5

Of the languages I know about, most of them (not Chinese, Japanese, etc.) only have characters or character groups for specific sounds, and also can have a single specific sound generated by placing two or more characters together, such as (in English) 'sh', 'th', 'ch', 'oo', and long vowel sounds caused by an 'e' at the end of the word/syllable.

Also, some characters can represent multiple sounds put together, such as 'x', which in English sounds like 'ks' (or in some cases 'z'). Also, many characters can have multiple sounds, such as 'c' — it can have a hard 'k' or soft 's' sound. Most languages that use characters for sounds have these rules where sound can change depending on other characters paired/near the affected character, and I have not heard of any of these types of languages that has a large set of sounds they use and a single character for each sound that cannot be affected.

I don't know any languages where the sounds created by pairing a character with an 'h' (ex. 'ch', 'sh', 'th', 'ph'...) have a single character that represents them, or a character for all vowel sounds (ex. 'oo', 'ee', 'uh'...). For example, let's say that 'u' represents the 'oo' sound, 'i' represents the 'ee' sound, 'a' represents the 'ah' sound, 'o' represents the 'oh' sound (without the 'oo' drop at the end) and ☺ represents the 'ch' sound. (Some letters have multiple mouth positions, such as 'o', where you move your lips closer together while saying its sound — I am talking about a character for every mouth position, so this letter would become two letters put together, 'ou'.)

We would spell some words differently, such as "hi" --> "hai", "cockroach" --> cacrou☺, and "bone" --> "boun". Are there any languages that work this way, with a single independent character for every single sound made in that language? This would make it possible to always pronounce a word correctly just by reading it. If no languages such as the one I have described exist, then why? Is it because of the incorporation of words from other languages or the changing of dialect over time, or is it just a question of origin? It would be very interesting if someone could explain this to me (since I don't know much about linguistics).

  • 8
  • 1
    A lot of it probably has to do with historical reasons, but phonology is also involved. For example, would you have a phonemic English writing system distinguish between [t]/[d] and the flap in writing? There's also a tradeoff in some sense, e.g. this history SE question: history.stackexchange.com/questions/11017/… – user3898238 Dec 16 '14 at 3:13
  • Definitely not 100%, but Romanian comes pretty close. – Lucian Dec 16 '14 at 4:38
  • 1
    Slavic languages are also good candidates. – Lucian Dec 16 '14 at 4:50
  • 2
    The premise of your question is wrong because many languages with recent orthographies do! – curiousdannii Dec 16 '14 at 8:29
12

There seem to be several common confusions in your question:

  1. Phonetic vs. phonemic

Phoneme is a collection of sounds that serve the same function. For example, English phoneme /p/ sounds like [p] in 'spit' but like [ph] (h represents aspiration) in 'pit'. Speakers don't recognize them as two different sounds, though, because they have function e.g. to differentiate from /b/ pit/bit.

Therefore it would make no sense at all to try to have a different character (letter) for each phonetic sound because there would be hundreds for each language and the speakers would not know how to use them.

  1. Phonemic orthography vs. perfect phonemic orthography

Many languages have the kinds of orthographies you want. Finish, Italian, Czech, etc. all have more or less one letter for phoneme (with the occasional digraph or other inconsistency). The problem arises when a phoneme changes because of context. A common situation is final devoicing of consonants. For example, English 'but' and 'bud' are pronounced differently but a German speaker would pronounce both 'butt'. But they would not pronounce 'buds' and 'buts' identically. This is incredibly common in any language. For example, the Czech word 'had' (snake) is pronounced /hat/ because the voiced 'd' is devoiced (into 't') when it's the last sound in a word. But in all its other forms, the 'd' remains voiced. For for instance, in 'hady' (snakes, pl. acc.) remains voiced. This is different from the 'pit'/'spit' example because both /t/ and /d/ are phonemes in Czech and many native speakers would write 'had' as 'hat'. However, the orthography chose lexical/morphological consistency over phonological consistency. As a result, children have to learn how to spell these words - generally, by asking themselves what the other forms of the word sound like. However, there's no inconsistency in pronunciation because the devoicing happens no matter what. There are some orthographies that may choose phonemic consistency in some cases, but no orthography will be perfectly consistent in all respects because there is simply too much that is context-dependent in pronunciation (these many rules of referred to as phonotactics).

|improve this answer|||||
  • +1 Very nice answer. But, I'm not sure 'phoneme' is a group of sounds at all ... And certainly - if a phoneme is sounds, which I very much doubt, I'm not sure 'a group of sounds' is an ideal description. For a phoneme to be a group of sounds, it would have to have an exhaustive inventory as it were. I doubt any two sounds are ever exactly the same in the first place, if you see what I mean? – Araucaria - Not here any more. Dec 22 '14 at 2:22
  • 2
    I absolutely agree. If I wanted to be 'accurate' I would say something like a phoneme represents a continuum of phonetic realizations. Or even something like 'an exemplar cloud'. But in practice, I notice, that linguists often talk about phonemes as basically clusters of allophones-depending on the level of detail they want to evoke. So when I try to explain a phoneme, I resort to the group/collection schema hoping to make up in clarity what I lose in accuracy. – Dominik Lukes Dec 22 '14 at 8:03
  • Even then, the ⟨p⟩ in "spɪ̈t" isn't even a normal /p/, but a /pʱ/. Spit: /spʱɪt/, pork /pʰɔɹk/, strap: /st͡ʃɹæp/ – Matthew T. Scarbrough Jan 17 '18 at 7:46
6

Another hurdle to a 1-to-1 correspondence between phonemes and graphemes in a language has to do with language change and the fact that written language is more resistant to change than spoken language (and writing systems are even more reluctant to change), in general. In fact, you'll often find that languages that come the closest to having such a system are usually either dead, or have developed a written form fairly recently, or their written form is purposefully made to reflect the sounds rather than made to be standardized and preserved (for instance, languages like Lushootseed)

For the sake of an argument, imagine that there's a language that just had its 1-to-1 grapheme to phoneme correspondence writing system finalized (and commemorated the occasion by releasing a dictionary, some spellers, learning materials for students, etc.). Individuals become literate, books are written, maybe some political documents are composed. Language changes, and the sound system of a language changes over time as well. If you were to examine this newly written language some number of centuries in the future, you will almost certainly find that the sound pattern has indeed changed, but that the writing system most likely hasn't kept up. As the language continues to change, you will probably see that the graphemes and phonemes are falling out of sync. Maybe some new phonology emerges (for example, devoicing in certain environments like word-finally). Chances are, this will not be reflected in the written language if it's been sufficiently standardized at some point. You might also see evidence of an effort to keep the spelling and pronunciation in a 1-to-1 correspondence, but the effort wasn't 100% successful, resulting even in more disparity. So, as the writing system matures, it tends to diverge from the sounds of the language it's meant to represent. At this point in the history of humanity, we have a lot of languages with written systems old enough to have done that.

This isn't even necessarily an inconvenience. For example, you consider the noun "sign" and the noun "signal". "sign" appears to have that obnoxious trait where "ign" is pronounced in an unintuitive way as "ayn" whereas "signal" looks much nicer, its "ign" sounds like it's spelled. These things aren't nice when you're learning a language, but the inconvenient differences in spelling preserve evidence of an etymological relationship between the two. It would have been more difficult to figure out such a relationship between e.g. "sayn" and "signal" if the spelling and pronunciation kept in lockstep.

(Of course, the value of the knowledge that "sign" is related to "signal" can be questioned, and whether the ability to construct a picture of these connections is worth the difficulty in spelling/reading can be debated.)

|improve this answer|||||
1

Serbian is my mother tongue and that's exactly the way the language works. There's a rule which says: Write the way you speak, and read as it is writen, i.e. every single sound has its own character. It makes learning the writing system so easy, only you have to learn cyrillic. I suppose there are other languages with the same system, just have never wondered about that.

|improve this answer|||||
  • I would suspect that "exactly" there. Phonemes are already kind of what you hear, that means if you write the way you speak, and pronounce it as it is written then it can be called a phonemic language, not phonetic (meaning that every single sound does not have its own character). – Yanek Yuk Nov 19 '19 at 10:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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