Week 12: What a designed language cannot fix

The closing argument, and the empty row from week 5 — there is still no phrase for "I think this clearance is wrong"

24 May 2027

All three of us are here, which given week 7 means you will not get a consensus. Good.

What the twelve weeks claimed#

That ICAO phraseology is a designed artefact: authored, argued over, revised, and answerable to evidence. Not a tradition, not a list, not common sense written down.

And that it works. The reserved word, the closed vocabulary, the mandatory readback, the respelled digits — these are real engineering against a real channel, and the accident rate over the period they were introduced is not a coincidence.

What it cannot do#

It cannot make the sentence type survive. Charkhi Dadri. The digits arrived; the difference between information and instruction did not. Readback tests the content and nothing tests the type.

It cannot supply urgency the speaker withholds. Avianca 052. The phrase existed. Having a word for the emergency does not make somebody say it.

It cannot close a loop that both parties close correctly on false premises. Air Canada 759. Question answered accurately, about the wrong runway, by a controller with no way to know.

It cannot cover the non-routine. Which is why plain language exists, and why week 8’s rating scale exists, and why the hardest part of speaking on a frequency is the part with no phrase for it.

And it has no phrase for doubt.

The empty row#

Week 5, the mapping table, the row that never fills: the KLM flight engineer asked whether the other aircraft was clear, was answered, and did not ask again.

Fifty years on, there is still no standard phrase for I believe the clearance we are acting on is not valid.

Consider what such a phrase would need to do. It has to be sayable by a junior crew member to a senior one, on an open frequency, in front of everybody, without being an accusation. It has to be short. It has to be unmisinterpretable. And it has to compel a response rather than invite one.

Aviation has attacked this problem hard, and almost entirely outside phraseology: crew resource management, the two-challenge rule, explicit authority-gradient training. The industry decided this was a culture problem rather than a vocabulary problem.

Is that right? Three positions, and you have heard all three this semester:

• Halvard: a phrase would be ignored under the same authority gradient that silenced the challenge. You cannot vocabulary your way out of a hierarchy.
• Priya: a phrase gives a junior speaker cover. Saying a designated form is much easier than composing an objection, because the form carries the authority instead of the speaker.
• Noor: the fact that fifty years of design attention has not produced one is itself the evidence — and we should be suspicious of a gap that everyone agrees is a gap and nobody fills.

Pick one on Wednesday. You will be asked.

Why the audit went outside aviation#

Because the argument either transfers or it was never an argument.

The reasoning you have — find the channel, find what it loses, find where the language is ambiguous, remove the ambiguous form rather than clarifying it, and ask what the messy version was doing for free — is not about aeroplanes. It is about any group of people speaking under pressure on a channel that loses things, which is most of the important talking anybody does.

Your audits are the evidence about whether you have it. They are also, most years, better than the aviation examples at showing where the reasoning breaks down, because nobody has smoothed those domains into a story yet.

One last thing#

The most useful two words in this language are SAY AGAIN, and the reason is not that they are clear. It is that they are cheap. A system where admitting you did not understand costs you nothing is a system that finds out about misunderstandings while they are still cheap.

That is the design principle worth taking out of here, and it has almost nothing to do with aviation.