I thought my conlang would need a NATO-esque spelling alphabet. In my current sketch, they are words borrowed from Korean and few other languages. They are to serve as mere spelling alphabets, and not to have any meaning in my conlang. (And sorry for changing the phonology and orthography again)

Spelling alphabet IPA Origin Meaning
AGAŠI ɑ gɑ ɕi Korean "아가씨" madam
ÄGUGGA æ guk kɑ Korean "애국가" anthem
CÌNAMI t͡sɯ nɑ mi Japanese "津波" tsunami
ČIČARÌM t͡ɕi t͡ɕɑ ɾɯm Korean "치찰음" sibilant
EIGO e i go Japanese "英語" English
ÈRINI ə ɾi ni Korean "어린이" child
GIRÈGI ki ɾə gi Korean "기러기" goose
ISUNŠIN i sun ɕin Korean "이순신" Yi Sunsin
ÌMGÌGSÈN ɯm gɯk sən Korean "음극선" electron beam
KARSENTÈR k͡xɑɻ sen təɻ Korean "카센터" garage
LISSÌNÈR lis sɯ nəɻ English "listener" -
MINARI mi nɑ ɾi Korean "미나리" dropwort
NAPOLI nɑ po li Italian "Napoli" Naples
OSAKA o sɑ k͡xɑ Japanese "大阪" Osaka
ÖSAMČON ø sɑm t͡ɕon Korean "외삼촌" maternal uncle
PAULA pɑ u lɑ German "Paula" Paula
ROMEO ɻo me o English "Romeo" -
SAMAGÜ sɑ mɑ gy Korean "사마귀" mantis
ŠIGTAGPO ɕik tɑk po Korean "식탁보" tablecloth
TEGAMI te gɑ mi Japanese "手紙" letter
URETAN u ɾe tɑn Korean "우레탄" urethane
ÜSSARAM ys sɑ ɾɑm Korean "윗사람" elder / senior

At first glance, this seems to work well, especially for the syllables are distinctive per alphabet. But how can I ensure this? Is there a caveat I've overlooked?

  • I would have thought that, more or less by defintion, a "nato-esque spelling alphabet" would use words from the language it's being used to spell; those words would have meanings normally; and the semantically-nulled, spelling interpretation requires context. Commented Jul 10, 2023 at 17:44

3 Answers 3


The best way is to do the same thing they did in real life: test it! Record yourself spelling something out with these words, distort it and add some static, then see how well you can make out the words. (Or better yet, see how well someone else can make out the words.) See if any of the words sound too similar, or if you could shorten some of them without causing a problem (the NATO one has several one-syllable words).


A NATO style phonetic alphabet needs three things:

  1. The words have to be clearly distinct from other words in the list.
  2. The words should ideally be short with uncomplicated sounds.
  3. The words should be unique enough that if you know the alphabet, you don't need to actually hear the entire word to know what it is when transmitted over a bad radio connection or in a noisy environment.

Give you an example: let's say you have a patrol out in a city and they report having come under fire from a building, and for whatever reason they don't use grid coordinates but instead give you the street address. This is what you hear over a bad radio connection:

"...number fife niner tree lima (static) -dia papa (static) -erra (static) -arlie alpha (static) -ovem- (static) india."

If you know the alphabet, you wouldn't have much of an issue decoding that even having only received fragments of the words: they're reporting the building as 593 Lipscani. This works because none of the words share a sequence of the same syllables, or nearly the same syllables, in the same order.

So looking at your word list, you have "OSAKA" and ÖSAMČON. You have both words starting with two similar sounding syllables, and both are three syllables long. That can cause confusion over a bad connection.

Your word list should also keep it simple as possible for non-native speakers. The first phoneme is obviously the most important as it represents the required letter, but the rest should be as "basic" as possible, using sounds that are as common in multiple languages as possible to make it as easy to pronounce by as many different speakers as possible.

Going to the NATO alphabet, take "whiskey". You need it to represent "w", but the actual phoneme /w/ isn't present in many languages, thus you don't see that phoneme present in any of the other words. Someone speaking who doesn't have that phoneme in their language doesn't even have to worry about trying to pronounce it: if you hear "viskee" or "hwiskee" or "isskay" or something, it doesn't matter because you know what word they are trying to pronounce, thus what letter they are indicating.

And that works for all of the words. "Quebec" is often pronounced "kwa-bek" by English speakers, but is properly "kay-bek". Doesn't matter, it's distinct enough from all the other words that the receiver knows that word it's supposed to be.


You should also consider if you expect people with assorted accents and/or dialects and/or other first languages to be using this alphabet. "OSAKA" and ÖSAMČON provide the example here again. The speaker and listener may not differentiate between those two initial vowels; their own language may pronounce them the same, or they may not have one or the other, or neither. So that's why you need the words representing them to be as different as possible. Consider, in the NATO alphabet, "KILO" and "QUEBEC". If you pronounce the latter properly, both have the same initial consonant but there is absolutely no way to confuse the two words, and thus what letter they represent. Similarly, in my accent, the initial sounds in "ALPHA" and "OSCAR" aren't that different, but the words are so distinct there isn't a problem in determining if I mean "A" or "O".


To extend on Keith Morrison answer:

Split up all the words per syllables and put them in a spreadsheet program. Sort alphabetically and manually go through to validate that there's not a repetitive sounding written patterns.

Then go over sounds in a similar manner.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.