Romanization challenge thread

[smii]

Consonants:
/pʰ~p͡xʰ p~p͡x pʰʲ~pʰʲj pʲ~pʲj b~bə bʲ~bʲj tʰ~t͡xʰ t~t͡x tʰʲ~tʰʲj tʲ~tʲj d~də dʲ~dʲj kʰ~k͡xʰ k~k͡x kʰʲ~kʰʲj g~g͡ɣ c͡ç~c ɟ͡ʝ~ɟ ʔ/
<ph p p' pj b bj th t t' tj d dj kh k k' g kj gj h>
/m~mə mʲ~mʲj n~nə nʲ~nʲj~ɲ ŋ/ <m mj n nj ŋ>
/ɾ~ɾə ɹ~ɹj~ʐ̺~z~ə/ <r rr>
/t͡sʰː~t͡sʰːx t͡s~t͡sx d͡z t͡sʰʲː~t͡sʰʲːj t͡sʲ~t͡sʲj t͡ʂ̺ʰː t͡ʂ̺ d͡ʐ̺ t͡ɕ̻ʰː t͡ɕ̻ d͡ʑ̻/ <tsh ts dz ts' tsj tśh tś dź tch tc dq>
/f~f͡x v~və fʲ~fʲj vʲ~vʲj s~s͡x z sʲ~sʲj ʂ̺ ʐ̺ ɕ̻ ʑ̻ x ɣ/ <f v fj vj s z sj ś ź zh c q hh x>
/j w/ <j w>
/lʲ~lʲj~ʎ ɫ~ɫ͡ə/ <lj l>
Vowels:
/ɪ u e ə o ɐ/ <i u e y o a>
/i: u: e: ə: o: ä:/ <ii uu ee yy oo aa> (length reversed at the beginning of a word)
/ɪ̃ ĩ: ũ ũ: ɛ̃ ɛ̃: ə̃˩ ə̃:˩ ɔ̃ ɔ̃: ɐ̃ ä̃:/ <í íí ú úú ...>
/ɪ˩ u˩ e˩ ə˩ o˩ ɐ˩/ <ih uh yh...>
/i:˩ u:˩ e:˩ ə:˩ o:˩ ä:˩/ <iih uuh yyh ...>
/ɪ̃˩ ĩ:˩ ũ˩ ũ˩: ɛ̃˩ ɛ̃˩: ɔ̃˩ ɔ̃˩: ɐ̃˩ ä̃˩:/ <íh íí úh úúh ...>
An actual sentence:
/fe:nʲɪʐ̺ɐ̃ ʂ̺ä:jɪŋ i:ʑ̻lʲjeʔ i:ɫozɐ̃ ʂ̺ä:jɪŋ tä:lʲkʲəf, o:˩ i:ʑ̻ t͡sʰːxi:dʲjũ˩: lɔ̃ɣɐʂ̺/
<Fenjiźá śajiŋ iqljeh ilozá śajiŋ taljkjyf, oh iź tshidjúúh lóxaś.>

[HazelFiver]

I don't have any good ideas (I think they're all taken), but I figured I should post a romanization before putting my own phonology.

Consonants:
/pʰ~p͡xʰ p~p͡x b~bə pʰʲ~pʰʲj pʲ~pʲj bʲ~bʲj tʰ~t͡xʰ t~t͡x d~də tʰʲ~tʰʲj tʲ~tʲj dʲ~dʲj kʰ~k͡xʰ k~k͡x g~g͡ɣ kʰʲ~kʰʲj c͡ç~c ɟ͡ʝ~ɟ ʔ/ <ph p b phy py by th t d thy ty dy kh k g khy ky gy '>
/m~mə mʲ~mʲj n~nə nʲ~nʲj~ɲ ŋ/ <m my n ny ng>
/ɾ~ɾə ɹ~ɹj~ʐ̺~z~ə/ <r zh>
/t͡sʰː~t͡sʰːx t͡s~t͡sx d͡z t͡sʰʲː~t͡sʰʲːj t͡sʲ~t͡sʲj t͡ʂ̺ʰː t͡ʂ̺ d͡ʐ̺ t͡ɕ̻ʰː t͡ɕ̻ d͡ʑ̻/ <tsh ts dz tshy tsy dzy txh tx dq tch tc dj>
/f~f͡x v~və fʲ~fʲj vʲ~vʲj s~s͡x z sʲ~sʲj ʂ̺ ʐ̺ ɕ̻ ʑ̻ x ɣ / <f v fy vy s z sy x q ch j c'h gh>
/j w/ <y w>
/lʲ~lʲj~ʎ ɫ~ɫ͡ə/ <ly l>


Vowels:
/ɪ u e ə o ɐ/ <i ou e u o a>
/i: u: e: ə: o: ä:/ <í oú é ú ó á>
/ɪ̃ ĩ: ũ ũ: ɛ̃ ɛ̃: ə̃˩ ə̃:˩ ɔ̃ ɔ̃: ɐ̃ ä̃:/ <iñ íñ ouñ oúñ eñ éñ uñ úñ oñ óñ añ áñ>
/ɪ˩ u˩ e˩ ə˩ o˩ ɐ˩/ <ì où è ù ò à>
/i:˩ u:˩ e:˩ ə:˩ o:˩ ä:˩/ <î oû ê û ô â>
/ɪ̃˩ ĩ:˩ ũ˩ ũ˩: ɛ̃˩ ɛ̃˩: ɔ̃˩ ɔ̃˩: ɐ̃˩ ä̃˩:/ <ìñ îñ oùñ oûñ èñ êñ òñ ôñ àñ âñ>

Nasal vowels and vowel + nasal sequences contrast, geminated consonants are common. Palatalised consonant and consonant + /j/ sequences never contrast, just like non-palatalised consonant and consonant + /x/ sequences - same as on Ilos. Also, all first vowels in a word are always long except in recent borrowings and /ɣ/ never occurs anywhere except before a vowel.

An actual sentence:
/fe:nʲɪʐ̺ɐ̃ ʂ̺ä:jɪŋ i:ʑ̻lʲjeʔ i:ɫozɐ̃ ʂ̺ä:jɪŋ tä:lʲkʲəf, o:˩ i:ʑ̻ t͡sʰːxi:dʲjũ˩: lɔ̃ɣɐʂ̺/
Fényiqañ xáying íjlye' ílozañ xáying tálykyuf, ô íj tshídyoûñ loñghax.
(This says something about Phoenician and Ilosean Shayana, doesn't it? I can't tell what, though.)

[mèþru]

I can make out the following:
/fe:nʲɪʐ̺ɐ̃ ʂ̺ä:jɪŋ i:ʑ̻lʲjeʔ i:ɫozɐ̃ ʂ̺ä:jɪŋ tä:lʲkʲəf, o:˩ i:ʑ̻ t͡sʰːxi:dʲjũ˩: lɔ̃ɣɐʂ̺/
Phoenician Shayana ? Ilosean Shayana ? all is ? ?.

[HazelFiver]

I'm considering making some changes to the phonology of my main project, but I haven't decided how I should spell the result, so I thought I'd post it here to see if anyone had what suggestions. (Yes, it's boring and SAE-ish. Don't worry, though, I know what I'm doing.)

/m n ŋ/
/p b t d k g tʼ kʼ/
/f v θ ð s z ɬ ʃ ʒ χ* h/
/l r j w/
/a aː e eː i iː o oː u uː ə/

*This is actually a chi.

(C)(C)V(C)(C) -- don't think I need to get into the details of which clusters are permissible. Any consonant except the plosives can be syllabic, but these don't contrast with their non-syllabic counterparts often, if at all. Geminated consonants do not exist except across morpheme boundaries. /i u/ and /j w/ don't contrast often but this is partly because the current orthography doesn't distinguish /i/ from /j/ and originally didn't distinguish /u/ from /w/ either (I know this sounds backwards, but I don't mean this to be diachronic). (Is this too much or too little information?)

More-or-less real sentences:
/tabir uːz maːr pfitʼ ve kairdið ʃo fuz ‖ biːr pkəi za χeːŋ za foːn fo parθot veu ʃeɬ za fəln sl̩ ksau ʒav fn̩kʼ ‖ ɬaʃir tʼep pʒ̍t ɬ̩ kʼfou/

[mèþru]

Good information. If I were to write it, I'd add if geminated vowels contrast with two of the same short vowel in hiatus. Most members who post such phonologies don't do that and I assume that there is no contrast. If there is a specific style you want or things for people to avoid, you should include that in background information.
<m n ŋ>
<p b t d k g t' k'>
<f v þ ð s z ł š ž x h>
<l r j v>
<a aa e ee i ii o oo u uu ĕ>

Tabir uuz maar pfit' ve kairdið šo fuz. Biir pkĕi za xeŋ za foon fo parþot veu šeɬ za fĕln sl ksau žav fnk'. Łašir t'ep pžt ł k'fou.

agefaqeg-style:
<m n ŋ>
<p b t d k g ṭ ḳ>
<f v þ ð s z ḷ š ž kh h>
<l r̀ r w>
<a aa e ee i ii o oo u uu ê>

tabir̀ uuz maar̀ pfiṭ ve kair̀dið šo fuz, biir̀ pkêi za kheŋ za foon fo par̀þot veu šeɬ za fêln sl ksau žav fnḳ, lašir̀ ṭep pžt ł ḳfou,

[opipik]

/m n ŋ/ <m n ng>
/p b t d k g tʼ kʼ/ <p b t d k g t' k'>
/f v θ ð s z ɬ ʃ ʒ χ* h/ <f v th dh s z lh sy zy x h>
/l r j w/ <l r y w>
/lh sj zj/ <l'h s'y z'y>

/a aː e eː i iː o oː u uː ə/ <a ä~a i e iy~yi ey~ye u o uw~wu~w ow~wo Ø>
/ei ou/ <ay aw~au~u>


*This is actually a chi.

More-or-less real sentences:
/tabir uːz maːr pfitʼ ve kairdið ʃo fuz ‖ biːr pkəi za χeːŋ za foːn fo parθot veu ʃeɬ za fəln sl̩ ksau ʒav fn̩kʼ ‖ ɬaʃir tʼep pʒ̍t ɬ̩ kʼfou/
Tabir woz mär pfit' vi kayirdyidh syu fwuz. Beyr pky za xeng za fon fu parthut viwu syilh za fln sl ksawu zyav fnk'. Lhasyir t'ip pzyt lh k'fu.

[Knit Tie]

This says something about Phoenician and Ilosean Shayana, doesn't it? I can't tell what, though.

It says "Phoenician Shayana isn't a dialect of Ilosean Shayana, but instead is a distinct language" with a very alpha-version grammar.

[HazelFiver]

Good information. If I were to write it, I'd add if geminated vowels contrast with two of the same short vowel in hiatus. Most members who post such phonologies don't do that and I assume that there is no contrast.

Ah, sorry. There is indeed no contrast -- two of the same vowel in hiatus aren't allowed. I should also probably have mentioned that /mh nh lh/ are the only allowable /h/ clusters (there aren't any words with those sequences yet, but nonetheless this means that say <th dh> are ok digraphs but <lh> probably isn't) and that /sj zj/ are allowed. This sounds like I made it up just now to spite opipik, but I swear I didn't.

If there is a specific style you want or things for people to avoid, you should include that in background information.

The style isn't important to me, and I don't want to seem too picky (I unfortunately don't like any of the usual ways of writing /ŋ/ and ejectives, which is why /ŋ/ and /ng/ conveniently don't contrast and I've been tentatively writing /tʼ kʼ/ as <tz k>. Maybe I'm too closed-minded.)

[mèþru]

<tz> means either /tz/ or /t͡s/. How do/did you distinguish between /k/ <kʼ>. Also, /ŋ/ <ŋ> is not really a common romanisation on this board. It's just my preference.

[HazelFiver]

<tz> means either /tz/ or /t͡s/.

Yes, that's the problem with it, though it isn't ambiguous because /ts/ is <ts> and /tz/ isn't allowed because of voicing assimilation. Nonetheless it looks pretty stupid. I suppose I would have to go with apostrophes.

How do/did you distinguish between /k/ <kʼ>.

/k/ is <c>.

[Chengjiang]

I'm considering making some changes to the phonology of my main project, but I haven't decided how I should spell the result, so I thought I'd post it here to see if anyone had what suggestions. (Yes, it's boring and SAE-ish. Don't worry, though, I know what I'm doing.)

Ejectives are a pretty major non-SAE feature; none of the language groups traditionally considered to contribute to the SAE archetype have phonemic (or, in most cases, even phonetic) ejectives. The only thing I'd call SAE about this is the largish fricative system, and even then while /θ ð/ are found in English and a few other European languages, they're rare enough in European languages (as they are in most other groups) that I don't think they can properly be called an SAE trait. Relatedly, /ɬ/ is actually much rarer in European languages than it is in, say, indigenous North American languages. Also, as far as phonotactics goes, having syllabic fricatives is extremely non-SAE. So I wouldn't actually call this especially SAE, if that's something you're at all concerned about one way or another.

Here's the sort of thing I'd probably do if this were my language. It uses several digraphs, and while it's pretty intuitive to, say, an English-speaking audience, it might not be to your taste. Some choices reflect your saying that the syllabic:non-syllabic contrast is rare, even for close vowels versus glides, and that geminates do not exist except across morpheme boundaries. I have <ll> for /ɬ/ since you said that /lh/ can occur and, I presume, isn't rare.

/m n ŋ/ m n ng
/p b t d k g tʼ kʼ/ p b t d k g tt kk
/f v θ ð s z ɬ ʃ ʒ χ h/ f v th dh s z ll sh zh kh h
/l r j w/ l r i u
/a aː e eː i iː o oː u uː ə/ a aa e ee i ii o oo u uu y

Here's an alternative with no digraphs, although it's a bit heavy on special characters. All of these characters are used for the sound in question in at least one natural language.

/m n ŋ/ m n ŋ
/p b t d k g tʼ kʼ/ p b t d k g ƭ ƙ
/f v θ ð s z ɬ ʃ ʒ χ h/ f v þ ð s z ł š ž x h
/l r j w/ l r i u
/a aː e eː i iː o oː u uː ə/ a ā e ē i ī o ō u ū ə

You could mix and match these systems, if you like. I think there's something to be said for combining the consonants from the second romanization with the vowels from the first and thus having only long segments be digraphs, for instance.

(Is this too much or too little information?)

It's definitely not too much information. Given anything that isn't a stop can be a syllable nucleus and given your sample, I'm going to presume that most CC sequences are valid onsets and valid codas, or that most non-valid onsets/codas would just syllabify one of the consonants anyway.

/tabir uːz maːr pfitʼ ve kairdið ʃo fuz ‖ biːr pkəi za χeːŋ za foːn fo parθot veu ʃeɬ za fəln sl̩ ksau ʒav fn̩kʼ ‖ ɬaʃir tʼep pʒ̍t ɬ̩ kʼfou/

First romanization: Tabir uuz maar pfitt ve kairdidh sho fuz. Biir pkyi za kheeng za foon fo parthot veu shell za fyln sl ksau zhav fnkk. Llashir ttep pzht ll kkfou.

Second romanization: Tabir ūz mār pfiƭ ve kairdið šo fuz. Bīr pkəi za xēŋ za fōn fo parþot veu šeł za fəln sl ksau žav fnƙ. Łašir ƭep pžt ł ƙfou.

[mèþru]

If you are concerned about phonological SAE-ness (which is OK, depending on what is the purpose of the language and realism of your setting), go this thread and take the test.

[Knit Tie]

Speaking of SAE, am I correct in thinking that both of my Shayanas aren't very SAE-like either?

[HazelFiver]

I put a disclaimer in case you thought I was totally new to this or just ignorant and only knew how to make stuff that contained the entire consonant inventory of English (it does, kind of). Also, the established phonology of this language ("Sparrowgrass" by the way) scores 73 on the test, which I thought was kind of high, so I thought I'd put in the ejectives to make it below 70 and also just for the heck of it since I can apparently pronounce them (though I still don't quite understand how they work so I'm not sure the sounds I'm using are real ejectives, but I don't know what else they could be). The syllabic fricatives don't affect the score, but maybe that's reasonable because they're also rare outside Europe. OTOH none of this technically matters since Sparrowgrass isn't set anywhere and is intended to be easy and intuitive for me to use rather than strictly realistic, so I suppose I should just pick something and stop fussing about it, and perhaps it doesn't even belong on this forum because the conlangs here all seem to aim for naturalism. That said, the existing orthography is kind of blah and also too Welsh, so thanks for the ideas.

I think I'm derailing the thread at this point, so feel free to ignore this post.

[mèþru]

The ejectives you use are pronounced like the hi-hat and snare drum of beatboxing, respectively.

I think I'm derailing the thread at this point, so feel free to ignore this post.

This thread has been derailed lots of times. Someone will always end up posting a new phonology to romanise and things will get r-railed in the end.

Speaking of SAE, am I correct in thinking that both of my Shayanas aren't very SAE-like either?

Aspiration instead of a voice distinction is very un-SAE. I suggest taking the test in the link I put for a more in-depth analysis.

[Chengjiang]

I put a disclaimer in case you thought I was totally new to this or just ignorant and only knew how to make stuff that contained the entire consonant inventory of English (it does, kind of). Also, the established phonology of this language ("Sparrowgrass" by the way) scores 73 on the test, which I thought was kind of high, so I thought I'd put in the ejectives to make it below 70 and also just for the heck of it since I can apparently pronounce them (though I still don't quite understand how they work so I'm not sure the sounds I'm using are real ejectives, but I don't know what else they could be). The syllabic fricatives don't affect the score, but maybe that's reasonable because they're also rare outside Europe. OTOH none of this technically matters since Sparrowgrass isn't set anywhere and is intended to be easy and intuitive for me to use rather than strictly realistic, so I suppose I should just pick something and stop fussing about it, and perhaps it doesn't even belong on this forum because the conlangs here all seem to aim for naturalism. That said, the existing orthography is kind of blah and also too Welsh, so thanks for the ideas.

I think I'm derailing the thread at this point, so feel free to ignore this post.

Honestly I think the SAE phonology test has some problems anyway, but sure, if you want it to sound less like an SAE language, ejectives definitely help. Also FWIW Sparrowgrass, at least as far as what I've seen here, doesn't sound non-naturalistic in any obvious way.

While the majority of conlangs here are intended to be naturalistic, I know some people here do make make non-naturalistic languages. One of the old members, JohnQPublik (who I'm not sure comes around here anymore), created Ithkuil, a non-naturalistic philosophical language.

[mèþru]

Ithkuil's creator used to come here? He was one of the people who inspired me to start conlanging!

[Chengjiang]

Ithkuil's creator used to come here? He was one of the people who inspired me to start conlanging!

Yes, he came here occasionally back in the day. I remember him doing a good thread on cognitive linguistics that's in the L&L Museum.

OK, I'm interested in how you guys would represent this in the Roman alphabet:

Tswana

/m n ɲ ŋ/
/b (d) dʒ/
/p t ts tɬ tʃ k/
/pʰ tʰ tsʰ tɬʰ tʃʰ kʰ qʰ/
/f s ʃ χ h/
/r/
/w l j/

/i u ɪ ʊ ɛ ɔ a/

High and low tones with high tones more marked than low

The syllable structure is (C)(w)V. [d] is an allophone of /l/ before /i u/.

[Knit Tie]

Roman? Not Cyrillic? Damn.

Anyway:

/m n ɲ ŋ/ <m n ñ ŋ>
/b (d) dʒ/ <bb (l) j>
/p t ts tɬ tʃ k/ <b d dz dlh dx g>
/pʰ tʰ tsʰ tɬʰ tʃʰ kʰ qʰ/ <p t ts tlh tx k q>
/f s ʃ χ h/ <f s x c h>
/r/ <r>
/w l j/ <w l y>

/i u ɪ ʊ ɛ ɔ a/ <ii uu i u e o a>

High and low tones with high tones more marked than low <V́ V>

[gufferdk]

/m n ɲ ŋ/ <m n ñ ng>
/b (d) dʒ/ <b (d) g>
/p t ts tɬ tʃ k/ <p t ts tl tc k>
/pʰ tʰ tsʰ tɬʰ tʃʰ kʰ qʰ/ <ph th tsh tlh tch kh q>
/f s ʃ χ h/ <f s c x h>
/r/ <r>
/w l j/ <v l j>

/i u ɪ ʊ ɛ ɔ a/ <i u y w e o a>

High and low tones with high tones more marked than low <V́ V>

[Chengjiang]

Roman? Not Cyrillic? Damn.

Go ahead and make a Cyrillic one.

[Knit Tie]

/m n ɲ ŋ/ <м н нь нг>
/b (d) dʒ/ <бб (л) ж>
/p t ts tɬ tʃ k/ <б д дз дл щ г>
/pʰ tʰ tsʰ tɬʰ tʃʰ kʰ qʰ/ <п т ц тл ч к къ>
/f s ʃ χ h/ <ф с ш х хъ>
/r/ <р>
/w l j/ <в л й>

/i u ɪ ʊ ɛ ɔ a/ <ие оу и у е о а>

High and low tones with high tones more marked than low

<ыэ ёю ы ю э ё я> for high tone vowels.

[mèþru]

Sitsvana
<m n ň ŋ>
<ḅ ž>
<b d ds dl dš g>
<p t ts tl tš k q>
<f s š x h>
<r>
<v l j>

<í ú i u e o a>

High and low tones with high tones more marked than low <V' V>.

agefaqeg-style:
sìtswana

<m n ň ŋ>
<b ž>
<p t ts tl ť k>
<ph th tsh tlh ťh kh q>
<f s š x h>
<r̀>
<w l r>

<i u ì ù e o a>

No tone markings

[opipik]

/m n ɲ ŋ/ <m n ny ng>
/b (d) dʒ/ <b̵ l d̶j>
/p t ts tɬ tʃ k/ <b d dz dl dj g>
/pʰ tʰ tsʰ tɬʰ tʃʰ kʰ qʰ/ <p t ts tl tj k q>
/f s ʃ χ h/ <f s sj x h>
/r/ <r>
/w l j/ <w l y>

/i u ɪ ʊ ɛ ɔ a/ <i u e o è ò a>

High and low tones with high tones more marked than low <á a> <ê è>

[mèþru]

Updated my romanisation.