Latin Orthography Help for ConPhonemes

[Arelvustẃ Zaveʒiyu]

Hello board!

It's been a long time and this is my first time posting on the new board, but I have been an on-off conlanger ever since middle school, though it is only this year that something serious is coming together!

I don't want to be "that guy" and be greedy for help right away but... I have been wrestling with something! I have a pretty well-organized and (finally!) final phonology for my conlang, but am REALLY struggling with some orthographic decisions and I just want some opinions/advice!

My orthography so far:

A - /a/
Á - /ɐu/
B - [ b ]
D - [d]
E - /e/
É - /ʏɐ/
F - /f/
G - [g]
H - /ɦ/
I - /i/
Í - /ɛi/
J - /ʒ/
K - /k/
L - /ɫ/
M - /m/
N - /n/
Ŋ - /ŋ/
O - /ʌ/
Ó - /ɤi/
P - /p/
Q - /ʔ/
R - /r/
S - /s/
T - /t/
U - /ɘ/
Ú - /aɨ/
V - /v/
W - /u/
Ẃ - /yu/
X - /ʃ/
Y - /ʝ/ & [ç]
Z - /z/
Ʒ - /ts/

Vowels are pretty much set; as you can see I have a short-long dichotomy in which the long vowels are really diphthongized, much like English (this has exciting implications for grammar!)

NOW - When it comes to consonants I also have the following phonemes: /tʃ/, /ð/ & [θ], /ɬ/, and /ɮ/. But I am SO lost with how to represent these in a way I'm happy with, so any suggestions are welcome.

[bradrn]

Hi Arelvustẃ! Welcome to the board! (EDIT: I’ve just reread your post and it looks like you’re actually been here before, on the old board — sorry for assuming you were new! Still, welcome.)

First, let me format your phonology and (current) orthography in a more convenient manner:

/m n ŋ/ ⟨m n ŋ⟩
/p t ts tʃ k ʔ/ ⟨p t ʒ ? k q⟩
/b d ɡ/ ⟨b d g⟩
/f θ s ʃ ç~ʝ/ ⟨f ? s x y⟩
/v ð z ʒ ɦ/ ⟨v ? z j h⟩
/ɫ ɬ ɮ/ ⟨l ? ?⟩
/r/ ⟨r⟩

/a e i ʌ ɘ u/ ⟨a e i o u w⟩
/ɐu ʏɐ ɛi ɤi aɨ yu/ ⟨á é í ó ú ẃ⟩

(If I’ve left anything out, please let me know!)

Now, on to your questions. As you can see, I’ve inserted question marks above for all the phonemes you’re not sure of. My suggestions are:

• For /tʃ/, you could use ⟨ǯ⟩, as a parallel to /ts/ ⟨ʒ⟩.
• For /θ ð/ you have several options, but I think that ⟨c ð⟩ would work best.
• For /ɬ ɮ/, ⟨hl dl⟩ are fairly common. If you want to stick to single letters, you can use ⟨ł ľ⟩.

But I do have a few critiques/questions about what you already have:

• ⟨ʒ⟩ for /ts/ seems quite odd to me, but I can’t think of anything better. I personally usually use ⟨c⟩ for /ts/, but ⟨c⟩ is already taken, since I suggested it for /θ/.
• I would write /a e i ʌ ɘ u/ with ⟨a e i o w u⟩, so using ⟨u⟩ for /u/ and writing the central vowel with ⟨w⟩ instead. (This is quite similar to what Welsh does.)
• Why do you write some phonemes as [d] and others as /t/? Usually you would reserve [] for phonetic transcription, and use // for phonemes.

Anyway, if you apply all the suggestions above, you get the following:

/m n ŋ/ ⟨m n ŋ⟩
/p t ts tʃ k ʔ/ ⟨p t ʒ ǯ k q⟩
/b d ɡ/ ⟨b d g⟩
/f θ s ʃ ç~ʝ/ ⟨f c s x y⟩
/v ð z ʒ ɦ/ ⟨v ð z j h⟩
/ɫ ɬ ɮ/ ⟨l ł ľ⟩
/r/ ⟨r⟩

/a e i ʌ ɘ u/ ⟨a e i o w u⟩
/ɐu ʏɐ ɛi ɤi aɨ yu/ ⟨á é í ó ẃ ú⟩

Which I think works pretty well as a romanization!

Oh, and a suggestion: If you run into a problem like this again, try submitting your phonology to the Romanization Challenge Thread; you’ll probably get plenty of suggestions! I’ve done that a couple of times, and it’s helped a lot.

[cedh]

I would write /a e i ʌ ɘ u/ with ⟨a e i o w u⟩, so using ⟨u⟩ for /u/ and writing the central vowel with ⟨w⟩ instead. (This is quite similar to what Welsh does.)

Actually, Welsh uses ⟨u⟩ for /ɨ/ and ⟨w⟩ for /u/, so the OP's scheme is closer to that than your suggestion is.

Here's my own suggestion for the consonants:

/m n ŋ/ ⟨m n ŋ⟩
/p t ts tʃ k ʔ/ ⟨p t c č k q⟩
/b d ɡ/ ⟨b d g⟩
/f θ s ʃ ç~ʝ/ ⟨f θ s š y⟩
/v ð z ʒ ɦ/ ⟨v ð z ž h⟩
/ɫ ɬ ɮ/ ⟨l ł j⟩
/r/ ⟨r⟩

[bradrn]

I would write /a e i ʌ ɘ u/ with ⟨a e i o w u⟩, so using ⟨u⟩ for /u/ and writing the central vowel with ⟨w⟩ instead. (This is quite similar to what Welsh does.)

Actually, Welsh uses ⟨u⟩ for /ɨ/ and ⟨w⟩ for /u/, so the OP's scheme is closer to that than your suggestion is.

Oh, I didn’t know that — thanks for the correction! I really should have checked with a source first…

[holbuzvala]

@Arelvustw

It seems you’re not keen on digraphs. If so, why?

[bradrn]

@Arelvustw

It seems you’re not keen on digraphs. If so, why?

I can’t speak for Arelvustẃ, but I don’t particularly like digraphs either; in particular, I don’t like the way that a digraph like ⟨th⟩ can be confused for the two individual letters ⟨t h⟩. Still, I’m willing to tolerate them when it’s unambiguous.

[Arelvustẃ Zaveʒiyu]

I would write /a e i ʌ ɘ u/ with ⟨a e i o w u⟩, so using ⟨u⟩ for /u/ and writing the central vowel with ⟨w⟩ instead. (This is quite similar to what Welsh does.)

Actually, Welsh uses ⟨u⟩ for /ɨ/ and ⟨w⟩ for /u/, so the OP's scheme is closer to that than your suggestion is.

Oh, I didn’t know that — thanks for the correction! I really should have checked with a source first…

Yes Welsh was actually my inspiration for giving w and u the values I have!

[Arelvustẃ Zaveʒiyu]

@Arelvustw

It seems you’re not keen on digraphs. If so, why?

I can’t speak for Arelvustẃ, but I don’t particularly like digraphs either; in particular, I don’t like the way that a digraph like ⟨th⟩ can be confused for the two individual letters ⟨t h⟩. Still, I’m willing to tolerate them when it’s unambiguous.

So basically my syllable structure only allows one consonant at the end of a syllable/word, and a max of two in a mid-word cluster. Thus, I was hoping to have one letter per phoneme, including affricates, for aesthetic reasons. (Actually, I'm pretty sure this was the ancient greeks' logic when they invented the letters xi and psi! )

Thank you all for so much help with this!! I might check out the romanization thread too.

I think I am actually going to consolidate /ɬ ~ ɮ/ into a single phoneme, thus I will only need one letter for this. Thinking of giving this c but if that's too weird I can always do ł. As of now ʒ "just feels right" for /ts/ - I loved this letter when researching Sami languages, and it has a similar sound there. And ð is feeling right for /θ ~ ð/.

I think I am going to keep x and j for /ʃ/ and /ʒ/ - since these are pretty prominent sounds in my conlang I kind of want them to have "equal dignity" to f/s/v/z as opposed to being any of those with diacritics. And I'm also really liking the look of tx as a digraph for /tʃ/ like Basque/Catalan, and since I was thinking of banning word-final /tʃ/ anyway, this can just be treated as a consonant cluster and thus follow my "rules" (/ts/ meanwhile shall truly be considered one sound, appearing word finally and in at least one possible cluster, so this definitely deserves its own letter).

[Arelvustẃ Zaveʒiyu]

[*] Why do you write some phonemes as [d] and others as /t/? Usually you would reserve [] for phonetic transcription, and use // for phonemes.
[/list]

[ b d g ] are word-final and voiced consonant cluster allophones of /p t k/ - it seems too "weird" to me not to assign them their own letters! But thank you for your thoughtful consideration of my work thus far!

[bradrn]

As of now ʒ "just feels right" for /ts/ - I loved this letter when researching Sami languages, and it has a similar sound there.

I just looked it up, and it looks like Sami and Laz both use ⟨ʒ⟩ for /d͡z/. This is new to me — I always thought it was used for /ʒ/! So thanks for teaching me this.

[ b d g ] are word-final and voiced consonant cluster allophones of /p t k/ - it seems too "weird" to me not to assign them their own letters!

If they’re merely allophones, then to me, it seems very weird to not give them their own letters! But obviously this is a matter of personal preference.

But thank you for your thoughtful consideration of my work thus far!

You’re welcome!

[Pabappa]

There's precedence for giving allophones separate letters in the orthography, but it seems to happen most often in non-native scripts, e.g. the Hepburn romanization of Japanese, and Rotokas getting an s just to write the allophone of /t/ that appears before /i/. Sanskrit did it too with its nasals, giving /ŋ ɲ/ their own consonants even though /ŋ/ was entirely predictable and i think /ɲ/ may also have been. Greek always marked aspiration on rho, even though that, too, was entirely predictable. So you can go either way. That said, I think the trend is towards phonemic orthographies .... e.g. the many languages of Australia that use only voiced stop letters or only voiceless ones.

[Richard W]

There's precedence for giving allophones separate letters in the orthography, but it seems to happen most often in non-native scripts, e.g. the Hepburn romanization of Japanese, and Rotokas getting an s just to write the allophone of /t/ that appears before /i/. Sanskrit did it too with its nasals, giving /ŋ ɲ/ their own consonants even though /ŋ/ was entirely predictable and i think /ɲ/ may also have been. Greek always marked aspiration on rho, even though that, too, was entirely predictable. So you can go either way. That said, I think the trend is towards phonemic orthographies .... e.g. the many languages of Australia that use only voiced stop letters or only voiceless ones.

Sanskrit has word-final /ŋ/ through the rule against word-final consonant clusters. The writing of /ɲ/ probably arises because writing is not native to Sanskrit, but borrowed from the Prakrits, which have phonemic /ɲ/. (<ñ> is a word-starting onset in Pali.) Vedic Sanskrit has separate symbols for two allophones of visarga.

In Greek, was aspiration on rho ever written using the letter eta? (Heta in some transcriptions.) Writing aspiration on rho served as word-boundary marker (except on double rho) , so may have had a contrastive phonetic correlate. Compare the two forms of lower case sigma.

[Arelvustẃ Zaveʒiyu]

There's precedence for giving allophones separate letters in the orthography, but it seems to happen most often in non-native scripts, e.g. the Hepburn romanization of Japanese, and Rotokas getting an s just to write the allophone of /t/ that appears before /i/. Sanskrit did it too with its nasals, giving /ŋ ɲ/ their own consonants even though /ŋ/ was entirely predictable and i think /ɲ/ may also have been. Greek always marked aspiration on rho, even though that, too, was entirely predictable. So you can go either way. That said, I think the trend is towards phonemic orthographies .... e.g. the many languages of Australia that use only voiced stop letters or only voiceless ones.

So basically my logic with the stops was to 1) make consonant clusters more clear with regard to voicing and 2) for my own english-oriented pronunciation, but thinking about this is making me realize the extra letters probably aren’t necessary and just add unnecessary complexity. After all, I’m already not distinguishing for voiceless allophones of /ð ɮ ʝ ɦ/ǃ Australian languages are a great comparison since my stops are dealt with a lot like those!

I think I’ll use p, t, k instead of b, d, g - my main means of practicing/testing/sorting out phonotactics and orthography so far has been translating people’s names in my Snapchat contacts and in doing this, I’ve usually assimilated English /pʰ tʰ kʰ/ to /p t k/ and English /b d g/ to nasals /m n ŋ/ or lateral combos /dl gl/ for d and g or /v/ for b.

Now, this actually leaves me with some extra letters to use! - maybe d for /ɮ/ and b for /ð/ (using b in this way is SUPER weird, I know, but might actually work since the way I speak many of my labials are actually labiodentals in practice, and I’ve been thinking of /ð/ more as a companion to those than the other coronals anyway). And c and g can be alternatives to ʒ and ŋ if I were to so choose or require.

Still thinking this through but having fun brainstorming!

[bradrn]

Now, this actually leaves me with some extra letters to use! - maybe d for /ɮ/ and b for /ð/ (using b in this way is SUPER weird, I know, but might actually work since the way I speak many of my labials are actually labiodentals in practice, and I’ve been thinking of /ð/ more as a companion to those than the other coronals anyway). And c and g can be alternatives to ʒ and ŋ if I were to so choose or require.

⟨d⟩ for /ɮ/ sounds fine — for a precedent, Marshallese uses ⟨d⟩ for /rʲ/, which is fairly similar. ⟨g⟩ for /ŋ/ is also attested, in Fijian. ⟨c⟩ for /ʒ/ is a bit more weird, but Berber uses ⟨c⟩ for /ʃ/, which is very similar. However, ⟨b⟩ for /ð/ seems to me to be much too weird to be reasonable, even with the labial/labiodental alternation you mentioned. I would recommend using ⟨ð⟩ instead, as a more reasonable alternative.

(Also, I’d be interested to know: do you also say /p/ as a labiodental? Or just the fricative /f/?)

[Xwtek]

I think this should be on Romanization Challenge Thread v2.0. Also, when asking the romanization, tell us the phonotactics too. For example, it's a good practice to use <ts> as the affricate, unless if that language has phonotactics like Polish, where /ts/ can be either consonant cluster or affricate.

[Birdlang]

My take on the consonants
I kept the original ones the same, but adjusted the sounds to a way I thought looked cool.
/m n ŋ/ ⟨m n ŋ⟩
/p t ts tʃ k ʔ/ ⟨p t ʒ ǯ k q⟩
/b d ɡ/ ⟨b d g⟩
/f θ s ʃ ç~ʝ/ ⟨f þ s x y⟩
/v ð z ʒ ɦ/ ⟨v ð z j h⟩
/ɫ ɬ ɮ/ ⟨l ś ź⟩
/r/ ⟨r⟩

Is this ok OP? It fits the no digraphs romanization.
If you want to make the lateral fricatives one sound, I’ve always used the character ŀ or ł or ś (last from Semitic).
Also for the lateral, I also think you could use d, for the /ʒ ŋ/ you can use c g for /ʃ/ ç to mirror with Turkish.
Edit: changed since two characters were too similar.

[Travis B.]

⟨ƚ ɫ⟩ seem too similar to me.

[bradrn]

⟨ƚ ɫ⟩ seem too similar to me.

What’s this in reply to? None of the orthographies proposed so far have both of ⟨ƚ ɫ⟩ in them.

[Travis B.]

⟨ƚ ɫ⟩ seem too similar to me.

What’s this in reply to? None of the orthographies proposed so far have both of ⟨ƚ ɫ⟩ in them.

Oh Birdlang changed his orthography so it now uses ⟨ś ź⟩ for /ɬ ɮ/.

[bradrn]

⟨ƚ ɫ⟩ seem too similar to me.

What’s this in reply to? None of the orthographies proposed so far have both of ⟨ƚ ɫ⟩ in them.

Oh Birdlang changed his orthography so it now uses ⟨ś ź⟩ for /ɬ ɮ/.

Thanks for clarifying! It looks like Birdlang edited it just before I looked at his reply.