To Get ICAO Level 6 the Examiner Must Have No Doubt

Most rating scales let an examiner place you somewhere. This one does not.

If the examiner has doubts about any single element, no level is recorded at all, and the candidate is sent elsewhere.

SpeakShark is free for the six things an expert rating actually describes, three AI conversation sessions a day with no card.

In this guide: the six descriptors · all or nothing · doubt · one outcome · accent · the training aid · what to take from it · limits · method · FAQ

Key takeaways

• Six descriptors: pronunciation, structure, vocabulary, fluency, comprehension, interactions.
• All must be demonstrated. Examiners must be in no doubt.
• Doubt in any element means no level is recorded, not a lower one.
• A flight examiner can award Level 6 only, and cannot upgrade from 4 or 5.
• Standardisation runs through the ICAO Rated Speech Samples Training Aid.

The six descriptors

The UK CAA publishes the expert level criteria that must be consistently met before a rating is awarded, taken from AMC 2 FCL.055.

Pronunciation. Assumes a dialect or accent intelligible to the aeronautical community. Pronunciation, stress, rhythm and intonation, though possibly influenced by the first language or regional variation, almost never interfere with ease of understanding.

Structure. Relevant grammatical structures and sentence patterns are determined by language functions appropriate to the task. Both basic and complex grammatical structures and sentence patterns are consistently well controlled.

Vocabulary. Vocabulary range and accuracy are sufficient to communicate effectively on a wide variety of familiar and unfamiliar topics. Vocabulary is idiomatic, nuanced and sensitive to register.

Fluency. Able to speak at length with a natural, effortless flow. Varies speech flow for stylistic effect, for example to emphasise a point. Uses appropriate discourse markers and connectors spontaneously.

Comprehension. Consistently accurate in nearly all contexts and includes comprehension of linguistic and cultural subtleties.

Interactions. Interacts with ease in nearly all situations. Is sensitive to verbal and non-verbal cues, and responds to them appropriately.

Those are the regulator's words rather than ours, reproduced in full because the rule attached to them turns on every clause.

Three of those six are about something other than correctness. Varying speech flow for effect, understanding cultural subtleties and responding to non-verbal cues are not accuracy criteria, and no amount of grammar study produces them.

All of them, or none of it

The instruction attached to that list is the part that changes what the scale is.

To receive a Level 6 rating, a candidate must demonstrate all aspects of the Level 6 descriptors in the rating scale during the assessment. Examiners must be in no doubt that a candidate is an expert speaker.

Compare that with how the other requirements in this series work.

Every one of the others is a compensatory design in some form. This one is not compensatory at all.

Doubt produces no rating

Read the failure condition carefully, because it is not what most people expect.

In cases where the examiner has doubts about the level of attainment in any element, then no language proficiency level should be recorded, and the candidate is referred to a CAA approved English language proficiency testing organisation for formal assessment.

Not a lower level. No level.

That is a genuinely unusual rule and it follows from what the informal route is for. The examiner is not running a graded test; they are confirming that somebody is an expert speaker. If that confirmation is not available, the examiner has no basis for placing the candidate anywhere on the scale, and the honest output is a referral rather than a guess.

For a candidate it means an informal assessment cannot go partly right. You leave with an expert rating or with nothing but an appointment.

The examiner can award one level

The single-outcome design is stated directly. The UK CAA approved examiner can only award Language Proficiency Expert Level 6, cannot upgrade from Level 4 or 5.

So the informal route has exactly two results.

And the route is closed in advance to some candidates. Informal assessments must no longer be used where a candidate has previously demonstrated Level 4 or Level 5 in a formal assessment, and anybody holding a Level 4 or 5 assessment from the UK or another member state must attend a CAA approved language assessment centre, or a specially trained examiner, to renew or upgrade.

Who may use the informal route, and the linguistic history evidence an examiner must retain to run one, is covered in aviation's top English level asks about your childhood.

What it says about accent

The pronunciation descriptor deserves quoting because it settles an argument people have constantly.

Level 6 assumes a dialect or accent intelligible to the aeronautical community, and accepts that pronunciation, stress, rhythm and intonation may be influenced by the first language or regional variation, provided they almost never interfere with ease of understanding.

So the standard is not an accent. It is intelligibility, and the scale explicitly anticipates first-language influence at its own top level.

That is worth holding onto, because the informal route is largely aimed at first-language speakers, and first-language speakers have strong regional accents too. The descriptor is written to include a Glaswegian and a Texan, and to exclude anybody in any accent whose delivery makes listeners work.

The same principle appears in how the wider aviation scale treats communication, where the top level rewards adapting to a less proficient speaker rather than displaying range, covered in ICAO Level 6 rewards adapting to a weaker speaker.

How examiners are kept consistent

An all-or-nothing rule is only as good as the consistency of the people applying it, and the CAA points to the mechanism.

Examiners must familiarise themselves with the descriptors at Expert Level 6 of the ICAO rating scale, and should familiarise themselves with the ICAO Rated Speech Samples Training Aid, which the CAA says is designed to:

• raise awareness of what constitutes an acceptable or unacceptable level of English for communication between pilots and air traffic controllers
• provide examples of performance at each of the ICAO levels 2 to 6
• serve as an accurate and reliable reference for users
• promote rating standardisation between different raters and test service providers and between different regions of the world
• act as a training tool for initial and recurring training of raters and examiners
• support efforts to enhance proficiency test standards

Recorded speech samples at every level, used to calibrate raters across countries, is the most direct answer to the standardisation problem we have seen in this series. Cambridge does something comparable by publishing scored candidate performances, which is why we use them in Cambridge speaking criteria by level and Interactive Communication.

What to take from this

• Level 6 is confirmed, not scored. There is no partial credit.
• Doubt anywhere ends it. Prepare for the whole descriptor set or take the formal route.
• Accent is not the issue. Interference with understanding is.
• Three descriptors are about interaction, not accuracy.
• If you hold Level 4 or 5, use the formal route. The informal one is closed.

The three interactive descriptors are the ones that only develop against another speaker: varying delivery for effect, reading cultural subtleties, and responding to cues. Reading quietly does not touch any of them. Role play scenarios and a daily speaking partner are the practical routes, testing your speaking level free shows where you stand, and CEFR levels for speaking explains the general scale. What C2 actually means covers the comparable top level elsewhere, and English certificates that expire covers validity across bodies.

What we could not verify

We did not read AMC 2 FCL.055. The descriptors are as the CAA reproduces them.

We did not access the Rated Speech Samples Training Aid. We have described what the CAA says it is for.

We do not know how often informal assessments end in referral. No figures are published.

We did not check the levels 2 to 5 descriptors. This page reproduces the expert level only.

We did not check other national authorities. This is the UK implementation of an international standard.

This is not licensing advice. It is a reading of a published guidance page, linked below.

How we researched this guide

The descriptors are worth reading in full, which is why we reproduced them rather than summarising, but the finding is the instruction underneath them.

Most published criteria describe levels so an assessor can choose between them. These describe one level and then say the examiner must be in no doubt, and that any doubt produces no rating at all. That converts a scale into a threshold, and it is the only instrument in this series that works that way.

We noticed it by asking a question we have asked at every regulator: what happens to somebody who falls slightly short. Everywhere else the answer is a lower number, a permitted concession, or a resit of one component. Here the answer is that nothing is recorded.

The accent line we quoted deliberately. Candidates worry about accent constantly, and a top-level descriptor that explicitly anticipates first-language and regional influence is the most authoritative answer to that worry we have found.

Reading the three interactive descriptors together was the last step. Varying speech flow for stylistic effect, comprehending cultural subtleties and responding to non-verbal cues are not knowledge, and an assessment built on them is measuring something no written exam reaches.

Practise speaking, from SpeakShark

SpeakShark is an AI English speaking practice app, and three of these six descriptors describe things that only happen in conversation.

You cannot practise varying your delivery for effect while reading. You cannot practise responding to cues without something to respond to. Those are the descriptors that separate an expert speaker from an accurate one, and they need a partner.

You talk, the AI answers what you actually said, and you get speaking feedback while the conversation is still running. The free tier gives basic feedback; the detailed pronunciation and grammar breakdown is on Premium.

Being straight about the limits: a free session runs five minutes with four turns, we issue no certificates, and we do not teach radiotelephony phraseology or assess against the ICAO scale.

Start free with three sessions a day and no card. Paid sessions run ten minutes with unlimited turns. Limits are on the pricing page, and how it works walks through a session. Sign up here if interaction is the part you have never practised.

We are a speaking improvement tool. We are not an exam preparation provider, and we are not affiliated with Cambridge English, the IELTS partners, Pearson or any exam board. For test format, booking and official practice material, go to the exam body directly. Use SpeakShark to make your spoken English stronger, and use official material to learn the test.

• Site: speakshark.com
• Support: speaksharksupport@gmail.com
• Free tier: 3 AI conversation sessions per day, no card

Sources

• UK Civil Aviation Authority, language proficiency: the six expert level descriptors for pronunciation, structure, vocabulary, fluency, comprehension and interactions from AMC 2 FCL.055, the requirement to demonstrate all aspects with the examiner in no doubt, the rule that doubt in any element means no level is recorded and the candidate is referred for formal assessment, the statement that a CAA approved examiner can only award Level 6 and cannot upgrade from Level 4 or 5, the revised procedure closing informal assessment to previously formally rated candidates, and the description of the ICAO Rated Speech Samples Training Aid
• ELPAC, English Language Proficiency for Aeronautical Communication: the recognised test suite and its account of what the top level assesses
• Cambridge English, assessing speaking performance: the comparable practice of publishing scored candidate performances for rater standardisation
• Council of Europe, CEFR Table 1, common reference levels global scale: the general level scale used for comparison across this series

FAQ