Romanization Challenge Thread v2.0

[bradrn]

Kalam

This is an interesting one. Phonemically, the consonant inventory is pretty simple:

/p t tʃ k/
/ᵐb ⁿd ᶮdʒ ᵑɡ/
/s/
/m n ɲ ŋ/
/l w j/

Phonetically, however, there is quite a bit of allophony. Firstly, the voiceless stops have several allophones depending on their position within the word:

Phoneme	Word-initial	Word-medial	Word-final
p	ɸ	β	p~b
t	t	ɾ	ɾ
k	k	ɣ	k~ɡ

Additionally, /tʃ s/ can become voiced medially and finally, while the prenasalised stops become voiceless finally, and lose their nasal after /p tʃ k ᵐb ⁿd ᶮdʒ ᵑɡ/.

Also of note are the semivowels /w j/. These have several allophones as well, which if my transcription is correct are:

• [w j] initially and in environments C_V and V_V
• [ʉ ɪj] before palatal consonants (as well as velars for /i/)
• [u i] before other consonants
• [uw ij] finally

The vowels provide similar problems. The basic vowel set is /a e o/ [ɐ~ə e o], which is easy enough. (You might also add /i u/ depending on how you analyse it; my sources are inconsistent.) However, there is also an epenthetic vowel inserted between consonant clusters, with the following allophones:

• [ə~o] when C_C(C)oC#
• [ɐ~ə] when C_C(C)aC#
• [e~ɛ~ə] when C_C(C)eC# or C_Cj
• [u] when C_CwC# or C_w or #w_[+velar]
• [u~ʉ] when #w_[-velar]
• ∅ when (C,V)w_C, though in this case /w/ is realised as [u~ʉ]
• ∅ when C_j, though /j/ is realised as [ɪ~i~ij]
• [ɨ] when #j_C
• [∅~ɨ] when (C,V)j_C
• [ɨ~ə] otherwise

Stress is non-phonemic, so I won’t bother indicating it. It’s debatable whether the notion of ‘syllable structure’ is applicable.

I don’t have any sample text, but luckily Pawley does quote the phonetic transcription of a few phrases:

/jaⁿd wj mⁿdenᵑɡajn/
[jaⁿt wij mɨⁿdenɨᵑɡajɨn]

/mⁿdp kotp bjaŋ/
[mɨⁿdɨp koɾɨp bjaŋ]

/… mⁿdnknŋ/
[… mɨⁿdɨnɨɣɨnɨŋ]

/woᵑɡ ᵑɡdj ownᵑɡajn/
[woᵑk ᵑɡɨⁿdij ownɨᵑɡajɨn]

[Echobeats]

Kalam

/p/ f, f, p (initial, medial, final)
/t/ t
/tʃ/ c
/k/ k, ğ, k
/ᵐb ⁿd ᶮdʒ ᵑɡ/ b d j gg (or gk with final devoicing)
/s/ s
/m n ɲ ŋ/ m n ñ g
/l w j/ l w y

/a e o i u/ a e o i u
epenthetic [o ɐ ɛ u ʉ ɨ] o a e u u ı

/jaⁿd wj mⁿdenᵑɡajn/
[jaⁿt wij mɨⁿdenɨᵑɡajɨn]
yad wiy mıdenıggayın

/mⁿdp kotp bjaŋ/
[mɨⁿdɨp koɾɨp bjaŋ]
mıdıp kotıp byag

/… mⁿdnknŋ/
[… mɨⁿdɨnɨɣɨnɨŋ]
mıdınığınıg

/woᵑɡ ᵑɡdj ownᵑɡajn/
[woᵑk ᵑɡɨⁿdij ownɨᵑɡajɨn]
wogk ggıdiy ownıggayın

[Nortaneous]

/p t tʃ k/ <p/f t x k>
/ᵐb ⁿd ᶮdʒ ᵑɡ/ <b d j g>
/s/ <s>
/m n ɲ ŋ/ <m n ñ ng>
/l w j/ <l u i>

/a e ə o/ <a e y o>

kajn ak nej awsek amwⁿb | ɲn ak oⁿgnap wtsek ⁿd ap tan ⁿd ap jap ⁿg ⁿg swp | ɲn ak oⁿgnap wtsek ⁿd ap tan ⁿd ap jap ⁿg ⁿg met nŋl | aⁿdkⁿd katp owp
Kain ak nei ausek amub, ñyn ak ognap utsek dy ap tan dy ap iap gy gy sup, ñyn ak ognap utsek dy ap tan dy ap iap gy gy met nyngyl, adkyd katyp oup.

[bradrn]

/p t tʃ k/ <p/f t x k>
/ᵐb ⁿd ᶮdʒ ᵑɡ/ <b d j g>
/s/ <s>
/m n ɲ ŋ/ <m n ñ ng>
/l w j/ <l u i>

/a e ə o/ <a e y o>

kajn ak nej awsek amwⁿb | ɲn ak oⁿgnap wtsek ⁿd ap tan ⁿd ap jap ⁿg ⁿg swp | ɲn ak oⁿgnap wtsek ⁿd ap tan ⁿd ap jap ⁿg ⁿg met nŋl | aⁿdkⁿd katp owp
Kain ak nei ausek amub, ñyn ak ognap utsek dy ap tan dy ap iap gy gy sup, ñyn ak ognap utsek dy ap tan dy ap iap gy gy met nyngyl, adkyd katyp oup.

Minor correction: I believe /ⁿd ap tan/ and /ⁿd ap jap/ are single words, although I haven’t seen any explicit confirmation of this point yet. (This is why for my sample texts I chose ones with phonetic transcriptions.)

[Echobeats]

kajn ak nej awsek amwⁿb | ɲn ak oⁿgnap wtsek ⁿd ap tan ⁿd ap jap ⁿg ⁿg swp | ɲn ak oⁿgnap wtsek ⁿd ap tan ⁿd ap jap ⁿg ⁿg met nŋl | aⁿdkⁿd katp owp

Kayın ak ney ausek amub. Ñın ak ogganap utsek d-af-tan d-af-yap ggı ggı sup. Ñın ak ogganap utsek d-af-tan d-af-yap ggı ggı met nıgıl. Adığıd katıp oup.


Thanks for finding a text.

[Karch]

/p t tʃ k/ <p~v t~r c k~g>
/ᵐb ⁿd ᶮdʒ ᵑɡ/ <mb~mp nd~nt nj~nc nk>
/s/ <s>
/m n ɲ ŋ/ <m n ny ng>
/l w j/ <l w y>

/a e o/ <a e o>
[i ɨ u] <ii i u>
/Vw Vj/ <Vu Vi>

/jaⁿd wj mⁿdenᵑɡajn/
[jaⁿt wij mɨⁿdenɨᵑɡajɨn]
Yant wii mindeninkain.

/mⁿdp kotp bjaŋ/
[mɨⁿdɨp koɾɨp bjaŋ]
Mindip korip mbyang.

/… mⁿdnknŋ/
[… mɨⁿdɨnɨɣɨnɨŋ]
Mindinigining.

/woᵑɡ ᵑɡdj ownᵑɡajn/
[woᵑk ᵑɡɨⁿdij ownɨᵑɡajɨn]
Wonk nkindii owninkain.

kajn ak nej awsek amwⁿb | ɲn ak oⁿgnap wtsek ⁿd ap tan ⁿd ap jap ⁿg ⁿg swp | ɲn ak oⁿgnap wtsek ⁿd ap tan ⁿd ap jap ⁿg ⁿg met nŋl | aⁿdkⁿd katp owp
Kain ak nei ausek amump. Nyin ak onknap utsek ndi ap tan ndi ap yap nki nki sup. Nyin ak onknap utsek ndi ap tan ndi ap yap ngi ngi met ningil. Andgint karip oup.

[Echobeats]

I'm bringing hwair and yogh out of retirement for Montana Salish.

/a e i o u (ə)/ a e i o u -

/p t (k) kʷ q qʷ ʔ/ b d - g q qw `
/p’ t’ kʷ’ q’ qʷ’/ p t k q` qw`
/ʦ ʧ/ z c
/ʦ’ ʧ’ tɬ’/ tz tc tl
/s ʃ ɬ xʷ χ χʷ h/ s x hl ƕ ȝ ȝw h
/m n/ m n
/ˀm ˀn/ `m `n
/j w l ʕ ʕʷ/ y w l r rw
/ˀj ˀw ˀl ˀʕ ˀʕʷ/ `y `w `l `r `rw

/jetɬ’xʷasq’t qe ˀjamnʦut u qe es wiʧstm ɬu nxʷlxʷiltn ta es hoj qe ʦxʷiʦ’ɬlt ɬu qeqɬ nxʷlxʷiltn iqs ʃiʔmnwexʷ l es yaʔ ɬu l es xʷlxʷilt put u yetɬ’xʷasq’t ɬu qe nkʷ’ɬaxmintn nkʷ’uʔ wilʃ ɬu ne qe es lmntmnwexʷ u qe es ʦuti lemlmt.ʃ ʧ’ es yaʔ sqlqelixʷ χʷˀl qe sqelixʷ/

yetlƕasq`d qe `yamnzud u qe es wicsdm hlu nƕlƕildn da es hoy qe zƕitzhlld hlu qeqhl nƕlƕildn iqs xi`mnweƕ l es ya` hlu l es ƕlƕild bud u yetlƕasq`d hlu qe nkhlaȝmindn nku` wilx hlu ne qe es lmndmnweƕ u qe es zudi lemlmd. X c` es ya` sqlqeliƕ ȝw`l qe sqeliƕ.

/imʃe kʷ χest m ʃe putʔentxʷ ɬu a snkʷsqelixʷ/

imxe g ȝesd m xe bud`endƕ hlu a sngsqeliƕ

ETA: I thought I'd see what it looked like with <ə> for /ə/.

yetlƕasq`d qe `yamnzud u qe es wicsdəm hlu nƕəlƕildən da es hoy qe zƕitzəhlld hlu qeqəhl nƕəlƕildən iqs xi`əmnweƕ l es ya` hlu l es ƕəlƕild bud u yetlƕasq`d hlu qe nkəhlaȝmindən nku` wilx hlu ne qe es lmndəmnweƕ u qe es zudi lemlmd. X c` es ya` sqəlqeliƕ ȝwə`l qe sqeliƕ.

imxe g ȝesd m xe bud`endƕ hlu a səngsqeliƕ

[bradrn]

Now that everyone else has had a go at it, I might as well reveal how those Kalam examples are usually transcribed:

Yad wy mdengayn. Mdp kotp byaŋ. Mdnknŋ. Wog gdy owngayn.

Kayn ak new awsek amub, ñn ak ognap wtsek d ap tan d ap yap g g suwp, ñn ak ognap wtsek d ap tan d ap yap g g met nŋl, adkd katp owp.

Despite the unusual look, this actually seems to be one of the more widely-used Papuan orthographies (e.g. in the works of the late I. S. Majnep).

_________________


Now, how about a new challenge: Modern Hebrew. I find myself persistently annoyed at the sort of unsystematic romanizations which get used for this one; I wonder if anyone else has any better ideas?

Consonants:

/m n/
/p b t d k ɡ/
/t͡s (t͡ʃ) (d͡ʒ)/
/f v s z ʃ (ʒ) χ h/
/l j (w) ʀ/

(Note that the phonemes in brackets are only present in loanwords, though the sequence /tʃ/ occurs occasionally in native words.)

There are a couple of problematic cases here. Most notably, there is a glottal stop [ʔ], the status of which is controversial: it freely alternates with ∅, is not obligatory, and has largely predictable distribution (common between vowels where the second vowel is stressed, rare elsewhere). Another problematic case is word-final /h/; this is absent in the spoken language, but commonly included in transliterations. I will transcribe word-final /h/ but not [ʔ]; feel free to romanize these however you wish.

Vowels:

The basic vowel system is /a e i o u/. There is also the ‘shva’ vowel, commonly transcribed as /ə/. Wikipedia claims this is consistently pronounced exactly the same as /e/, though I’m fairly sure I’ve heard /ə/ occasionally. I’ll transcribe the two differently anyway.

Stress: Usually on the last syllable, though sometimes on the penult. Stress is phonemic.

Sample text:

I don’t happen to have a Modern Hebrew sample text on me right now, so I’ll just use a Biblical Hebrew text in the modern pronunciation (specifically, Genesis 1:1–3):

/bəʀeˈʃit baˈʀa eloˈkim et haʃaˈmajim vəˈet haˈaʀet͡s ‖ vəhaˈaʀet͡s hajəˈtah ˈtohu vaˈvohu | vəˈχoʃeχ al pəˈnej təˈhom | vəˈʀuaχ eloˈkim məʀaˈχefet al pəˈnej haˈmajim ‖ vaˈjomeʀ eloˈkim | jəˈhi oʀ | vajəˈhi oʀ/

And while I’m on the religious theme, I might as well include a few other religious terms, most notable for being romanised in nearly every way imaginable:

/ˈtoʀah ˈχalah ˈpesaχ χanuˈkah/

And for reference, here’s the sort of unsystematic romanization which is commonly used today by Hebrew-speakers:

Bereshit bara Elokim et hashamayim v’et ha’aretz. Veha’aretz hayetah tohu vavohu, v’choshekh al-p’nei tehom; v’ruach Elokim merachefet al-p’nei hamayim. Vayomer Elokim, ‘Y’hi or.’ Vayehi or.

Torah, challah, Pesach, Channukah. (Though you’ll also see Tora, chala, halla, Pesakh, as well as nearly every variant of (C)ha(n)nuka(h).)

[Echobeats]

Modern Hebrew

/m n/ m n
/p b t d k ɡ ʔ/ p b t d k ɡ '
/t͡s (t͡ʃ) (d͡ʒ)/ c tsh dj
/f v s z ʃ (ʒ) χ h/ f v s z sh j kh h
/l j (w) ʀ/ l y u r

/a e i o u ə/ a e i o u ë

Stress is usually on the final syllable; when elsewhere, I will mark it with an acute accent.

/bəʀeˈʃit baˈʀa eloˈkim et haʃaˈmajim vəˈet haˈaʀet͡s ‖ vəhaˈaʀet͡s hajəˈtah ˈtohu vaˈvohu | vəˈχoʃeχ al pəˈnej təˈhom | vəˈʀuaχ eloˈkim məʀaˈχefet al pəˈnej haˈmajim ‖ vaˈjomeʀ eloˈkim | jəˈhi oʀ | vajəˈhi oʀ/

/ˈtoʀah ˈχalah ˈpesaχ χanuˈkah/

Bëreshit bara Elokim et hashamáyim vë et ha'árec. Vë ha'árec hayëta tóhu va vóhu, vë khóshekh al pënei tëhom, vë rúakh Elokim mërakhéfet al pënei hamáyim. Va yómer Elokim: "Yëhi or", va yëhi or.

Tóra, Khála, Pésakh, Khanuka.

Out of curiosity, what determines the alternation of /və/ and /va/ for "and"? Some kind of vowel harmony?

(I was tempted to put my tongue firmly in my cheek and say that vowels would not have their own letters but instead be marked by configurations of dots below the consonants...)

[Karch]

/m n/ <m n>
/p b t d k ɡ/ <p b t d k g>
/t͡s (t͡ʃ) (d͡ʒ)/ <c ch j>
/f v s z ʃ (ʒ) χ h/ <f v z s x zh q h>
/l j (w) ʀ/ <l y w r>

/a e i o u ǝ/ <a e i o u ∅>
The schwa is marked with an apostrophe before a vowel.
Stress is marked with an acute accent when not on the last syllable

/bəʀeˈʃit baˈʀa eloˈkim et haʃaˈmajim vəˈet haˈaʀet͡s ‖ vəhaˈaʀet͡s hajəˈtah ˈtohu vaˈvohu | vəˈχoʃeχ al pəˈnej təˈhom | vəˈʀuaχ eloˈkim məʀaˈχefet al pəˈnej haˈmajim ‖ vaˈjomeʀ eloˈkim | jəˈhi oʀ | vajəˈhi oʀ/
Brexit bara Elokim et haxamáyim v'et haárec. Vhaárec haytah tóhu vavóhu, vqóxeq al pney thom, vrúaq Elokim mraqéfet al pney hamáyim. Vayómer Elokim: "Yhi or". Vayhi or.

/ˈtoʀah ˈχalah ˈpesaχ χanuˈkah/
Tórah, qálah, Pésaq, Qanukah

[bradrn]

Out of curiosity, what determines the alternation of /və/ and /va/ for "and"? Some kind of vowel harmony?

Hmm… I’ve been wondering the same myself, actually, so let me check…

So, as usual, the rules are different in Biblical and Modern Hebrew. In Biblical Hebrew (e.g. the text I gave), it depends to a large extent on whether the next vowel is long or short. However this distinction has been lost in the modern Israeli pronunciation, which is why the alternation in the text looks apparently random. By contrast, in Modern Hebrew, it’s simply pronounced /u-/ if the next consonant is labial or followed by a vowel, and /və-/ otherwise. (In informal contexts, apparently even that is lost.)

[Nortaneous]

Revised & Newly Endorsed Hebrew Romanization

/m n/ <m n>
/p b t d k ɡ/ <ph p th t kh k>
/t͡s (t͡ʃ) (d͡ʒ)/ <tsh tśh tśj>
/f v s z ʃ (ʒ) χ h/ <xw w s z ś ź x x>
/l ʔ (w) ʀ/ <l · w ɣ>

Adjacent to uvulars:
/a e ə o i u/ <a ej ə o e u>

Otherwise:
/a e ə o i u/ <ja jij jɨ jo ji ju>

/am em om im um/ <jã jɨj jow jĩ jũ>
/aʀ eʀ əʀ oʀ iʀ uʀ/ <ia iej iə io ie u>
/Vh/ <V̠>; /Vχ/ is written as /Vʀh/
Other coda consonants are written as if followed by schwa.

Stress, word boundaries, and /j/ aren't written.

/bəʀeˈʃit baˈʀa eloˈkim et haʃaˈmajim vəˈet haˈaʀet͡s ‖ vəhaˈaʀet͡s hajəˈtah ˈtohu vaˈvohu | vəˈχoʃeχ al pəˈnej təˈhom | vəˈʀuaχ eloˈkim məʀaˈχefet al pəˈnej haˈmajim ‖ vaˈjomeʀ eloˈkim | jəˈhi oʀ | vajəˈhi oʀ/

pə ɣej śji thjɨ pa ɣa ·jij ljo khjĩ ·jij thjɨ xja śja mja ·jĩ wjɨ ·jij thjɨ xja ·a ɣej tshy. wjɨ xja ·a ɣej tshy xja jɨ thja̠ thjo xju wja wjo xju, wə xo śie̠j ·ja ljɨ phjɨ njij jɨ thjɨ xjow, wə ɣu ·ia̠ ·jij ljo khjĩ mə ɣa xej xjwij thjɨ ·ja ljɨ phjɨ njij ·jɨ xja mja ·jĩ. wja ·jo miej ·jij ljo khjĩ, ·jɨ xji ·io, wja ·jɨ xji ·io

/ˈtoʀah ˈχalah ˈpesaχ χanuˈkah/
tho ɣa̠ xa lja̠ phjij sia̠ xa nju kja̠

[Echobeats]

bəʀeˈʃit
Brexit

You monster

[Richard W]

Why is there /k/ rather than /h/ in the word for 'God'?

[Nortaneous]

/pʰ p ⁿb tʰ t ⁿd tsʰ ts ⁿdz tʃʰ tʃ ⁿdʒ kʰ k ⁿg/ <ph p b th t d tsh ts dz ch c j kh k g>
/s ɬ ʃ h/ <s lh sh h>
/v l z ʒ ɣ/ <v l z zh x>
/r/ <r>
/m n ŋ/ <m n ng>
/ʔ/ <'> (0 word-initially)

Medials:
/w/ <w>

/a e ø ə o i y u/ <a e eu y o i iu u>
(uvularized, epiglottalized) <Vg aV>
(creaky, breathy) <Vq Vh>
(nasal) <Vm>
(rhotacized) <Vr>

/˧ ˩˥/ <V V́>

ⁿbu˧ ⁿda˩˦ pʰi˩˦ŋøʶʴ˩˦ a˧ɬḭ˩˥ɣaʶ˩˥ʃwo˧ o^ʢ˧ uʶ˩˥ ⁿdʒe̤˩˦ ta˧ mi˧ve˧ tʰa˧ suʶ˧ ⁿbə˧ⁿbe˩˦ ʔa̤ʴ˩˥ reʶʴ˩˦raʶʴ˩˦ kuʶ˧ raʴ˧ kwəʶʴ˧ tsʰəʶ˧ ʃə^ʢ˧nə˩˦ ʔe˧ⁿdzaʶ˩˦tʃḛ˩˥ || ɬøʶ˧meʴ˩˥ ⁿde˩˥viʴ˩˥ tʰa˧ kə˧ zø̤ʶ˩˥ ɬøʶ˧ kõʶ˩˥ su˩˥la̰˧tʃʰu˧ kuʶ˧ nõ̤^ʢ reʴ˩˥ o̰^ʢ˧ kʰaʶ˧ ve˩˥nõʴ˩˥vi˧
Bu dá phíngéurg alhíqxágshwo ao úg jéh ta mive tha sug bybé áhr régrág kug ra kwyrg tshyg shayný edzágcéq. Lheugmér dévír tha ky zéuhg lheug kómg súlaqchu kug naomh ré aoq khag vénórmvi.

with /kʰ k g ɣ ŋ/ <qh q qg xh nx> before uvularized vowels (instead of -g) and <-f> for creaky voice
Bu dá phínxeúr alhífxháshwo ao úg jéh ta mive tha sug bybé áhr régrág qu ra qwyr tshyg shayný edzágcéf. Lheugmér dévír tha ky zeuhg lheug qóm súlafchu qu naomh ré aof qha vénórmvi.

with grave accent
Bu dá phíngêur alhíqxâshwo ao û jeh ta mive tha sù bybe áhr rêrâ kù ra kwỳr tshỳ shayný edzâcéq. Lhèumér dévír tha ky zêuh lhèu kôm súlaqchu kù naomh ré aoq khà vénórmvi.

[dɮ the phoneme]

As a nebulously pseudo-Eastern-European language:

/pʰ p ⁿb tʰ t ⁿd tsʰ ts ⁿdz tʃʰ tʃ ⁿdʒ kʰ k ⁿg/ <p b mb t d nd c z nz cs j nj k g ng>
/s ɬ ʃ h/ <sz ł s h>
/v l z ʒ ɣ/ <v l z zs gh>
/r/ <r>
/m n ŋ/ <m n gn>
/ʔ/ <q>

Medials:
/w/ <w>

/a e ø ə o i y u/ <a e ö ë o i ü u>
(uvularized, epiglottalized) <Vg Vʿ>
(creaky, breathy) <Vʾ Vh>
(nasal) <Vn>
(rhotacized) <Vr>

/˧ ˩˥/ <V V́ (V̋)>

ⁿbu˧ ⁿda˩˦ pʰi˩˦ŋøʶʴ˩˦ a˧ɬḭ˩˥ɣaʶ˩˥ʃwo˧ o^ʢ˧ uʶ˩˥ ⁿdʒe̤˩˦ ta˧ mi˧ve˧ tʰa˧ suʶ˧ ⁿbə˧ⁿbe˩˦ ʔa̤ʴ˩˥ reʶʴ˩˦raʶʴ˩˦ kuʶ˧ raʴ˧ kwəʶʴ˧ tsʰəʶ˧ ʃə^ʢ˧nə˩˦ ʔe˧ⁿdzaʶ˩˦tʃḛ˩˥ || ɬøʶ˧meʴ˩˥ ⁿde˩˥viʴ˩˥ tʰa˧ kə˧ zø̤ʶ˩˥ ɬøʶ˧ kõʶ˩˥ su˩˥la̰˧tʃʰu˧ kuʶ˧ nõ̤^ʢ reʴ˩˥ o̰^ʢ˧ kʰaʶ˧ ve˩˥nõʴ˩˥vi˧
Mbu ndá pígnőrg ałíʾghág swo oʿ úg njéh da mive ta szug mbëmbé qárh rérg kug rar kwërg cëg sëʿne̋ qenzágcséʾ. Łögmér ndévír ta kë zőrh łör kóng szúlaʾcsu kul nonhʿ rér onʿ kag vénórnvi.

-----------

Unnamed conlang

Root-initials:
/p t tʃ k/ <p t c k>
/s ʃ h/ <s sy h>
/l ʎ/ <l ly>
/m n ɲ/ <m n ny>

Root-medials:
/p t tʃ tɬ cʎ̥˔ k/ <p t c tl cl k>
/pʰ tʰ tʃʰ kʰ/ <ph th ch tlh clh kh>
/p' t' tʃ' k'/ <p' t' c' k'>
/b d dʒ dɮ dʎ̥ g/ <b d j dl jl g>
/bʱ dʱ dʒʱ gʱ/ <bh dh jh gh>
/ɓ ɗ ʄ ɠ/ <b' d' j' g'>
/ᵐb ⁿd ⁿdʒ ⁿdɮ ⁿdʎ̥ ᵑg/ <mb nd nj ndl njl ng>
/ᵐbʱ ⁿdʱ ⁿdʒʱ ᵑgʱ/ <mbh ndh njh ngh>
/ᵐɓ ⁿɗ ⁿʄ ᵑɠ/ <mb' nd' nj' ng'>
/f s ɬ ʃ ʎ̥˔ h/ <f s lh sy lhy h>
/s' ʃ'/ <s' s'y>
/v z l ʒ ʎ/ <v z l zy ly>
/ᵐv ⁿz ⁿl ⁿʒ ⁿʎ/ <mv nz nl nzy nly>
/m n ɲ/ <m n ny>

Vowels:
/i u e o a/ <i u e o a>

(C)V syllables

/ha imana tʃuⁿdʱeva la pe tovaʎ̥˔e sugazuga la ha iedʒʱu so | pea ʎemegala sodezidiʄo amano la haᵑɠe etɬo ha inaⁿza/
<ha imana cundheva la pe tovalhye sugazuga la ha iejhu so, pea lyemegala sodezidij'o amano la hang'e etlo ha inane.>

[bradrn]

Why is there /k/ rather than /h/ in the word for 'God'?

The custom in Judaism is to avoid pronouncing or writing the name of God outside the context of prayer. In particular, ⟨ʔlhm⟩ is replaced with ⟨ʔlqm⟩ /elokim/ (and similarly ⟨yhwh⟩ is replaced by ⟨hšm⟩ /haʃem/). In deference to this custom, I did the same when transcribing that text. (Not that it makes much difference to the text anyway.)

[Nortaneous]

/pʰ p ⁿb tʰ t ⁿd tsʰ ts ⁿdz tʃʰ tʃ ⁿdʒ kʰ k ⁿg/ <p b bb t d dd c z zz q j jj k g gg>
/s ɬ ʃ h/ <s lh x h>
/v l z ʒ ɣ/ <w l ss y hh>
/r/ <r>
/m n ŋ/ <m n ng>
/ʔ/ <0>

Medials:
/w/ <u>

/a e ø ə o i y u/ <a ie eo e o i iu u>
uvularized: <ak iek...> (tone 1) / <ag ieg...> (tone 2)
epiglottalized: <aa ai ao ae au ei eu ou>
(rhotacized) <Vr>

/Ṽ/ (modal) <Vn>
/Ṽ/ (uvularized) <Vnk> / <Vng>

/˧ ˩˥/ (modal) <-0 -f>
/˧ ˩˥/ (creaky) <-x -v>
/˧ ˩˥/ (breathy) <-h -l>

ⁿbu˧ ⁿda˩˦ pʰi˩˦ŋøʶʴ˩˦ a˧ɬḭ˩˥ɣaʶ˩˥ʃwo˧ o^ʢ˧ uʶ˩˥ ⁿdʒe̤˩˦ ta˧ mi˧ve˧ tʰa˧ suʶ˧ ⁿbə˧ⁿbe˩˦ ʔa̤ʴ˩˥ reʶʴ˩˦raʶʴ˩˦ kuʶ˧ raʴ˧ kwəʶʴ˧ tsʰəʶ˧ ʃə^ʢ˧nə˩˦ ʔe˧ⁿdzaʶ˩˦tʃḛ˩˥ || ɬøʶ˧meʴ˩˥ ⁿde˩˥viʴ˩˥ tʰa˧ kə˧ zø̤ʶ˩˥ ɬøʶ˧ kõʶ˩˥ su˩˥la̰˧tʃʰu˧ kuʶ˧ nõ̤^ʢ reʴ˩˥ o̰^ʢ˧ kʰaʶ˧ ve˩˥nõʴ˩˥vi˧
Bbu ddaf pif·ngeorg a·lhix·hhag·xuo au ug jjiel da mi·wie ta suk bbe·bbief arl rieg·rag guk ra guerk cek xae·nef ie·zzag·jievg. Lheok·mierg ddief·wirf ta ge sseolg lheok kong suf·lax·qu guk naunh rief aux kak wief·nornf·wi.

Root-initials:
/p t tʃ k/ <p t c k>
/s ʃ h/ <s x h>
/l ʎ/ <l r>
/m n ɲ/ <m n v>

Root-medials:
/p t tʃ tɬ cʎ̥˔ k/ <pp tt cc ttl ccr kk>
/pʰ tʰ tʃʰ kʰ/ <spp stt scc skk>
/p' t' tʃ' k'/ <hpp htt hcc hkk>
/b d dʒ dɮ dʎ̥ g/ <p t c tl cr k>
/bʱ dʱ dʒʱ gʱ/ <sp st sc sk>
/ɓ ɗ ʄ ɠ/ <hp ht hc hk>
/ᵐb ⁿd ⁿdʒ ⁿdɮ ⁿdʎ̥ ᵑg/ <np nt nc ntl ncr nk>
/ᵐbʱ ⁿdʱ ⁿdʒʱ ᵑgʱ/ <nsp nst nsc nsk>
/ᵐɓ ⁿɗ ⁿʄ ᵑɠ/ <npp ntt ncc nkk>
/f s ɬ ʃ ʎ̥˔ h/ <ff ss ll xx rr h>
/s' ʃ'/ <hs hx>
/v z l ʒ ʎ/ <f s l x r>
/ᵐv ⁿz ⁿl ⁿʒ ⁿʎ/ <nf ns nl nx nr>
/m n ɲ/ <m n v>

Vowels:
/i u e o a/ <i u e o a>

(C)V syllables

/ha imana tʃuⁿdʱeva la pe tovaʎ̥˔e sugazuga la ha iedʒʱu so | pea ʎemegala sodezidiʄo amano la haᵑɠe etɬo ha inaⁿza/
ha imana cunstefa la pe tofarre sukasuka la ha iescu so, pea remekala sotesitihco amano la hankke etlo ha inansa

[dɮ the phoneme]

Unnamed conlang v. 2:

Root-initials:
/p t tʃ k/ <p t c k>
/s ʃ h/ <s sy h>
/l ʎ/ <l ly>
/m n ɲ/ <m n ny>

Root-medials:
/p t tʃ tɬ cʎ̥˔ k/ <p t c tl cl k>
/pʰ tʰ tʃʰ kʰ/ <ph th ch tlh clh kh>
/p' t' tʃ' k'/ <p' t' c' k'>
/b d dʒ dɮ ɟʎ̞ g/ <b d j dl jl g>
/bʱ dʱ dʒʱ gʱ/ <bh dh jh gh>
/ɓ ɗ ʄ ɠ/ <b' d' j' g'>
/ᵐb ⁿd ⁿdʒ ⁿdɮ ⁿɟʎ̞ ᵑg/ <mb nd nj ndl njl ng>
/ᵐbʱ ⁿdʱ ⁿdʒʱ ᵑgʱ/ <mbh ndh njh ngh>
/ᵐɓ ⁿɗ ⁿʄ ᵑɠ/ <mb' nd' nj' ng'>
/m n ɲ/ <m n ny>
/m̥ n̥ ɲ̊/ <mh nh nhy>
/m̰ n̰ ɲ̃/ <m' n' n'y>
/f s ɬ ʃ ʎ̥˔ h/ <f s lh sy lhy h>
/s' ʃ'/ <s' s'y>
/v z l ʒ ʎ/ <v z l zy ly>
/ᵐv ⁿz ⁿl ⁿʒ ⁿʎ/ <mv nz nl nzy nly>
/l̰ ʎ̰/ <l' l'y>

Vowels:
/i u e ɔ a/ <i u e o a>

(C)V syllables

There's now explicit rules for the (very complex) system consonant mutation, which occurs both within words and across words, triggered by particular morphemes. Morphemes beginning with one of /p t tʃ k s ʃ h l ʎ/ undergo mutation according to the following 2x5 grid of consonant grades:

Code: Select all

		plain				voiced			aspirate				glottalic 1			glottalic 2
oral		p t tʃ k s ʃ h l ʎ		b d dʒ g z ʒ v l ʎ	pʰ tʰ tʃʰ kʰ s ʃ f ɬ ʎ̥˔    	    
														p' t' tʃ' k' s' ʃ' f~∅ tɬ cʎ̥˔	p' t' tʃ' k' s' ʃ' f l̰ ʎ̰ 
nasal			ᵐb ⁿd ⁿdʒ ᵑg ⁿz ⁿʒ ᵐv ⁿl ⁿʎ			ᵐbʱ ⁿdʱ ⁿdʒʱ ᵑgʱ s ʃ f ɬ ʎ̥˔     

I'm not sure if it's clear, but basically the plain/voiced nasal grades are identical and the glottal grades are identical in both nasal and oral.

Morphemes that begin with a nasal consonant or a vowel are more complicated. They fall into two classes, which undergo mutation differently:

Code: Select all

Class 1:
			plain/voiced	aspirate		glottalic
oral/nasal		m n ɲ ∅		m̥ n̥ ɲ̊ s~(ʃ)/f	 	m̰ n̰ ɲ̃ b/d~(dʒ)/g

Class 2:
			plain		voiced		aspirate		glottalic
oral			m n ɲ ∅		b d dʒ g	bʱ dʱ dʒʱ gʱ		ɓ ɗ ʄ ɠ
nasal				ᵐb ⁿd ⁿdʒ ᵑg		ᵐbʱ ⁿdʱ ⁿdʒʱ ᵑgʱ	ᵐɓ ⁿɗ ⁿʄ ᵑɠ

The class 1 aspirate and glottalic grades of the zero initial vary by the triggering morpheme: some trigger the s~(ʃ) set, others the f set, etc. Among the s~(ʃ) set and d~(dʒ), /ʃ/ and /dʒ/ appear before /e a/, while /s/ and /d/ appear before other vowels.

-----

Well, I'd really like to come up with a romanization that's systematic with respect to all these consonant gradations, but honestly working all this out was tiring so I'm just gonna hope somebody else here thinks that sounds like a fun task.

-----

Ok, actually, here are some simple examples for romanization:

/pe/, def.sg.m article, glottalic 1 grade
/ʎɔ/, def.sg.f article, oral voiced grade
/pep'e/ def.pl.m article, glottalic 1 grade
/ʎɔʎɔ/, def.pl.f article, oral voiced grade
/e-/, pl. suffix, nasal plain/voiced grade, class 2
/lɔga-/, augmentative, oral aspirate grade
/la/, acc., plain grade

/taha/, "cat"
/muⁿɟʎ̞ete/, "palm frond", class 1
/ɔᵑgu/, "female panther", class 2
/nɔmɔ/, "man", class 2

/pe t'aha/, /ʎɔ daha/, /eⁿdaha/, /pep'e ɠeⁿdaha/, /ʎɔʎɔ geⁿdaha/, /lɔgatʰaha/, /pe tɬɔgatʰaha/

/pe m̰uⁿɟʎ̞ete/, /emuⁿɟʎ̞ete/, /pep'e ɠemuⁿɟʎ̞ete/, /lɔgam̥uⁿɟʎ̞ete/

/ʎɔ gɔᵑgu/, /eᵑgɔᵑgu/, /ʎɔʎɔ geᵑgɔᵑgu/, /lɔgagʱɔᵑgu/

/pe dɔmɔ/, /eⁿdɔmɔ/, /la nɔmɔ/, /lɔgadʱɔmɔ/

[Killerwhale]

Hey folks!

I used to have an account here, but I forgot the password, so, new account, new nickname.

Anyway,

Code: Select all

m n
p b t̪ d̪ t d ʈ ɖ k g
ɓ ɖ tʼ ʈʼ ʄ~ɠ̪ kʼ
tɕ dʑ ʈʂ ɖʒ
tɕʼ ʈʂʼ
f v s z ɕ ʂ x
w ɾ l j

a e i o u ɛ~ɜ

Here's me with my stuff again. A variant of American English that was adopted as a second language by space colonists (with a large Nilo-Saharan speaking population) and has subsequently evolved into something quite different. Diachronically, the retroflexes come from /ɻ/ + alveolar plosive sequences, with alveolar plosives followed by /ɻ/ becoming retroflex affricates instead; the dentals come from th-stopping and the glottalized series - influenced by the rich glottals of Nilo-Saharan languages - comes from nasal+plosive sequences through an intermediate stage where the nasal was debuccalised to a glottal stop. Initial and intevocalic /ɻ/ became /ʂ/ instead, pushing the preexisting postalveolar fricative into /ɕ/. /ɛ~ɜ/ used to be /ɨ/, which used to be modern AmE's /ɪ/, /ʉ/ and /ʊ/.

The most complex possible syllable structure is CCV[jw]C, though CV syllables are heavily preferred, and most coda consonants are either fricatives or sonorants.