Linguistic competence in radiotelephony: what is assessed and why it expires

Cabine insonorisée avec casque-micro suspendu à son support, murs de mousse grise et lampe ambrée

A pilot who unthinkingly handles a sequencing constraint on a saturated platform, jokes with North American control, and no longer needs to repeat anything is given a level 4 — sometimes a 3. He hasn’t regressed: he was evaluated on something other than what he thought he mastered. The language proficiency rating doesn’t measure fluency in everyday English or the phraseology learned in school. It measures what remains when phraseology is no longer enough, and it has an expiration date because this reserve, precisely, fades without anyone noticing.

What the rating certifies, and what it does not

The requirement for language proficiency arose from a repeated investigative finding: in a series of major events, language was not the primary cause but the channel through which a recoverable situation became irrecoverable. The common factor in these cases was never the standard phraseology — it was known on both sides of the microphone. The common factor was the moment when the situation fell outside the grid: a failure, a refusal, an occupied runway, a request with no predefined wording. At that instant, everyone shifts to what ICAO calls plain language, and that’s exactly where the evaluation takes place.

Hence a first dissociation, which many candidates discover in the room. Mastery of phraseology is not graded in the language test. It falls under the practical license exams and the competency check, where the examiner judges the conduct of in-flight communication. The language test does the opposite of what is expected: it deliberately pushes you out of phraseology and observes what you do when the rails disappear. A candidate who responds to an unusual situation with a pile of standard formulas has not answered the question asked.

Second dissociation, less well accepted: what is certified is not the resemblance to a native speaker, but intelligibility within an international community. The implicit reference of the test is not an English speaker from London, but a controller or crew whose English is a second language, on a busy frequency and a poor signal. A very strong regional accent, a fast pace, and massive use of idiomatic expressions degrade this intelligibility — and are penalized in the grading, even for a native speaker. Being born into the language does not exempt anyone from evaluation and does not guarantee any level.

The test doesn’t check if you can speak English like a native speaker. It checks if you can communicate clearly in unexpected situations, even with poor radio signals and different accents. This is why even native speakers can fail.

An A320 captain with a level 4 rating might struggle during a go-around at a busy airport when ATC suddenly asks for a different approach due to wake turbulence. The pilot needs to quickly explain the new procedure to the first officer in plain English, not just standard phraseology.

Six descriptors, one score: the rule of the lowest

The scale has six levels, from pre-elementary to expert, and six independent descriptors: pronunciation, structure, vocabulary, fluency, comprehension, and interactions. Each is graded separately, and the level awarded is the lowest of the six. Not the average, not the dominant: the minimum.

Six scores, a single level awardedSix vertical bars represent the scores obtained on each descriptor. Five reach or exceed level four, marked by a dashed line, but the comprehension bar stops at level three and alone sets the level awarded.The level awarded is the lowest of the six60seuil 4455435Pronon.Struct.Vocab.AisanceCompr.Interac.level awarded: 3
A single weak descriptor brings down the whole endorsement. Five scores of 4 or 5 never make up for comprehension at 3: the rule is that of the minimum, because one broken link is enough to break the loop of an exchange.

This rule is not gratuitous severity; it describes the physics of radio communication. An exchange is a loop: transmit, be understood, understand in return, confirm. A vocabulary level of 6 is useless if comprehension is at 3, because the loop breaks at the same point in both cases. Each descriptor is a link, and the chain is worth its weakest link.

The six do not degrade in the same way. Pronunciation and vocabulary are relatively stable: they develop slowly and fade slowly. Structure — the ability to maintain a coherent past tense to report what happened, a conditional to negotiate an intention — is the first to give way under pressure. Fluency drops as soon as you have to build rather than recall. And interactions, often the lowest score among experienced practitioners, measure something no one trains for: explicitly asking for clarification instead of guessing.

  • Pronunciation and vocabulary degrade slowly
  • Structure fails first under pressure
  • Fluency drops when you must create new sentences
  • Interactions are often the weakest point

Why a certified competency has an expiration date

The question comes up at each deadline: why retake the test when you speak English every day in flight? Because “every day in flight” means, for most crews, a narrow and repetitive use — a limited repertoire of turns, a channel, a register. This use maintains phraseology and automation, not the reserve of plain language. The resource that serves once a year, the one for explaining a diversion or describing a failure, degrades in silence, and nothing in normal operations signals its loss.

Two men in a cockpit, one pointing at a display while wearing a headset, engaged in communication.
Clear communication is critical for safe and efficient operations.
Decline of plain language with intensity of useTwo curves start from the same point after the assessment. The solid curve, corresponding to broad and frequent use of the language, barely declines. The dashed curve, corresponding to use limited to routine phraseology, falls steadily and crosses the operational threshold before the end of the period shown.plain language reservetime since assessmentoperational thresholdfranchissementusage largeroutine use
The expiry date is not distrust, it is an estimate of the rate of loss. Two holders at the same level do not decline at the same rate; as regulation cannot measure actual use, it sets an interval that covers the unfavourable case.

The mechanism of the deadline then becomes clear: the interval is not indexed on the prestige of the level, but on the margin above the threshold. A level 4 holder is exactly at the threshold: the slightest erosion puts them below, so they are reviewed often. A level 5 has a full notch of margin, it can degrade a bit before becoming non-compliant, so it is spaced out. Level 6 describes a mastery reputed to be stable, that of a speaker whose competence no longer depends on active maintenance — hence, in the ICAO system, the absence of required re-evaluation.

Margin above the threshold and reassessment interval: the principle common to both regimes, and where Europe departs from it
Niveau Title Margin above the threshold Reassessment under ICAO Validity in the European system
4 Operational None: exactly at the threshold Periodic, the shortest of the three The shortest, around four years
5 Extended One step About twice the level 4 interval Around six years
6 Expert Two steps, proficiency deemed stable Not required Long validity, around nine years
3 and below Pre-operational and lower Below the threshold Not applicable: no international privilege Sans objet
Validity periods compared by levelThree pairs of horizontal bars, one pair per level, compare the reassessment interval set by ICAO as a dashed line and the validity in the European system as a solid line. Durations lengthen with the level, and the dashed line for the expert level runs to the edge of the frame to indicate no expiry.EuropeOACIN43 ans4 ansN56 ans6 ansN6no reassessment required9 ans0246810years
The European system lengthens the intervals at the bottom of the scale and does not exempt the expert level. A level 6 holder who thought they were set for life often discovers the expiry date when preparing a revalidation. The durations shown here are orders of magnitude: the date on your licence is authoritative.

The trap. An expired rating does not “partially” suspend the license: it immediately and without notice removes the privileges that depend on it. No upgrade flight or employer’s certificate replaces the re-evaluation. And because the score retained is the lowest of the six, a candidate who retakes the test thinking they only need to strengthen comprehension may come out with a lower level than before if another descriptor has slipped in the meantime.

Anatomy of a failure in a fluent speaker

The profiles that disappoint in the room are remarkably consistent, and none matches the image of the candidate who “doesn’t speak English.” The first is the one whose entire resource is based on routine: flawless phraseology, confident delivery, until the question with no formula. The request for description — what happened, what you plan to do, why you refuse — sends him back to a syntax he hasn’t constructed in years, and fluency collapses along with the structure.

Phraseology and plain language, four candidate profilesTwo-way matrix. The horizontal axis represents command of standard phraseology, the vertical axis the resource available in non-routine plain language. The bottom-right quadrant, solid phraseology and poor plain language, is highlighted as the profile that fails while believing itself at ease.non-routine plain languagebroad rangeweak phraseologyboth heldcompliant profileclear gaplong refresherbreaks down outside routinethe surprising profilecommand of standard phraseology
The dangerous quadrant is not that of the weak candidate. It is that of the practitioner whose phraseology masks a plain language reserve that has worn thin: nothing in daily operations signals it to them, and the test takes them there from the first question that falls outside the grid.

The second profile is the native or near-native speaker. He loses on pronunciation as understood by the test — intelligibility for an international audience — and on interactions: too fast a pace, contractions, idiomatic expressions, humorous turns, anticipated responses before the end of the question. He also loses because he never asks for clarification: for him, guessing works. The evaluator, on the other hand, is precisely looking for the opposite — the ability to detect a misunderstanding and explicitly resolve it.

The third is the candidate strong in transmission and fragile in reception. Comprehension is the most discriminating descriptor because it’s the only one you don’t control: you choose your words, but not the interlocutor’s accent, the quality of the signal, or the fact that they’re talking about something you weren’t expecting. The listening exercises deliberately include varied accents and degraded quality because that’s the real condition.

The right reflex. Prepare the test as an exercise in storytelling and negotiation, not as a general language oral or a phraseology review. Train three actions, out loud, in English: recounting a past technical event you’ve experienced; announcing an intention and defending it in the conditional against an objection; and formulating an explicit request for clarification rather than blindly reformulating. These three actions cover structure, fluency, and interactions — the three descriptors that cause experienced practitioners’ scores to drop.

What the evaluation actually constrains

The formats vary depending on the accredited organization and the authority, but the architecture is stable because it stems from the six descriptors. A first part tests comprehension alone: listening to recorded exchanges, often with unfamiliar accents and imperfect signals, and restating what was requested. A second part places the candidate in interaction on a non-routine situation: they must ask questions, obtain missing information, explain a constraint. A third part takes the form of an interview on professional topics, where lexical resources and the ability to develop are observed.

Sequence of an assessment and scoring ruleThree successive blocks represent the phases of the test: comprehension of recordings, interaction on a non-routine situation, topic-based interview. They converge on a block scoring the six descriptors, from which two outcomes follow: the endorsement with its expiry date if the lowest score reaches four, or no endorsement if a single descriptor remains below.1. comprehensiondegraded signal2. interactionhors routine3. interviewjob-related topicssix descriptors scored separatelythe lowest sets the levellowest at least 4endorsement and expiry datea single descriptor below 4no endorsement
Only two outcomes, and a single aggregation rule. The three phases carry no relative weighting: they feed the same six scores, the lowest of which decides everything, including the next expiry date.

Two characteristics of the procedure should be known before entering. First, grading is usually carried out by two people with distinct roles — the one who conducts the exchange and the one who evaluates — which makes any strategy of winning over the interlocutor ineffective. Second, the test is generally recorded. This recording is the basis for traceability, supervision of the organization by the authority, and any appeal: contesting a score without requesting a review of the recording has little effect.

A practical consequence follows: an aviation language test is not transferable from one profession to another. The situational content of a test intended for air traffic control is not the same as that of a test intended for flight crews, even though the scale and descriptors are identical. It’s the scenario that changes, not the grading.

What the type of operation changes

The same level is not maintained at the same pace depending on how you fly. What maintains the reserve is not the volume of exchanges but their variety: number of different interlocutors, diversity of accents, frequency of situations with no predefined wording. A long-haul crew that changes regions with each rotation receives this variety without effort; a short-haul crew that operates on a domestic network, mostly in French, accumulates thousands of exchanges without gaining much for the test. The latter speaks much more and erodes faster, which explains the surprise of the deadline for practitioners with very high activity.

The composition of the crew weighs just as much, and it’s the point that regulation cannot see. With two in the cockpit, a failing comprehension is constantly caught: the other pilot reconstructs the message, completes the readback, restarts the frequency. This redundancy masks individual drift until the day of evaluation, when it disappears. In single-pilot operations — aerial work, light business aviation, instruction in the right seat — there’s no safety net: comprehension becomes the sole barrier, and it’s precisely the descriptor you don’t control.

Two populations are exposed differently. The instructor and examiner bear the opposite burden of the practitioner: they must be understood by an interlocutor whose reserve is thin, which truly exercises interactions — reformulating, slowing down, verifying that the message has passed. The controller, on the other hand, changes the scene: his difficulty is not to describe a situation but to handle multiple exchanges at once and spot an incorrect readback in a continuous flow. That’s the fundamental reason why his test is not the same as a crew’s, even though the scale and descriptors are the same.

The mechanism that, on a daily basis, best indicates the real state of one’s own comprehension: the readback and its verification. A correct readback does not prove that you understood; it proves that you restated — a crew can faithfully repeat an instruction whose intent they didn’t grasp. The useful signal is elsewhere: the frequency with which you ask for confirmation, and especially how you ask. Asking for the entire message again is an admission that the loop has broken; precisely questioning the doubtful element — the level, the call sign, the reporting point — is exactly the trace the evaluator is looking for, and it’s also what costs the least airtime on a busy frequency.

Maintaining your rating rather than retaking it

The last deadline that is truly prepared is that of the file, not the test. Three checks are worth the time they take. The first: the date that counts is the one on the license, not the one on the certificate issued by the organization, and the transcription by the authority is not instantaneous. A test taken the day before the deadline can therefore leave a window, sometimes several days, during which the rating is not yet recorded: it’s the recording that is enforceable during a check, not the result obtained. The second: check what your operator does with this date. Some qualification management systems block scheduling before the deadline, others don’t monitor it, and the crew discovers the problem during a ramp check. The third concerns holders of licenses issued outside the European system: recognition of the rating during a conversion or validation is not automatic, and the host authority may require its own evaluation.

There remains an implicit trade-off that no one explicitly poses, and which has economic implications. Staying at the lowest level is accepting the most frequent re-evaluation cadence: each deadline mobilizes an entry, a day of unavailability, and the risk of not finding a session available when needed. Durably reaching the next level spaces these deadlines by as much, and the avoided expense accumulates over the course of a career. The transition is rarely about vocabulary: it’s about maintaining structure under pressure and conducting interactions, two habits that are built through voluntary exposure spread over time, not by a last-minute course.

Finally, an acknowledged gray area: the scale describes language behaviors, not mechanical scores, and the boundary between 4 and 5 on a given descriptor retains a part of subjective judgment, hence the importance of double grading and recording. For those preparing for an instrument rating, the point is unambiguous in the European system: English proficiency there conditions the exercise of privileges, it is not an add-on — the rating must be acquired and valid, independently of the language used on the usual frequencies.

The applicable text, the decision of your issuing authority, and the approved documentation of your operator prevail in all circumstances over this article, which has no other purpose than to explain the mechanics.

Commentaires

Leave a Reply

Découvrez l’écosystème FL330

FL330, c’est toute l’aviation professionnelle — explorez nos autres espaces.