I have made an alphabet, but I think that I need to add more letters. I don't want to invent random letters.

Is there a letter that doesn't exist in most other languages?

My language's alphabet has 16 consonants and 5 vowels, but I think that I need to add a new letter like ñ or ç, but I don't want to add a letter that exists in common languages.

Is there any list of uncommon letters?

(My conlang's alphabet doesn't include b, g, v, y, z).

  • 2
    You should study a little phonetics; you don't seem to understand the difference between a phoneme and a letter, or an alphabet and a phonology.
    – prosfilaes
    Jul 14, 2022 at 20:42
  • 2
    So you want a letter that doesn't exist in other languages, but don't want to invent one? I think it would help if you clarify what you're looking for exactly. For example, why do you want a new letter? Do you already have a sound picked out for it?
    – Draconis
    Jul 14, 2022 at 21:11
  • @Draconis I want to invent a new letter, but I don't want to invent. I need a letter to pick a sound.
    – Fmbalbuena
    Jul 14, 2022 at 21:23
  • @Fmbalbuena But surely if you don't want to invent a new letter, it needs to already exist.
    – Draconis
    Jul 14, 2022 at 21:32
  • @Draconis I can't invent, but I need a letter that exists in some languages, but not in most common languages.
    – Fmbalbuena
    Jul 14, 2022 at 21:35

3 Answers 3


If you're looking for letters that are obscure but still used somewhere, the Case Variants of IPA Letters table can be a good resource. This lists IPA symbols that someone's invented a capital form of, which usually means they've been integrated into the orthography of some language (used somewhere), but are still an IPA symbol first and foremost (not too common).


Unicode is basically a list of characters that you can use on computers. The code charts offer a full list; if the rest of your alphabet is Latin-based, the various combining marks offer unlimited flexibility, and Latin Extended-B through -G offer many rarely-seen characters, including some never used in actual languages and some only used in long obsolete orthographies.

Making more interesting orthographies can be quite fun and make a conlang jump off the page. I quite encourage the use of ñ and ŋ for the nasal sounds normally associated with them, ʒ or ȥ instead of zh, even ƿ or ȣ for w. But while the first four letters don't really need explanation, the last two would probably make more sense in some sort of historical context.

I'd recommend picking your sounds and then finding appropriate letters for them, instead of the other way around, unless you have a historical reason in world otherwise. I'd also think about goals; if this is a language set in a setting where the writers use the Latin alphabet, most such alphabets use one or two diacritics and often diphthongs where Latin's alphabet doesn't cover it. If this is just for fun, it's a balance of interesting and visually attractive versus ease of understanding versus ease of typing. ñ isn't that far away on any normal system, and any of the other characters is a simple alt-xxxx sequence on Windows. But characters like ȥ, ƿ and ȣ noticeably stand out for me on this system, like they're being pulled from a different font. Your font selection may be limited if you go beyond Latin Extended A, and some other people may have a harder time seeing the characters (though that could be overstated; I'm seeing them on a fairly default Windows 10 box, and up-to-date Linux and MacOS boxes should display pretty much everything but the latest additions as well.)


Well, the number of letters that do not exist in any languages (or better, their orthographies) is potentially unbounded. As for uncommon letters, you might want to look into scripts used only for one (or few) languages, such as Armenian or Georgian, or uncommon diacritics, such as some Slovak and Czech letters (ř, ť, ď, ň, ľ, ĺ...) or Vietnamese for a diacritical explosion. But generally, if you want your language to be consistent and aesthetically pleasing, you should perhaps avoid mixing letters from different scripts (but there are exceptions, e.g. common transliteration of Old Church Slavonic into Latin script uses ь ъ and they fit the orthography nicely); or using diacritics haphazardly, or adding new letters just for its own sake - there should be some (perhaps unstated, perhaps not outlined in all the details) reason behind the orthography.

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