Most people who fail the NAATI CCL speak the language beautifully. They argue in Hindi, joke in Hindi, ring their family in Hindi every Sunday. Then a recording plays thirty-five words of an English speaker, the beep sounds, and something goes wrong that has nothing to do with vocabulary. This is what that something usually is.
- The CCL scores accuracy and completeness under memory load, not eloquence.
- Qualifiers — 'about', 'usually', 'if', 'only' — get dropped even more often than numbers, and change the meaning when they go.
- Reported speech ('he says that...') is a silent tax: more words, longer segments, more to drop.
- Register is a scored dimension: aap is the safe default in professional settings, and over-formalising costs as much as being too casual.
- English loanwords are normal in spoken Hindi. Panicking and reaching for a literary substitute is usually the actual error.
What the test actually measures
Two dialogues. Each around 300 words, delivered to you in segments of roughly 35 words. Each dialogue is marked out of 45, so 90 in total. To pass you need 63 out of 90 and at least 29 out of 45 on each dialogue — one strong dialogue cannot carry a weak one.
Read those numbers again with fluency in mind. Nothing there rewards a rich vocabulary. The marks follow whether every unit of meaning the speaker produced arrived in the other language, unchanged. A fluent speaker and a hesitant one can lose the same segment for the same reason.
Habit 1: The quiet omission
Everyone expects to lose numbers, and numbers do go missing. But the detail fluent speakers drop most reliably is smaller than that: the qualifier. The word that narrows a claim.
“You should usuallytake it with food.” “The appointment is provisionallyon the 14th.” “We can sometimeswaive the fee.” “Bring the form onlyif you've already lodged.” Drop “usually”, “provisionally”, “sometimes” or “only” and the sentence still sounds fine — which is exactly why nobody catches themselves doing it. It also now says something the speaker did not say.
The fix
When you take notes, give qualifiers a symbol the same way you give numbers one. A tilde for “about / roughly”, a question mark for “might / provisionally”, an underline for “only”. Two seconds of writing protects a mark you would otherwise never know you lost. And when you replay your own practice recordings, do not listen for whether it sounded good — listen with the source text in front of you and tick off units of meaning one by one.
Habit 2: Sliding into reported speech
You interpret in the first person. The doctor says “I need to examine you”, and you say “I need to examine you” — as the doctor, not about the doctor. Almost every candidate knows this rule and a surprising number of them break it by the third segment.
It happens because reported speech is how you naturally relay a conversation at home. Under pressure you revert to the habit, and “the doctor is saying that he wants to…” starts creeping in. Two things then go wrong at once.
- 1Every segment gets longerYou have added a reporting frame to each sentence. More words to produce means more chances to lose the tail of the segment.
- 2Grammar has to be rebuilt on the flyShifting a first-person statement into a reported frame means re-choosing pronouns and verb forms mid-sentence. That is cognitive work spent on scaffolding rather than on remembering what was said.
- 3Some nuance quietly flattensReported speech tends to smooth out tone — urgency, hesitation, a direct request. Those are part of the message you were asked to carry.
If the speaker said “I”, you say “I”. You are not describing the conversation. You are the channel it passes through.
For the mechanics of holding first person steady across a long dialogue, see direct speech and first-person interpreting.
Habit 3: The register slip — aap, tum, tu
Hindi is generally described as having three levels of second-person address: aap, tum and tu, running from most formal to least. Descriptions differ in the details, and usage varies by region, family and generation — but the broad ordering is not controversial.
CCL dialogues are set where interpreters actually work: a clinic, a bank counter, a school office, a Centrelink desk, a solicitor's meeting. In those rooms, the participants are addressing each other as professionals or as strangers. aap is the safe default, and it stays the default for the whole dialogue.
| Form | Broadly used for | In a CCL dialogue |
|---|---|---|
| aap | Polite, respectful, formal address; strangers, elders, professional settings | Your default. Almost every CCL scenario sits here. |
| tum | Familiar address; friends, peers, often family | Rarely right. Reads as too casual for a clinic or an office. |
| tu | Intimate or very informal; can read as blunt depending on context and region | Effectively never appropriate in these settings. |
Habit 4: Over-formalising into literary Hindi
This is the trap that catches careful, well-read candidates, and it is counter-intuitive enough that it deserves its own section. Hindi has a well-documented stylistic range: a heavily Sanskritised register associated with literature, formal writing and some broadcast news, and a colloquial spoken register that most people actually use day to day.
Under exam pressure, candidates reach for the first one. It feels like showing respect for the test. It is not what the test wants.
A nurse asking a patient a routine question, rendered in the vocabulary of a written government notice. It is comprehensible, and it is wrong for the room — no nurse speaks that way at a bedside. Quality of language is a scored dimension, and “quality” here means fitting, not elevated.
There is a second cost. Reaching for the literary form takes time and working memory — and both are being spent while the rest of the segment is still sitting in your head, decaying. The elegant word is often purchased with a dropped clause at the end of the sentence.
Habit 5: Panicking over English loanwords
A candidate hears “appointment”, or “bank account”, or “bus stop”. They know that in ordinary speech they would simply use the English word inside a Hindi sentence, as speakers routinely do. Then a small alarm goes off: this is a Hindi test, I should use a proper Hindi word. They stall, hunt for a literary equivalent, and lose three seconds and the end of the segment.
English loanwords in everyday spoken Hindi are normal and thoroughly documented, both in India and among Hindi speakers in Australia. Using one is not, by itself, an error. The alarm is the error.
Find out which of the five is yours
These five habits do not affect everyone equally. Most candidates have one dominant leak and a couple of minor ones, and the fastest improvement comes from knowing which. An afternoon is enough to find out.
- 1Record yourself on any segment you have a transcript forAny English passage of about 35 words works. Play it once, close it, interpret into Hindi, record the result.
- 2Transcribe your own renderingIn writing. Painful, and non-negotiable — you cannot hear your own omissions while listening, because your brain fills them in from memory.
- 3Number the units of meaning in the sourceEvery fact, name, figure, qualifier and condition gets a number. Then tick each one that appears in your transcript.
- 4Tag every missOmission, addition, reported speech, register slip, or hesitation. Do this for six segments and the pattern is unmistakable.
Quick check
The five-habit checklist
Fixing the habit you found
Once you know which of the five is yours, the work is narrow and it responds fast. NaatiGuru's Hindi CCL preparation is built around exactly these leaks — qualifiers, first person, register and the loanword freeze — with practice dialogues across the community domains and a marking routine that shows you the leak instead of leaving you to guess at it.
The general interpreting guides on this site are language-neutral and worth your time too. Start with how the marking actually works, because knowing where marks come from is what turns vague practice into targeted practice. Note also that NaatiGuru is independent and is not affiliated with or endorsed by NAATI.