Phrase evolution game

[Rounin Ryuuji]

• Short vowels in stressed open syllables lengthen;
• [ui uː oː ɔː eː ɛː ai] > [yː yː uː oː aː eː ɛː]
• Non-aspirated stops and affricates and intervocalic [s] voice; prenasalisation is lost from voiced stops that have it; [ŋ] denasalises to [g]
• Coda [f] vocalises to [u], [l] to [i u] depending on what precedes, disappearing after diphthongs ending in either; [ɣ] devoices to [x~ç]
• Remaining postconsonantal [l] vocalises to [j], disappearing before [i];

[ˈsaː.zad ˈtʰoːm ˈd͡ʒoːuli ˈgɛːlid tʰoː ˈweːi ˈçyːlut ˈfoːd͡ʒi | ˈnuː ˈkʰaːmb t͡ʃʰeːl boːi'diː ‖ ˈgantʰiɡ ˈsiːd͡ʒ la'meː ˈjyːzaːid͡ʒ ‖ giːˈjyːsaː ˈwiːd͡ʒi ˈfoːid͡ʒ d͡ʒah ˈneːk we: ˈgantʰ tʰoː ˈsiːu la'meː ˈgul | geː niː kʰaː]

<Sásat tóm djóuli gêlid tó wéi hýlut fódji, nú kámb tjél póití. Gantig sídj lamé jýsáidj, gíjýsá wíldji fóidj djah nék wé gant tó síu lamé gul, gé ní ká.>

[bradrn]

Interesting! Let me do some grammatical changes before it gets distorted too much:

• ‘Say’ [foːdʒ] becomes a quotative; ‘follow’ [dʒoːu] becomes a conjunction, reduces and cliticises to previous word
• All 2P clitics delete except for past *=mu (now /-m(u)/) and question *=be (now /-b/)
• Prefixation of pronominal clitics to all verbs, except in some SVCs – if I’ve applied the previous changes right, they should now be /b- beː- ni- nd- ki- ɡiː-/. Due to how clitics work this sandwiches the past clitic /mu/ as a prefix, which grammaticalises to a true past tense
• Articles cliticise to noun with some fusion

[ˈsaːzadoːdʒo ˈɡeːlidoː kimuˈweːi ˈçyːlut foːdʒ | nuː ˈkʰaːmb kitʃʰeːl boːiˈdiː | ˈɡantʰ kimuˈsiːdʒ laˈmeː kimuˈjyːzaːidʒ ‖ ɡiːmuˈjyːsa ˈwiːdʒi foːdʒ dʒah neːk kiˈweː ˈɡantʰo kiˈsiːu laˈmeː ɡul | geː kiˈniː kʰaː]

Sásatódjo gêlito kimuwêi hýlut fódj, ‘nú kámb kitjél póití’, gant kimusídj lamé kimujýsaidj. Gîmujýsá wíldji fódj djah nék kiwé gantó kisíu lamé gul, gé kiní ká.

sásat=tó=djo

wind=DEF.SG=CONJ

gêlid=tó

sun=DEF.SG

ki-mu-wêi

3s-PST-do.IPFV.CONT

hýlut

argument

fódj,

QUOT,

‘nú

‘who

kám=b

strong.NMLZ=Q

ki-tjél

3s-leave

póití’,

all?’,

gant

person

ki-mu-sí-dj

3s-PST-have.PFV-DIM

lamé

cloak

ki-mu-jýsái-dj

3s-PST-come.PFV-DIM

Gî-mu-jýsá

3p-PST-fall.PFV

wíl-dji

contact.PFV

fódj

QUOT

djah

if

nék

someone

ki-wé

3s-do

gant=tó

person=DEF.SG

ki-síu

3s-remove

lamé

cloak

gul,

move.off,

gé

3s

ki-ní

3s-COP

ká

strong

[Rounin Ryuuji]

Interesting! Let me do some grammatical changes before it gets distorted too much:

• ‘Say’ [foːdʒ] becomes a quotative; ‘follow’ [dʒoːu] becomes a conjunction, reduces and cliticises to previous word
• All 2P clitics delete except for past *=mu (now /-m(u)/) and question *=be (now /-b/)
• Prefixation of pronominal clitics to all verbs, except in some SVCs – if I’ve applied the previous changes right, they should now be /b- beː- ni- nd- ki- ɡiː-/. Due to how clitics work this sandwiches the past clitic /mu/ as a prefix, which grammaticalises to a true past tense
• Articles cliticise to noun with some fusion

Okay, I like this version of the game — I was pondering if it wouldn't go this way earlier, with somebody who intimately understands the language making the grammatical changes every few rounds of sound changes. Anyway...

• Short vowels between long vowels, and at the end of long words, tend to reduce; [a o] > [ɐ], [i] > [ɪ], [u] > [ɵ]; some pretonic vowels in medial syllables also reduce;
• [z] > [ʁ]
• [tʰ d] > [t͡sʰ d͡z] before front vowels;
• [kʰ g] > [cʰ ɟ] before [e i y]; [tʃʰ dʒ] > [c͡ɕ ɟ͡ʑ] in the same environment, > [t͡s d͡z] elsewhere; [k c] > [kʰ cʰ] where they still occur;
• [ah] > [aː]
• Terminal [l] after front vowels becomes [i], long elements in diphthongs shorten; the quotative particle begins to show a strongly reduced vowel, despite having a historically long one;

[ˈsaːʁɐdoːd͡zɐ ˈɟeːlɪdoː cʰi.mɵˈwei ˈçyːlɵt fɐd͡z | nuː ˈkʰaːmb cic͡ɕʰei boiˈd͡ziː ˈɡantʰ cʰi.mɵˈɕiːd͡z laˈmeː cʰi.mɵˈjyːʁaid͡z]
Sáratódza Gêlidó kimùwêi hýlùt fàdz, ‘Nú kámb kichei boidzí’, gant kimùshídz lamé kimùjýraidz.

[ɟiːmuˈjyː.ʁa ˈwiːɟ͡ʑi fɐd͡z d͡zaː neːk cʰiˈweː ˈɡantʰo cʰiˈsiu laˈmeː ɡul | ɟeː cʰiˈniː kʰaː]
Gîmujýrá wíldzhi fàdz dzá nék kiwé gantó kisiu lamé gul, gé kiní ká.

sárat=tó=dza

wind=DEF.SG=CONJ

gêlid=tó

sun=DEF.SG

ki-mù-wêi

3s-PST-do.IPFV.CONT

hýlut

argument

fàdz,

QUOT,

‘nú

‘who

kám=b

strong.NMLZ=Q

ki-chei

3s-leave

boidzí’,

all?’,

gant

person

ki-mu-shí-dz

3s-PST-have.PFV-DIM

lamé

cloak

ki-mu-jýrai-dz

3s-PST-come.PFV-DIM

Gî-mu-jýrá

3p-PST-fall.PFV

wíl-dzhi

contact.PFV

fàdz

QUOT

dzá

if

nék

someone

ki-wé

3s-do

gant=tó

person=DEF.SG

ki-siu

3s-remove

lamé

cloak

gul,

move.off,

gé

3s

ki-ní

3s-COP

ká

strong

[/quote]

[bradrn]

Okay, I like this version of the game — I was pondering if it wouldn't go this way earlier, with somebody who intimately understands the language making the grammatical changes every few rounds of sound changes. Anyway...

Yeah, that’s what I was hoping as well, but alas, it was not to be. I think I’ll stay out of the phonology for this one, but I’m happy to be official ‘grammar-poker’ if everyone remembers to keep the glosses up to date.

I will make a phonological/orthographic change just this once, though:

1. [i] → [y], allowing ⟨j⟩ → ⟨y⟩

[ˈsaːʁɐdoːd͡zɐ ˈɟeːlɪdoː cʰi.mɵˈwei ˈçiːlɵt fɐd͡z | nuː ˈkʰaːmb cic͡ɕʰei boiˈd͡ziː ˈɡantʰ cʰi.mɵˈɕiːd͡z laˈmeː cʰi.mɵˈjiːʁaid͡z]
Sáratódza Gêlidó kimùwêi hílùt fàdz, ‘Nú kámb kichei boidzí’, gant kimùshídz lamé kimùyíraidz.

[ɟiːmuˈjiː.ʁa ˈwiːɟ͡ʑi fɐd͡z d͡zaː neːk cʰiˈweː ˈɡantʰo cʰiˈsiu laˈmeː ɡul | ɟeː cʰiˈniː kʰaː]
Gîmuyírá wíldzhi fàdz dzá nék kiwé gantó kisiu lamé gul, gé kiní ká.

sárat=tó=dza

wind=DEF.SG=CONJ

gêlid=tó

sun=DEF.SG

ki-mù-wêi

3s-PST-do.IPFV.CONT

hílut

argument

fàdz,

QUOT,

‘nú

‘who

kám=b

strong.NMLZ=Q

ki-chei

3s-leave

boidzí’,

all?’,

gant

person

ki-mu-shí-dz

3s-PST-have.PFV-DIM

lamé

cloak

ki-mu-yírai-dz

3s-PST-come.PFV-DIM

Gî-mu-yírá

3p-PST-fall.PFV

wíl-dzhi

contact.PFV

fàdz

QUOT

dzá

if

nék

someone

ki-wé

3s-do

gant=tó

person=DEF.SG

ki-siu

3s-remove

lamé

cloak

gul,

move.off,

gé

3s

ki-ní

3s-COP

ká

strong

[Otto Kretschmer]

Anyone for Mongol, Evenki or Proto Indo Anatolian?

[Rounin Ryuuji]

I think we still have quite a bit in this fictitious language first.

[Neonnaut]

• n → ɲ / _V[+high]
• ai ei oi iu → ɑː eː uj uj
• e(ː) o(ː) ɵ → i(ː) u(ː) e
• ˈa ˈɐ aː a ɐ → ɑ ɑ ɑː æ æ
• c͡ɕ(ʰ) → t͡ɕ(ʰ)
• c(ʰ) → s(ʰ) /_i
• d͡z → t͡s /_#
• d͡z → z / elsewhere
• ç → hj
• ɑ → o / _N
• l → o _#
• Unsure how there could be one lonely /ɪ/. So /ɪ/ → /i/.

[ˈsɑːʁæduːzæ ˈɟiːliduː sʰi.meˈwiː ˈhjiːlet fɐt͡s | ɲuː ˈkʰoːmb sit͡ɕʰiː bujˈziː ˈɡontʰ sʰi.meˈɕiːt͡s læˈmiː sʰi.meˈjiːʁɑːt͡s]
Sáratódza Gêlidó kimùwêi hílùt fàdz, ‘Nú kámb kichei boidzí’, gant kimùshídz lamé kimùyíraidz.

[ɟiːmuˈjiː.ʁæ ˈwiːɟ͡ʑi fɑt͡s zɑː niːk sʰiˈwiː ˈɡontʰu sʰiˈsuj læˈmiː ɡuo | ɟiː sʰiˈɲiː kʰɑː]
Gîmuyírá wíldzhi fàdz dzá nék kiwé gantó kisiu lamé gul, gé kiní ká.

sárat=tó=dza

wind=DEF.SG=CONJ

gêlid=tó

sun=DEF.SG

ki-mù-wêi

3s-PST-do.IPFV.CONT

hílut

argument

fàdz,

QUOT,

‘nú

‘who

kám=b

strong.NMLZ=Q

ki-chei

3s-leave

boidzí’,

all?’,

gant

person

ki-mu-shí-dz

3s-PST-have.PFV-DIM

lamé

cloak

ki-mu-yírai-dz

3s-PST-come.PFV-DIM

Gî-mu-yírá

3p-PST-fall.PFV

wíl-dzhi

contact.PFV

fàdz

QUOT

dzá

if

nék

someone

ki-wé

3s-do

gant=tó

person=DEF.SG

ki-siu

3s-remove

lamé

cloak

gul,

move.off,

gé

3s

ki-ní

3s-COP

ká

strong

[Rounin Ryuuji]

• Massive vowel chain shift:
  • [æ ɑ ɑː] > [ɛ ɔ ɔː]
  • [iː] > [ai] > [aː]
  • [u uj uː] > [y yː yː]
  • [o oː] > [u uː]; [uu] > [uː]
  • [e] > [i]
• Remaining long vowels other than [a] diphthongise: [ɔː yː uː ] > [wɛ ɥi ou]; [aː] shortens to [a]
• The function word fats reduces to a simple suffix -[(ə)s], graphically ès
• [tʰ d s sʰ] merge with [c͡ɕʰ ɟ͡ʑ ʑ ɕʰ] before [ɥ y i], remaining non-aspirated [s] and [ɟ] > [z]; non-aspirated [ɕ] > [ʑ]; [h] becomes [ç] before front vowels and [j]
• Remaining [ts] > [s]

[ˈzwɛ.ʁɛ.ɟ͡ʑɥi.zɛ ˈzaliɟ͡ʑɥi ɕʰi.miˈwa ˈçjalitəs | ɲɥi ˈkʰoumb ʑic͡ɕʰa byˈza ˈɡuntʰ ɕʰi.miˈʑas lɛˈma ɕʰi.miˈjaʁwɛs]
Zuerejuize zalijui shimiwa hyalitès, nyi koumb zhicha büza gunt shimizhas lema shimiyarwes

[zamyˈja.ʁɛ ˈwa.ɟ͡ʑis zwɛ nak ɕʰiˈwa ˈɡunc͡ɕʰy ɕʰiˈʑɥi lɛˈma ɡou | za ɕʰiˈɲa kʰwɛ]
Zamüyare wajis zwe nak shiwa gunchü zhui lema gou, za shinya kue

Zuere=jui=za

wind=DEF.SG=CONJ

zalid=jui

sun=DEF.SG

shi-mi-wa

3s-PST-do.IPFV.CONT

hyalit

argument

(è)s,

QUOT,

‘nui

‘who

koum=b

strong.NMLZ=Q

zhi-cha

3s-leave

büza’,

all?’,

gunt

person

shi-iu-zha-s

3s-PST-have.PFV-DIM

lema

cloak

shi-mi-yarwe-s

3s-PST-come.PFV-DIM

Za-mü-yarwe

3p-PST-fall.PFV

waji-s

contact.PFV

ès

QUOT

zwe

if

nak

someone

shi-wa

3s-do

gunt=chü

person=DEF.SG

zhui

3s-remove

lema

cloak

gou,

move.off,

za

3s

shi-nya

3s-COP

kue

strong

I think this language is very nice now.

[bradrn]

Time for another grammar change!

• Definite article =jui/=chü [=ɟ͡ʑɥi ~ c͡ɕʰy] → -jui [-ɟ͡ʑy] after vowels, → -ui [y] after consonants
• Quotative [-s] disappears, replaced with periphrastic construction =za -wa
• Variant 3s prefix zhi- (of uncertain etymology) becomes established in the nonpast, as opposed to the past 3s prefix shimi-
• Under Neo-Hlʉ̂ influence SVCs with sa ‘take’ becomes more and more common eventually giving SOV word order with an ergative clitic =sa
• yarwe [ˈja.ʁɛ] ‘fall/come’ (the meanings conflated by means of a typo in the original text, sorry) becomes a general andative prefix yar- [jaʁ-]; gou [gou] ‘move off’ similarly becomes venitive suffix
• shiwa [ɕʰiˈwa] grammaticalises to causative particle with some word order change
• Some lexical replacement (Neo-Hlʉ̂ again)

[ˈzwɛʁɛɟ͡ʑyzɛ ˈzalidy ɕʰimiˈwa ˈçjalitzɛ ɕʰimiˈwa | ɲɥi ˈkʰoumsab byˈza ʑic͡ɕʰa | ˈɡuntʰsa dop ɕʰimijaʁˈʑas]
Zuerejuiza zalidui shimiwa hyalitza shimiwa, ‘nui koumsab büza zhicha?’, guntsa dop shimiyarzhas.

[zamyjaʁˈwaɟ͡ʑiszɛ zamyˈwa zwɛ ˈnaksa ˈɡunty ɕʰiˈwa dop ʑiˈʑɥigou | za ʑiˈɲa kʰwɛ]
Zamüyarwajisza zamüwa zwe naksa guntui shiwa dop zhizhuigou, za zhinya kue.

Zuere-jui=za

wind-DEF.SG=CONJ

zalid-ui

sun=DEF.SG

shimi-wa

3s.PST-do.IPFV

hyalit=za

argument=CONJ

shimi-wa,

3s.PST-do.IPFV,

‘nui

‘who

koum=sa=b

strong.NMLZ=ERG=Q

büza

all

zhi-cha?’,

3s.NPST-leave?’,

gunt=sa

person=ERG

dop

cloak

shimi-yar-zha-s.

3s.PST-AND-have.PFV-DIM

Zamü-yar-wajis=za

3p.PST-AND-contact.PFV=CONJ

zamü-wa

3p.PST-do

zwe

if

nak=sa

someone=ERG

gunt-ui

person=DEF.SG

shiwa

CAUS

dop

cloak

zhi-zhui-gou,

3s.NPST-remove-VEN,

za

3s

zhi-nya

3s.NPST-COP

kue

strong

I think this language is very nice now.

As its creator I respectfully disagree.

[Rounin Ryuuji]

• Affricates disaffricate; assimilation of certain morphemes;
• Surface filters reaffricate dentals before front vowels;
• Final [p b] vocalise to [u], [au] merges with [ou];
• Nasal consonants in coda are lost, nasalising the preceding vowel or diphthong; short vowels become long, and tend to lower somewhat, during nasalisation;
• [w] > [v]
  
  [ˈzvɛʁɛʑyzɛ ˈza.li.ɟ͡ʑy ɕʰimiˈva ˈçja.li.sɛ ɕʰi.miˈva | ɲɥi ˈkʰõũsou byˈza ʑiɕʰa | ˈɡõ.sʰa dou ɕʰimijaʁˈʑas]
  Zuerezhuiza zalizhui shimiva hyalissa shimiva, ‘Nui Kounsou büza zhicha?’, gonsa dou shimiyarzhas.
  
  [za.my.jaʁˈva.ʑis.sʰɛ za.myˈva zvɛ ˈnas.sʰa ˈɡõ.c͡ɕʰy ɕʰiˈva dou ʑiˈʑɥigou | za ʑiˈɲa kʰvɛ]
  Zamüyarvajissa zamüva zve nassa gonchui shiwa dou zhizhuigou, za zhinya kve.
  
  

  Zvere-zhui=za

  wind-DEF.SG=CONJ

  zalid-ui

  sun=DEF.SG

  shimi-va

  3s.PST-do.IPFV

  hyalit=za

  argument=CONJ

  shimi-va,

  3s.PST-do.IPFV,

  ‘Nui

  ‘who

  koum=sa=u

  strong.NMLZ=ERG=Q

  büza

  all

  zhi-cha?’,

  3s.NPST-leave?’,

  
  
  

  gont=sa

  person=ERG

  dou

  cloak

  shimi-yar-zha-s.

  3s.PST-AND-have.PFV-DIM

  
  
  

  Zamü-yar-wajis=za

  3p.PST-AND-contact.PFV=CONJ

  zamü-wa

  3p.PST-do

  zwe

  if

  nak=sa

  someone=ERG

  gont-ui

  person=DEF.SG

  shiva

  CAUS

  dou

  cloak

  zhi-zhui-gou,

  3s.NPST-remove-VEN,

  za

  3s

  zhi-nya

  3s.NPST-COP

  kve

  strong

  
  

  I think this language is very nice now.

  As its creator I respectfully disagree.

  What's wrong with it? I think it sounds kind-of French. Besides, we both know you could do worse to one of mine.

[bradrn]

I think this language is very nice now.

As its creator I respectfully disagree.

What's wrong with it? I think it sounds kind-of French.

Well, it has /y/, and /dʑ/, and /ɥ/, and /cɕ/, and /ç/… basically, palatals and affricates up the wazoo. It makes it sound all mushy. (All the languages in this thread have required extensive demushification after you’ve gotten to them.)

Besides, we both know you could do worse to one of mine.

Oh, no ‘could’ there. I do do worse! I do!

[Rounin Ryuuji]

As its creator I respectfully disagree.

What's wrong with it? I think it sounds kind-of French.

Well, it has /y/, and /dʑ/, and /ɥ/, and /cɕ/, and /ç/… basically, palatals and affricates up the wazoo. It makes it sound all mushy. (All the languages in this thread have required extensive demushification after you’ve gotten to them.)

It isn't mushy! It's ~sparkling~. Now that I try reading it aloud, it sounds a bit like how Pinyin might lead you to believe Mandarin sounds (as opposed to how it actually does). [y] and the palatals are very bright, pleasant sounds. I don't fill it full of dark goopy vowels like [ə ʋ ʌ ɵ ɒ] or nasty glottal or gulpy retroflexes.

Besides, we both know you could do worse to one of mine.

Oh, no ‘could’ there. I do do worse! I do!

I haven't supplied a language for this thread (yet). You haven't played with one of my languages (that I know of) (yet).

If it's wanted, however...

[Otto Kretschmer]

Who agrees to this?

Proto Indo Anatolian 4000 BC

ɣweu̯is hku̯es-kwe

ɣweu̯is kwoi χu̯elhn̥ ne hest
hkums ɣwekwet;
to ˀdenso u̯ogom ugentm̥,
to mgeχ borom,
to u̯ihrom hohku brentm̥.
u̯ēukwt ɣweu̯is hku̯os:
“χedgor hme kērˀd,
χnerm̥ u̯iˀdenti hkums χˀgentm̥.”
u̯eukwn̥t hku̯es: “klu ɣwu̯e!
χedgor n̥sme kērˀd u̯iˀdenti,
χnēr, ʔesos, ɣwu̯i̯om χu̯lhenm̥
su̯e gwermom u̯esti kwr̥néuti,
ɣwu̯i̯om-kwe χu̯elhn̥ ne hesti.”
To keklu̯us ɣweu̯is pleχnom bēuˀgt.

If you do, do this and then a few more conlangs. It is easier to make a gloss in a language that one has crrated

[Karch]

It isn't mushy! It's ~sparkling~. Now that I try reading it aloud, it sounds a bit like how Pinyin might lead you to believe Mandarin sounds (as opposed to how it actually does). [y] and the palatals are very bright, pleasant sounds. I don't fill it full of dark goopy vowels like [ə ʋ ʌ ɵ ɒ] or nasty glottal or gulpy retroflexes.

You should! Deriving from ""sparkling"" languages gets boring very quickly, after all, and it's not like you're creating a conlang.

[bradrn]

It isn't mushy! It's ~sparkling~. Now that I try reading it aloud, it sounds a bit like how Pinyin might lead you to believe Mandarin sounds (as opposed to how it actually does). [y] and the palatals are very bright, pleasant sounds. I don't fill it full of dark goopy vowels like [ə ʋ ʌ ɵ ɒ] or nasty glottal or gulpy retroflexes.

*gasp* Blasphemy! [ə ʌ ʔ] and kin are not goopy; they are clean and sharp. There is a difference.

(I fully endorse your opinion of retroflexes, though.)

[Rounin Ryuuji]

It isn't mushy! It's ~sparkling~. Now that I try reading it aloud, it sounds a bit like how Pinyin might lead you to believe Mandarin sounds (as opposed to how it actually does). [y] and the palatals are very bright, pleasant sounds. I don't fill it full of dark goopy vowels like [ə ʋ ʌ ɵ ɒ] or nasty glottal or gulpy retroflexes.

You should! Deriving from ""sparkling"" languages gets boring very quickly, after all, and it's not like you're creating a conlang.

I was speaking facetiously, I did introduce reduced vowels into one stage of this one, in fact, but they were undone by somebody else somewhere. I don't actually start out with any idea of where I'm going with these — I just go with what suggests itself instinctually.

It isn't mushy! It's ~sparkling~. Now that I try reading it aloud, it sounds a bit like how Pinyin might lead you to believe Mandarin sounds (as opposed to how it actually does). [y] and the palatals are very bright, pleasant sounds. I don't fill it full of dark goopy vowels like [ə ʋ ʌ ɵ ɒ] or nasty glottal or gulpy retroflexes.

*gasp* Blasphemy! [ə ʌ ʔ] and kin are not goopy; they are clean and sharp. There is a difference.

(I fully endorse your opinion of retroflexes, though.)

A glottal stop is choking in phoneme form.

[Otto Kretschmer]

Bump
3500 BC
- Introduction of prothetic schwa into initial and final consonant clusters to break them

ɣəweu̯is kwoi χu̯elhən̥ ne hest
həkums ɣəwekwet;
to ˀdenso u̯ogom ugentəm̥,
to məgeχ borom,
to u̯ihrom hohku brentəm̥.
u̯ēukwət ɣweu̯is həku̯os:
“χedgor həme kērˀd,
χənerəm̥ u̯iˀdenti həkums χˀəgentəm̥.”
u̯eukwn̥ət həku̯es: “kəlu ɣəwu̯e!
χedgor n̥əsme kērˀd u̯iˀdenti,
χənēr, ʔesos, ɣəwu̯i̯om χu̯lhenəm̥
su̯e gəwermom u̯esti kəwr̥néuti,
ɣəwu̯i̯om-kəwe χu̯elhən̥ ne hesti.”
To keklu̯us ɣəweu̯is pleχnom bēuˀgət.

[Karch]

To finish the last challenge off (and to "desparkle" the language) I had some fun with the result of two rounds of Rounin:

1. y ɥi > wi
2. ɕw cɕw ʑw ɟʑw > xw kʰw ɣw ɡw
3. sʰ ɕʰ > h hʲ
4. ṼC > ṼNC
5. mf ns ŋh > m̥ n̥ ŋ̊
6. ç > x
7. V1ʁV1 Vʁ > Vˠ, ʁ > ɣ
8. v z ʑ ɣ > b dz dʑ ɡ
9. w disappears after consonants
10. velarized vowels receive stress, pretonic and posttonic vowels are deleted
11. tonogenesis: vowels before voiced consonants receive falling tone, vowels before voiceless consonants receive rising tone, voicing contrasts are neutralized, velarized vowels, alongside with the rest of the vowels become low
12. approximants and fricatives disappear word-finally and before a consonant

[ˈtspɛ̀ktsɛ̀ ˈtsâkì hʲîˈmpà ˈxjâsɛ̀ hʲîˈmpà | ŋì ˈkʰǒunòu ˈptsà tɕǐ | ˈkǒŋ tòu hʲîˈmjàtɕǎ]
<Tspektse tsàki hyìmpa xyàse hyìmpa, ngi khóunou ptsa chí, kóng tou hyìmyachá.>

[Rounin Ryuuji]

• [ɣ χ r] > [g ʁ̞ ʁ̞]
• Medial and terminal voiceless [m̥ n̥ r̥] > [p t x]; initially > [m n]
• A consonant followed by [u̯ w] becomes labialised, absorbing the glide; remaining [u̯] > [w];
• Coda nasals are lost; lowering, lengthening and nasalising the preceding vowel;
• Coda [h] disappears, lengthening and pharyngealising the preceding vowel; vowels following [h] also become pharyngealised;

[gəwewis kʷoi ʁ̞ʷelhəˤt ne heˤst]
Gèwewis kwoi rwelhèt ne hest

[həˤkõːs gəwekʷet]
Hèkôns gèwekwet;

[to ˀdãːso wogõː ugãːtəp,]
To 'dânso wogôn ugântèp

[to məgeʁ̞ boʁ̞õː,]
To mèger borôn

[to wiˤːrõː hoˤːku bʁ̞entəp.]
To wîhrôn hôhku brentèp.

[weːukʷət gʷewis həˤkʷos]
Wêukwèt gwewis hèkwos:

[ʁ̞edgor hˤəme keːrˀd]
Redgor hème kêr'd,

[ʁ̞ənerəp wiˀdãːti həˤkõːs ʁ̞ˀəgãːtəp]
Rènerèp wi'dânti hèkôns r'ègântèp."

[weukʷtət həˤkʷes | “kəlu gəwwe]
Weukwtèt hèkwes, "Kèlu gèwwe!

[ʁ̞edgor nəsme keːʁ̞ˀd wiˀdãːti]
Redgor nèsme kêr'd wi'dânti,

[ʁ̞əneːr | ʔesos, gəwwjõː ʁ̞ʷlhˤenəp]
Rènêr, 'esos, gèwwyon rwlh'enèp

[sʷe gəweʁ̞mõː westi kəwxneːuti]
Swe gèwermôn westi kèwkhnêuti

[gəwwjõː-kəwe ʁ̞ʷelhəˤt ne heˤsti]
Gèwwyôn-kèwe rwelhèt ne hesti."

[To keklʷus gəwewis pleʁ̞nõː beːuˀgət.]
To kekwus gèwewis plernôn bêu'gèt.

[Karch]

1. retranscribed the text in IPA, thus undoing some of the previous changes - labialized consonants are no longer broken up
2. more schwa epenthesis in consonant clusters, w j > u i / C_C
3. e e: > a e
4. a > ɔ after labio-velars, o > ɔ after uvulars

ɣʷɔwis kʷoi χwɔlhən na hast
həkums ɣʷɔkʷɔt;
to ˀdanso wogom ugantəm,
to məgaχ borom,
to wihrom hohku brantəm.
weukʷət ɣʷɔwis həkwos:
“χadgor həme kerˀd,
χənarəm wiˀdanti həkuməs χəˀgantəm.”
waukʷnət həkwas: “kəlu ɣʷəwa!
χadgor nəsma kerˀd wiˀdanti,
χəner, ʔasos, ɣʷujom χulhanəm
swa gʷɔrmom wasti kʷərˈnauti,
ɣʷujom-kʷɔ χwɔlhən na hasti.”
to kakləwus ɣʷɔwis plaχnom beuˀgət.

EDIT: fuck, ninja'd