Most European languages use some variation of the Latin alphabet. However, while most of them seem to broadly agree on what sounds most of the individual letters represent (with some minor differences, like whether or not /b d g/ are voiced, the exact phonetic value of /a/, etc.), some letters (including some digraphs) seem to have a lot of variance across languages. Some examples between English, German, Polish and (European) Spanish:

  • ⟨j⟩ in Spanish represents /x/, which in German and Polish is written ⟨ch⟩ (also ⟨h⟩ in Polish); meanwhile, Polish and German ⟨j⟩ represents /j/, which is written ⟨y⟩ in English and Spanish; and English and Spanish ⟨ch⟩ represents /tʃ/, which is written ⟨tsch⟩ in German.
  • Polish and German ⟨w⟩ represents /v/ like English ⟨v⟩, while English ⟨w⟩ represents /w/, written ⟨ł⟩ in Polish.
  • ⟨x⟩ represents /ks/ in Polish, German and English (also /gz/ in English), but in Catalan it (mostly) represents /ʃ/ – which is written ⟨sh⟩ in English and ⟨sch⟩ in German.
  • ⟨z⟩ represents /z/ in Polish and English, but /ts/ in German and /θ/ in Spanish; in English, /ts/ is written ⟨ts⟩ and /θ/ is ⟨th⟩, while in Polish, /ts/ is ⟨c⟩ (which in turn represents a mixture of /k/ and /s/ in English, /k/ and /θ/ in Spanish, and /k/ and /ts/ in German).

It has interested me for some years how many separate (single- and multi-language) groups there are of ways of pronouncing Latin letters (and whether that’s unique to Latin or is common also for, say, Cyrillic), but I can’t for the life of me even find the starting point to learn more.

So - what is the name of subdiscipline or theory that explores these differences in what pronunciations the letters of a single alphabet represent in different languages?

I don’t think the term I’m looking for is ‘orthographic depth’, since that concerns a single language in isolation, whereas I’m looking for commonalities and differences between several languages.

  • 1
    Start with grammatology.
    – Colin Fine
    Commented Oct 28, 2023 at 10:29
  • 2
    I don't know if any of it is online, but Bright & Daniels, The World's Writing Systems certainly discusses families of scripts together, detailing the different uses that language make of the same basic script.
    – Colin Fine
    Commented Oct 28, 2023 at 13:17
  • 1
    While Bright and Daniels is indeed the gold standard, Geoff Sampson's little book Writing Systems is a very nice look at the overall history, with careful attention to some of the more interesting cases, like Hebrew, Japanese, and Korean, among others.
    – jlawler
    Commented Oct 28, 2023 at 16:17
  • 2
    It doesn’t really affect the question itself, but you have some mistakes in which sounds letters actually represent in different languages: Firstly, German and Polish have significant differences between them: ⟨bdg⟩ Pl. voiced, Ge. unvoiced; ⟨y⟩ Pl. /ɨ/, De. /y/; ⟨z⟩ Pl. /z/, De. /ts/; ⟨c⟩ Pl. /ts/, De. mostly /k/. But also, Spanish ⟨j⟩ is /x/, not /h/; the sound /j/ is written ⟨y⟩ in Spanish, not ⟨ll⟩ (that represents /ʎ/); Spanish ⟨x⟩ is mostly one of /ks gz s/, in some words /x/, only rarely /ʃ/ (though it is /ʃ/ in Catalan). Commented Oct 28, 2023 at 19:20
  • 2
    No, its not necessarily poor hearing - it's an incomplete understanding of German, Polish, and Spanish phonology and phonetics that got the better of you.
    – Graham H.
    Commented Oct 28, 2023 at 23:21

1 Answer 1


I think the term is graphematics.

I found the following definition in a German textbook (Busch & Stenschke, Germanistische Linguistik, Tübingen 2013, p. 59): G. "Graphematik" = "the discipline that determines the distinctive units of writing systems of a particular language, and, based on that, the possible written representations of spoken language." ("written representation" is a clumsy way of rendering G. "Schreibungen", but I find it hard to translate; perhaps "spellings"?)

The phenomenon which you describe is referred to as the problem of "grapheme-phoneme-correspondence"

  • It indeed seems that "grapheme-phoneme correspondence" is a very useful term for my search - thank you. As per StackExchange custom, I'll give my question a few more days before I accept any answers - especially since that while I'm sure gpc will eventually lead me to proper research (if it exists), cursory search reveals many single-language papers and some multi-language papers that don't try to group languages. Either nobody bothers with gpc categorisation, the subject is an even more obscure sub-branch of graphematics, or there are no noticeable groups. Oh well.
    – Dragomok
    Commented Oct 28, 2023 at 18:45
  • 2
    It appears to have no practical applications. Ordinary spelling (even English spelling) can be figured out by AIs, if not by humans. There could be correlations with phonemes, but nobody cares.
    – jlawler
    Commented Oct 28, 2023 at 19:43
  • Alazon, I'd appreciate your input on this meta question: linguistics.meta.stackexchange.com/questions/2078/…
    – Dragomok
    Commented Oct 29, 2023 at 7:14

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.