Question - Languages with inconsistent spelling systems?

[Rui]

And where are you getting this 'can't have more vowels that consonants' thing?

WALS - Chapter 3 - there is no ratio lower than 1.11 - ergo, no value lower than 1

Correct me if I'm wrong, but it seems as though they are counting vowel qualities (well, they actually say that specifically), which I believe means they aren't counting length, tone, phonation, nasality, etc. but just the pure height/backness/roundedness of the vowels

[Wattmann]

And where are you getting this 'can't have more vowels that consonants' thing?

WALS - Chapter 3 - there is no ratio lower than 1.11 - ergo, no value lower than 1

Correct me if I'm wrong, but it seems as though they are counting vowel qualities (well, they actually say that specifically), which I believe means they aren't counting length, tone, phonation, nasality, etc. but just the pure height/backness/roundedness of the vowels

That is how I count them too.
If there's a series of long vowels, I consider them /V + :/, while in the case of Persian I won't (there are no corresponding opposite lengths)

[Chagen]

Wel, I consider things like long vowels to be seperate vowels in their own right. So this is merely a case of differnt analyzation styles leading to different transcriptions (though I know you'll just insist that yours is right).

[Wattmann]

Wel, I consider things like long vowels to be seperate vowels in their own right. So this is merely a case of differnt analyzation styles leading to different transcriptions (though I know you'll just insist that yours is right).

You're annoying, therefore you're wrong.
Did you even read what I had said?

[Rui]

That is how I count them too.
If there's a series of long vowels, I consider them /V + :/, while in the case of Persian I won't (there are no corresponding opposite lengths)

No, I mean they aren't counting length distinctions as separate vowel qualities. So if a language has /i i:/ then only /i/ is counted in the WALS page. This tends to reduce the amount of "vowels" that a language has, making the ratios larger

[Wattmann]

I understood you, and said that I count it that (their) way too, unless the length is not part of a full series X)
It depends on my mood though - I would count secondary consonant features as seperate...

[Chagen]

Wel, I consider things like long vowels to be seperate vowels in their own right. So this is merely a case of differnt analyzation styles leading to different transcriptions (though I know you'll just insist that yours is right).

You're annoying, therefore you're wrong.
Did you even read what I had said?

I did. You consider vowel qualities to be "base vowel + suprasegmental quality" while I consider them to be seperate phonemes in their own right.

To you, /i/ and /i:/ are just variants of /i/, I consider them to be two seperate phonemes.

[Rui]

I understood you, and said that I count it that (their) way too, unless the length is not part of a full series X)
It depends on my mood though - I would count secondary consonant features as seperate...

Then you apparently didn't understand what Drydic Guy was saying, because he was saying that Chagen's language could have had 10 vowel QUALITIES with several different PHONATIONS that total 50 vowels, whereas you were saying how having 50 vowels and only 23 consonants is impossible because no language has more vowels qualities than consonants. Not to mention the fact that Chagen flat out said he counted diphthongs as single vowel elements, which would contribute to the 50 vowel total.

Or else you are conflating vowel and vowel quality, which is incorrect in this case

(and yeah, no language has 50 distinct vowel QUALITIES because that's physically impossible)

I think you need to take your vendetta against Chagen somewhere else, because following him around every thread is pointless and just irritates everyone else.

[Wattmann]

I understood you, and said that I count it that (their) way too, unless the length is not part of a full series X)
It depends on my mood though - I would count secondary consonant features as seperate...

Then you apparently didn't understand what Drydic Guy was saying, because he was saying that Chagen's language could have had 10 vowel QUALITIES with several different PHONATIONS that total 50 vowels, whereas you were saying how having 50 vowels and only 23 consonants is impossible because no language has more vowels qualities than consonants. Not to mention the fact that Chagen flat out said he counted diphthongs as single vowel elements, which would contribute to the 50 vowel total.

Or else you are conflating vowel and vowel quality, which is incorrect in this case

(and yeah, no language has 50 distinct vowel QUALITIES because that's physically impossible)

I think you need to take your vendetta against Chagen somewhere else, because following him around every thread is pointless and just irritates everyone else.

First, I'm not following him around. I'm just visiting threads, likely like you, and, when I see something, I respond to it.

I didn't even see DG's post.
My point went null after Chagen said there's phonation - but you lot pressed on vowel length. My latest posts didn't even relate to Chagen's lang

[Legion]

Wattmann, I am not defending Chagen, whom I have already put on my ignore list, but I am beginning to consider adding you too on it.

[Pole, the]

@Chagen:

Of course /a/, /a:/, /ã/, /aˤ/, /a˞/, /à/, /a̋/ etc. are all different vowels on their own, but using totally different symbol for each of these would be highly impractical, making use of the written language very difficult.

Consider a language having /a i a: i:/ as its only vowel phonemes. It's extremely simple case - 4 is a relatively small number and no user of the lang would mind using separate letters for long and short variants.

But then see a language with /a e i ø y o u ã ẽ ĩ ø̃ ỹ õ ũ a: e: i: ø: y: o: u: ã: ẽ: ĩ: ø̃: ỹ: õ: ũ:/. It would be unlikely to have 28 signs for 28 phonemes, when they can be easily cathegorized. Using seven letters for different vowel qualities will be enough.

Phonology is only a tool.
As the whole science is.

Have in mind that people try to systematize much more complicated vowel sets, like this found in English.
The language has 20-30 vowel (dia)phonemes. Would you write each with separate letter (not a digraph)? Would it be more useful than that one with digraphs and diacritics?

See, that's not inconsistency. Not a bug but a feature.

[Chagen]

The main reason I'm iffy against tons of vowel digraphs is becasue of their ambiguity.

If you go all latin and have <oe> represent [oI], what do you do if the lang allows a /oe/ sequence?

Is <poe> /poI/ or /poe/?

Legion: Annoucing that you are blocking a member in public makes you look like an attention whore AND a twat. You want a fucking medal or something?

[Pole, the]

1. Then you actually use another digraph, say, <oi> or <oy>. Or use trema for vowel sequence: <oë> [oe]. There are many possibilities.

2. Ancient Latin <oe> was more like [oe̯]. Later it became [e].

[Shihali]

IIRC, Occitan uses the interpunct · to separate vowels in hiatus so they aren't interpreted as a diphthong. There's also the diaeresis (two dots above, like naïve) and, in more recent orthographies and many conlangs including mine, the apostrophe.

So yes, there are lots of methods to indicate non-coalescence.

[Whimemsz]

Ancient Latin <oe> was more like [oe̯]. Later it became [e].

Are you sure about that? My understanding was always that Latin <ae> and <oe> represented [ai] and [oi] in early Latin. But I'll admit this is not my area of expertise.




Incidentally, as for the OP question, Ojibwe (like many other languages which were unwritten until recently) is another example. Sort of. Not as much anymore, actually. The most commonly-used system today is the Fiero or Double-Vowel system, which at this point is pretty close to being the unofficial standard orthography, in that it's what all the major recent dictionaries, grammars, workbooks, etc. etc. use, as well as what's taught in schools, used in official publications on the language, and so on. But there are still people who use an ad hoc English-based orthography. And in some northern communities the syllabary is more common than (or used in addition to) the Latin orthographies.

[Drydic]

Ancient Latin <oe> was more like [oe̯]. Later it became [e].

Are you sure about that? My understanding was always that Latin <ae> and <oe> represented [ai] and [oi] in early Latin. But I'll admit this is not my area of expertise.

Yes, in Old Latin inscriptions we see the diphthongs being written as ai oi, and in Classical times they were written ae oe, to note a different pronunciation from ai oi. Not sure what Feles means by 'Ancient Latin'.

[Whimemsz]

Not sure what Feles means by 'Ancient Latin'.

Maybe, earlier stages of Latin, before <ae> and <oe> became monophthongs?

[Zontas]

Not sure what Feles means by 'Ancient Latin'.

Maybe, earlier stages of Latin, before <ae> and <oe> became monophthongs?

Old Latin? Before Rotacism and Diphthong Simplification?

[Torco]

Mapudungun/Mapuzungun/ has like six different spellings: one used by the mapuche, other used by other mapuche, other by anthropologists and linguists, another by the chilean state [perhaps another accepted by the argentinians as well], and one a mapuche dude came up with, plus another optimized for the interwebz. Plus plenty of variation within those systems.

fifteen century Spanish was ripe with orthographic variety.

[GrinningManiac]

People writing in Hindi on the internet will tend to use English since they either can't be bothered with Devanagari or don't know it.

Hence some really confused spellings. "हूँ" (meaning "am") can be spelt as hoon, hun, hoong, hu, hung, houn, hum, hūm̐, Hūm, hū and a load others. There's also a lot of historical disagreement on the pronunication of वो ("he/she/it") which can vary from Wo to Vah so they're often spelt all over the place.

But it's generally okay.

[zelos]

How on earth is that a problem? That's not even an irregularity!

Maybe it's just me, but when I make con-alphabets, I have a strict "each phoneme gets its own unique letter, no exceptions" policy.

If you skip the whole natural script comming from previous scrips, usually when adopted/made the differens between the script and spoken is rather small for that particular dialect. But as we all know, time has its effect on everything.

[Curlyjimsam]

It's non-contemporary but I'm surprised it hasn't been mentioned - English, of course, did not have any standard orthography from the end of the Old English period up until a few hundred years ago. I suspect other European languages are similar - there was certainly a lot of debate about spelling reform across Europe at about the time of the Renaissance, often with several very different proposals made (e.g. regarding the relative worth of etymological and phonological bases for spelling).

Perhaps it's also easy to overlook the fact that whilst a language like English currently has a standard spelling, it's by no means the case that everybody is capable of using it and there appear to be certain contexts, many of them in the online domain, where it is considered normal not to use standard spellings. I've heard it was commonplace even for educated persons to use non-standard spellings in personal handwritten letters up until the early 1800s.

You've also got cases where the boundary between "language" and "dialect" is fuzzy - so e.g. you could argue that Swedish and Danish orthographies are two competing standards for the same language. There are also cases where the same language is written in more than one script, e.g. Hindi-Urdu (if this is considered one language) or mainland/Taiwanese/pinyin Chinese.

[Torco]

Come to think of it, I don't think any language had a unified orthography before the age of widespread printed materials, and perhaps even electronic communication [the telegraph and stuff].

[zelos]

Learn to spell and let the grown-ups do the talking, Zelos.
Edit: Actually, learn some decent English, for a change.

You're not one to claim that you are a "grown-up" yourself, which should be obvious but clearly isn't to you...

Thank you.

You've also got cases where the boundary between "language" and "dialect" is fuzzy - so e.g. you could argue that Swedish and Danish orthographies are two competing standards for the same language. There are also cases where the same language is written in more than one script, e.g. Hindi-Urdu (if this is considered one language) or mainland/Taiwanese/pinyin Chinese.

Agreed.

It's non-contemporary but I'm surprised it hasn't been mentioned - English, of course, did not have any standard orthography from the end of the Old English period up until a few hundred years ago. I suspect other European languages are similar - there was certainly a lot of debate about spelling reform across Europe at about the time of the Renaissance, often with several very different proposals made (e.g. regarding the relative worth of etymological and phonological bases for spelling).

Yes but One would suspect that no matter which area you went to, if they defined spelling within that region with nothing previous that they would attempt to have consistency where all sounds they heard that are the same are spelled the same. Though this of course falls apart if you got other influences.

[clawgrip]

If we want to get back to the original topic, Japanese, even in its modernized orthography, has ambiguous spelling for /oː/. The standard way to spell it is おう <ou>, but it also occasionally appears as おお <oo> for preserved historical reasons, or when two morphemes come together. However, <ou> sometimes really is pronounced /ou/, usually with verb endings, e.g. 思う /omou/, or when morphemes come together, e.g. 丸の内 (まるのうち) /marunouʨi/.