Phraseology Audit
Take a spoken protocol from outside aviation, audit it against what this course taught you, and propose a phraseology for it
At a glance
- Due
- 21 May 2027
- Weight
- 35% of the course
- Submit
- An audit of 1500 to 2000 words, a table of proposed phrases with the failure each one is designed against, and a limits section
- Format
- One PDF, with the proposed phraseology as a table rather than as prose
- Where
- The course submission folder, by Friday noon of week 11
Before that: Clear your domain. Week 9, in the sim session. Domains are cleared before you gather evidence, not after. The usual reason one is rejected is that it cannot be observed ethically, and that is far cheaper to find out in week 9 than in week 11. Where that happens.
Find a place outside aviation where people speak to each other under pressure on a channel that loses things. Audit it. Propose a phraseology.
This is the largest piece of work in the course and the only one that leaves aviation. That is deliberate: a course about one designed language is worth very little if the reasoning does not transfer, and the way to find out whether you have the reasoning is to take it somewhere it has not been done for you.
Choosing a domain#
It has to be a real protocol, with real speakers, that you can actually observe or read transcripts of. Candidates that have worked:
- Clinical handover — nursing shift handover, ambulance-to-ED handoff, surgical count. Some of these already have designed protocols; auditing an existing one is entirely fair game and often stronger.
- Rail — signaller-to-driver exchanges, degraded-mode working.
- Maritime — bridge-to-bridge, pilotage, VTS.
- Incident response — the conference bridge on a major software outage, where the channel is voice, the participants are tired, and nobody can see each other.
- Fireground and emergency dispatch.
What does not work: a domain you cannot get evidence about, or a domain with no pressure in it. A café order is a protocol but nothing is lost when it fails.
Clear your choice by week 9. Two people may take the same domain; you will be marked separately and you may not share evidence.
What you submit#
An audit, 1500–2000 words. What the protocol is, who speaks, what the channel loses, where the current language is ambiguous, and what the consequences of ambiguity are.
A proposed phraseology. A table of phrases with defined meanings. For each one: the failure it is designed against. A phrase with no named failure is decoration and will be marked as such.
A limits section. Explicitly: how could this proposal make things worse?
How it is marked#
Fit of the chosen domain (15). Is there real pressure, a real channel, and real evidence available?
Evidence gathered (30). First-hand observation, published transcripts, or documented incidents. Not what you imagine a nurse says. If you observed, say when, where, how much, and what you could not see.
The proposed phraseology (35). Judged on whether each phrase is designed against a failure you actually demonstrated, whether the set is small enough to learn, and whether it removes ambiguity rather than relabelling it.
Honesty about limits (20). Twenty per cent, because it is the criterion most likely to be skipped and the one that separates a design proposal from advocacy. Every constraint this course admires costs something: expressiveness, speed, the ability to say a thing the vocabulary does not cover. A proposal that claims no cost has not been thought about.
The strongest submissions in this criterion have argued that their own proposal should not be adopted.
Ethics#
If you observe people at work, you need their knowledge and consent, and you must not record without it. Handover conversations contain patient information; do not transcribe it. If your domain requires an ethics application, it is the wrong domain for a twelve-week course — choose published transcripts instead. Ask in week 9 if you are unsure.
The spec
The brief poses the problem and leaves room for a response. The spec is the fixed part: what a marker checks. Some lines can be checked mechanically; the rest call for judgement.
- the chosen domain is a real spoken protocol with real speakers
- the evidence is first-hand observation or published transcript, not recollection
- every proposed phrase states the failure it is designed against
- the limits section names at least one way the proposal could make things worse
How it is marked
| Criterion | Weight |
|---|---|
| Fit of the chosen domain | 15% |
| Evidence gathered | 30% |
| The proposed phraseology | 35% |
| Honesty about limits | 20% |
What each grade looks like
The criteria above say what is being marked. These say what each level of it looks like. Each band assumes the one before it.
| Grade | What work in this band does |
|---|---|
| PPass50–59 | The domain is a real spoken protocol with real speakers under real pressure, the evidence is first-hand or published rather than remembered, and a limits section exists. |
| CRCredit60–69 | Every proposed phrase names the failure it is designed against, and the evidence gathered is enough to show that failure actually happens in this domain. |
| DDistinction70–79 | The phrase set is small enough to learn and removes ambiguity rather than relabelling it, and the limits section names a specific way the proposal could make things worse. |
| HDHigh Distinction80–100 | Everything in Distinction, and the audit transfers this course's reasoning instead of transplanting aviation's vocabulary. The strongest work in this band has argued that its own proposal should not be adopted. |