planning around it.
Try one
[Try a Speaking task →] — one full task, marked, no card required.