If natlangs were conlangs

[Whimemsz]

I will say I don't actually think it's that weird a convention. A voiced velar fricative is vaaaaaaguely-ish similar to [e] in contact with another vowel (not as similar as [a], but whatever -- /e/ only occurs in loanwords in Aklanon while /a/ occurs in native words, which may have been part of the original thinking), and this is a useful way of avoiding having to use more characters than necessary in the orthography (<eV> and <Ve> always represent the fricative, <eC> and <Ce> always represent the vowel, so there isn't ambiguity). This also suggests it's closer to an approximant than a fricative ("a voiced-velar spirant (with only some friction)"), which would further help explain the use of <e>: a velar approximant is somewhat closer perceptually and articulatorily to [e] (or [ɤ], see below) next to another vowel.

I do note with amusement that Wikipedia contradicts itself within this very short article and at the top says "Its unique feature among other Bisayan languages is the close-mid back unrounded vowel [ɤ] occurring as part of diphthongs and traditionally written with the letter E such as in the name Akeanon (Aklanon)." Which at least the source it cites shows is wrong, someone just skimming it or something evidently somehow confused the modified gamma used in the IPA for the voiced velar fricative as the rams' horns, even though the source (a Peace Corps manual) doesn't use the IPA...!! This could have been compounded by the just-noted fact of the perceptual/articulatory similarity of [ɰ] and [ɤ]/_V_. (Alternately, maybe it really is pronounced [ɤ] and was poorly described in older works, but in that case Wikipedia is failing to accurately cite a source for its claim.) Anyway, just another reminder *not to pay attention to Wikipedia* except as a way of finding primary sources that you then use to get actual, more reliable information.

[bradrn]

Regarding Aklanon, see https://en.wikipedia.org/wiki/Aklanon_l ... #Phonology where they mention the orthography.

That was where I originally noticed the weird orthography.

[Richard W]

Regarding Aklanon, see https://en.wikipedia.org/wiki/Aklanon_l ... #Phonology where they mention the orthography.

But the second sentence of the article says <e> represents a vowel.

[bradrn]

Regarding Aklanon, see https://en.wikipedia.org/wiki/Aklanon_l ... #Phonology where they mention the orthography.

But the second sentence of the article says <e> represents a vowel.

That second sentence is wrong — it should be [ɣ] rather than [ɤ], and then it was made worse by including the phonetic name of the already-miswritten IPA letter.

[Richard W]

Regarding Aklanon, see https://en.wikipedia.org/wiki/Aklanon_l ... #Phonology where they mention the orthography.

But the second sentence of the article says <e> represents a vowel.

That second sentence is wrong — it should be [ɣ] rather than [ɤ], and then it was made worse by including the phonetic name of the already-miswritten IPA letter.

As we have established. The point is that that sentence destroys the credibility of the article.

[bradrn]

But the second sentence of the article says <e> represents a vowel.

That second sentence is wrong — it should be [ɣ] rather than [ɤ], and then it was made worse by including the phonetic name of the already-miswritten IPA letter.

As we have established. The point is that that sentence destroys the credibility of the article.

Sorry! I think I misinterpreted what you were originally trying to say. I agree that this completely destroys the credibility of that article.

[Birdlang]

What about some natlangs that have completely off the wall orthographies?
I know Rohingya has ç for /ɽ/.
Fijian has c for /ð/.
And there is Hmong.

[bradrn]

What about some natlangs that have completely off the wall orthographies?
I know Rohingya has ç for /ɽ/.
Fijian has c for /ð/.
And there is Hmong.

The very worst one is Saanich (a.k.a. SENĆOŦEN):

That's a standard Salishan orthography; it only occurs in <t̓ᶿ> because there's no /tθ/ in the language, only /tθ’/. The Salishan language with the actually weird (aka stupid) orthography is Saanish (SENĆOŦEN), which (almost) only uses capital letters.

Oh yes, I completely forgot about Saanich! That has to be the weirdest Latin-script orthography I know of! For those who aren’t acquainted with the sheer bizarreness (is that a word?) of Saanich orthography (a.k.a. SENĆOŦEN), here are some details:

• All capital letters, except <-s> for some reason
• Stroked letters ȺȻꝀȽŦȾ (yes, T is stroked two different ways!)
• Comma for glottal stop (why? good question)
• Acute accent used with A/Á, C/Ć, K/Ḱ, S/Ś for no apparent reason (e.g. A and Á are the same, except the latter is used after post-velar consonants (what sort of language distinguishes palatal, pre- and postvelar?)) EDIT: palatal/pre/postvelar turns out to be the Americanist terminology for palatal/velar/uvular, which is indeed somewhat common.

Resulting in the following easy-to-read text:

SI,SI,OB BE₭OȻBIX̲ ,UQEȾ. ,ESZUW̲IL ELQE,. ,ESTOLX ELQE, ESDUQUD ,ESXEĆBID ȽṮUBEX̲ ELQE, ŚÍISȽ ,ÁL,ÁLŦ.

(That weird triangle X thing should actually be X with line below.)

Oh, and according to Wikipedia, Saanich uses regular metathesis for aspect. All in all, it makes the rest of Salishan look positively sane… which is an impressive achievement. (I do wonder sometimes why Salishan got all the crazy stuff… extreme polysynthesis, weird orthography, nounlessness, occasional vowellessness…)

In the ensuing discussion, we also complained about Tlingit (how does using ⟨.⟩ for the glottal stop make sense‽), Zhuang (just as bad as Hmong), Gwoyeu Romatzyh (a truly awful romanization system for Standard Mandarin which respells the vowels depending on tone) and Mediaeval Arpitan (with its word-final stress letters). And elsewhere in this thread I’ve also complained about Kiowa (which uses ⟨f j ch c⟩ for /p t ts k/) and Northern Qiang, which (in)famously uses ⟨v⟩ for /χ/.

[Linguoboy]

I've belatedly come to appreciate Gwoyeu Romatzyh. By far my #1 issue with Pinyin is that I'll remember the initial, the final, and forget the tone. Since I also sometimes have an issue with forgetting accents in French or umlauts in German (despite how these affect the pronunciation), I ascribe this to the fact that I was raised on an orthography without diacritics. So I'm very sensitised to the letterforms in an orthographic word but I've never gotten over the mindset that diacritics are a sort of extra decoration.

All this is to say that I would've had an easier time memorising shan, sharn, shaan, shann and associating each with a different character than I do with shān, shán, shǎn, shàn--despite how straightforwardly iconic in nature these diacritics are. Sure, it makes for a deeper orthography, but I've learned Irish for Chrissakes, that doesn't frighten me (and, once again, my biggest issue there is remembering when a vowel takes the síneadh fada[*] and not when a consonant is broad or slender).

Sadly, this ship has sailed; all my learning has been with Pinyin or Bopomofo (which uses the same tone diacritics and thus runs into the same issues). Gwoyeu Romatzyh learning materials aren't easy to find anyway (though the Taiwanese dictionary I've used for most of my life includes GR in its entries and in its indices). Still, I don't see it as an inherently unsound system, especially in the more consistent form it's been implemented for Hmong, Zhuang, et al.

[*] As if to prove my point, I had to look up síneadh just now because I was 90% the first vowel was long but I wasn't certain.

[bradrn]

I've belatedly come to appreciate Gwoyeu Romatzyh. By far my #1 issue with Pinyin is that I'll remember the initial, the final, and forget the tone. Since I also sometimes have an issue with forgetting accents in French or umlauts in German (despite how these affect the pronunciation), I ascribe this to the fact that I was raised on an orthography without diacritics. So I'm very sensitised to the letterforms in an orthographic word but I've never gotten over the mindset that diacritics are a sort of extra decoration.

All this is to say that I would've had an easier time memorising shan, sharn, shaan, shann and associating each with a different character than I do with shān, shán, shǎn, shàn--despite how straightforwardly iconic in nature these diacritics are. Sure, it makes for a deeper orthography, but I've learned Irish for Chrissakes, that doesn't frighten me (and, once again, my biggest issue there is remembering when a vowel takes the síneadh fada[*] and not when a consonant is broad or slender).

Interesting! I’m absolutely fine with diacritics, as far as I can tell, but I have never been able to figure out Irish or other languages with irregular spelling (like French). My biggest issue with Pinyin is not the diacritics, but the spelling issue remembering which of ⟨zh j ch q sh x⟩ is which. (Although admittedly I’ve never seriously tried to learn Standard Mandarin, so potentially the diacritics could prove to be a bigger issue if I were to try.)

[Linguoboy]

My biggest issue with Pinyin is not the diacritics, but the spelling issue remembering which of ⟨zh j ch q sh x⟩ is which. (Although admittedly I’ve never seriously tried to learn Standard Mandarin, so potentially the diacritics could prove to be a bigger issue if I were to try.)

I guess because you pronounce both as their nearest English equivalents rather than pronouncing one series as palatal and the other as retroflex? It helped me to learn that native speakers are far more likely to confuse zh with z than with j (and so forth).

French orthography is more consistent than English, and Irish orthography is far more consistent than either (even when you take into account considerable dialect variation). But the last of these only makes sense if you understand the phonological structure (which, honestly, is how it should be; orthographies serve speakers, not foreign learners).

[mèþru]

Pinyin may not be my style orthographically but I think it is a really well made system.

[Linguoboy]

Pinyin may not be my style orthographically but I think it is a really well made system.

It took me a while to get over using a symbol I associate with uvular sounds and labialised velars (q) for a palatal affricate, but once I managed that I was golden. And initial x is undeniably cool.

[zompist]

which, honestly, is how it should be; orthographies serve speakers, not foreign learners.

Amen. I rather like pinyin, not least because the consonants, at least, are well designed. Bopomofo handles the vowels better, but pinyin isn't too awful there.

You may remember back on sci.lang in the 1990s several people criticized pinyin basically because it was invented by commies.

[Birdlang]

I like pinyin too.
The only problem is I associate the macron with vowel length usually, not high tone.
I spell the /ɕ/ sound in one Birdish dialect with x because it is like Pinyin.
Also, there are some inconsistencies in the Manx, Irish, and Scottish Gaelic alphabets which I’m not a fan of, mostly because of the multiple spellings for one sound.
There are some languages with weird spellings though that seem to be just as weird as Aklanon like Drehu and some other New Caledonian and even some Papuan languages.

[Linguoboy]

Also, there are some inconsistencies in the Manx, Irish, and Scottish Gaelic alphabets which I’m not a fan of, mostly because of the multiple spellings for one sound.

That's not something you can fully get away from with an interdialectal orthography. Of course, the logical alternative is an orthography strictly based on a single dialect (such as Ó Cuív's proposal for Irish), but I don't think that's desirable for endangered varieties.

[Pabappa]

I use an orthography for tones in my conlangs that might seem like i chose to be perverse on purpose.
the short high tone is marked by a grave accent (à),
a long high tone, which is sometimes followed by downstep, is marked by a macron (ā),
a long high tone that is not followed by downstep is marked by acute (á),
a short mid tone is marked by breve (ă), and
a long, low, pharyngealized tone is marked by circumflex (â).

the logic behind the grave is that this tone is marked by a following glottal stop in most of the languages that have it, and i see the grave as illustrating how the syllable is "chopped", or checked, by the glottal stop.

I wonder if some similar thought process may have happened in Lithuanian, because Lithuanian also uses the grave for a short high tone, at least in dictionaries. I believe all three of its tones are high tones ... essentially the tonal pattern of the stressed syllable can be H, HL, or LH. Thus it makes sense to assign the grave to at least something.

Oh, the pinyin tone marks look perfect to me .... I didnt realize it was that simple until now. By contrast with something like Vietnamese, youve got to make tough decisions, and its possible even that the tone inventory has changed in just the short period of time since we chose the tone markers, so what looks like a mess right now might have been the best option at the time.

[Kuchigakatai]

The very worst one is Saanich (a.k.a. SENĆOŦEN)

It appears no one in this thread or the thread where you first posted that (the Conlang Random Thread) mentioned it, but in case anyone doesn't know, Saanich got this orthography because most of the documentation is derived from a huge corpus that was typed out by one native speaker, Dave Elliott. He felt that good, fluent Saanich was about to die soon, so in an attempt to preserve what he had in his head, he grabbed a typewriter and began putting what he could on paper. Knowing basically nothing but English and Saanich, he adapted English orthography/ies to write his other language.


He disliked the ambiguity of digraphs because separating them may not always be obvious, so he chose to render each sound he felt he was making with one letter. Typewriters allow you to go back to the previous character to add a diacritic on top, so that's the technique he used to make up new letters.

He assigned the vowel symbols based on the conventions of macrons in both English bibles (when it comes to difficult names*) and North American dictionaries like the American Heritage Dictionary and the Merriam-Webster Dictionary, where macrons are used for the traditional "long" vowels, e.g. <ā> for /eɪ/, using respellings like <sā’-shē-āt> for "(to) satiate". It seems his typewriter didn't have the ability to type a macron though, so he replaced it with an acute accent. (Other diacritics he used would be the superimposable underline, a superimposed hyphen, and a superimposed slash "/".)

He did this depending on how he perceived his own vowels. He felt that Saanich /e/ was "like English /eɪ/" (note how many English speakers learning Spanish do the same), so he wrote it by adapting biblical/dictionary <ā> as <Á>. However, for the phonetically longer /ej e:j/, he used <Ⱥ>, basically by thinking of the superimposed slash "/" as a "longer macron" than the macron-turned-acute-accent of <Á>. Same goes for /i/ <I> and /ǝj aj a:j/ <Í>. He apparently had a rounded and somewhat high allophone of /ǝ/ in /ǝw/ ([ʊw]?), which explains why it's written <U> (the same as the uncommon /u/, mostly found in borrowings) and not <Ó> (i.e. adapting biblical/dictionary <ō> /oʊ/). The exceptional use of <A> for /e/ next to a uvular is due to him perceiving his allophone there (which linguists describe as [ɜ] or [ɛ]) "like English /æ/".

The consonants are just modifications of similar English sounds, quite naturally, even though the diacritics are not used for phonetic features in a consistent way aside from <Ć Ś> /tʃ ʃ/ (which he might have seen in newspapers for Polish or Serbo-Croatian or something). The comma is actually a decent choice for the glottal stop given an ignorance of normal linguistic notation, if you think of a glottal stop as being a "pause", especially when you hear it between vowels. Compare this with the name "saltillo" ("little jump") for the Nahuatl glottal stop sound in coda position (/tɬeʔko/ <tlehco tle'co tlèco tleco> 'in a fire').


* Although dictionaries tend to use respellings, bibles tend to just put breves/macrons on top plus apostrophes for stress, as if they were ambiguous Hebrew vowel marks. I've seen KJV copies where e.g. Nebuchadnezzar appears as Nĕ’bŭchadnĕ’zzar, or Jeremiah as Jĕremī’ah. If you ever see a newbie using an apostrophe after a vowel to mark stress, as someone did just a couple days ago on the CBB, now you know where that comes from!

[Pabappa]

I used to think that the diacritic overload on those names was real, and that e.g. Hebrew really had 20 vowels, 40 consonants, etc. I made some very early conlang sketches with huge phonologies, figuring that if they could talk like that, so could I. But I was young and I didnt pursue any of these projects to the point where I had a definite phonology ... I just threw diacritics around.

[bradrn]

My biggest issue with Pinyin is not the diacritics, but the spelling issue remembering which of ⟨zh j ch q sh x⟩ is which. (Although admittedly I’ve never seriously tried to learn Standard Mandarin, so potentially the diacritics could prove to be a bigger issue if I were to try.)

I guess because you pronounce both as their nearest English equivalents rather than pronouncing one series as palatal and the other as retroflex? It helped me to learn that native speakers are far more likely to confuse zh with z than with j (and so forth).

No; I just keep on forgetting which one of ⟨zh j ch q sh x⟩ is palatal and which one is retroflex. (Yes, I know that all the digraphs are retroflex and all the single letters are palatal, but I still keep on getting those two sets mixed up.)

French orthography is more consistent than English, and Irish orthography is far more consistent than either (even when you take into account considerable dialect variation). But the last of these only makes sense if you understand the phonological structure (which, honestly, is how it should be; orthographies serve speakers, not foreign learners).

I know that they’re more consistent than English, but that has never seemed to help when I’ve tried to figure it out… Maybe it’s because I don’t know the languages, and if I had a better understanding of them the orthographies would be easier to understand.

The very worst one is Saanich (a.k.a. SENĆOŦEN)

It appears no one in this thread or the thread where you first posted that (the Conlang Random Thread) mentioned it, but in case anyone doesn't know …

Thanks for the excellent explanation Ser! I was aware that it was devised by a native speaker, but I didn’t know the details. The orthography does make a lot more sense now that you’ve explained it.

(Of course, that doesn’t take away from the fact that SENĆOŦEN is still very, very weird by the standards of Latin-script romanizations…)