Proto-Language Reconstruction Thread

[CapriceFrenata]

I have no idea if this was on the old forum or not, but I thought it'd be fun to try. The following five words in fifteen languages have been derived from words generated using one of the gen.html presets. Derivation was inspired by Indo-European. Their order has been randomized.

베디너 업레네더 두디딕 레밉러 싯룰
pennin ablonnan torredigo rammib simmurro
pehantine abonahanto tiudreotuiucuis ramiuboe sosmeosror
pazzinë åblånnazzë dutrerigë rammiblë remmërror
pendin ablonnand tidretics rammibyl sormyrror
պիդենե աբլոնադե դիտրեդեգեզ րամիբլե զեմլոր
ペㇴチニ アㇷ゚ロ ㇺナㇺチ ト゚ㇳレチカㇱ ラㇴミㇷ゚リ シㇱマㇱルㇽ
ペンキン ンロナメン ツケキン ラミビイ ヒマル
pennine aplonanse toredige ramivle shimsler
𐌘𐌄𐌄𐌍𐌕𐌉𐌉𐌍𐌉 𐌀𐌀𐌐𐌋𐌏𐌏𐌍𐌀𐌀𐌍𐌕𐌉 𐌈𐌀𐌏𐌕𐌓𐌄𐌄𐌕𐌉𐌉𐌊𐌀𐌔 𐌓𐌀𐌀𐌌𐌉𐌉𐌁𐌋𐌄𐌄 𐌔𐌉𐌉𐌔𐌌𐌀𐌔𐌓𐌏𐌓
바아덴 어볘에노옫 도뒈데것 로오미버 센소오
pfenzine apfonnanze tuzrezike ramible sosmesler
ペセネ エポルンナメセ トツレセケ ランミベ セマル
φιτιν αφλυνατ θυθρεθικς ραμιυλ σεμλλο
ピチニ エペロナメチ ツテリエチケシ ラミベヂ シシマシルハ

[Tropylium]

Don't feel like transcribing the Korean/Japanese/Armenian rn, but let's start by roughly aligning the rest and sketching a rough initial reconstruction. Greek/Old Italic has been mostly translitterated.

Code: Select all

4.  p  a zzinë   åb lånn a zzë   du tre r i gë     rammi b lë    r emmë rror
-----
5.  p  e ndin    ab lonn a nd    ti dre t i c  s   rammi byl     s ormy rror
2.  p  e nnin    ab lonn a n     to rre d i go     rammi b       s immu rro
9.  p  e nnine   ap lon  a nse   to  re d i ge     ram i v le    shi m  sler
---
3.  p ehantine   ab  on ahanto   tiudreotuiucuis   ram iub  oe   s osmeosror
---
12. pf e nzine   apf onn a nze   tu zre z i ke     ram i b le    s osme sler
10. f  ē ntīni   āp lōn  ā nti   θaotrē t ī ka s   rām ī b lē    s īsma sror
14. f  i  tin    af lun  a  t    θu θre θ i k  s   ram i u l     s e m  llo
   #p  a̤ ntine  #ap lon  a̤ nte  #teutre t i kV s  #ram i b le   #s esmV sror

Segments clearly evident: *p *b *t *k *m *n *s *r *a *e *i *o. #l in #-pl, #-bl- could be an allophone of *r, given some cases of apparent *sr > *sl. #pa̤ntine seems to have widespread umlaut of *a to /e/, perhaps blocked in #ramible by *-m-. Vowels in #sesmVsror look very unstable.

Tentative subgrouping: 4 is an outgroup distinguished by lack of a-umlaut, *s- > r-, *nt > zz. 2+9 share several lenitions: *tr > *dr > *rr, *nt > nn and *t *k > d g / V_V. 5 has *tr *nt > dr nd, and apocope similar to 2, and could be a sister of these. 10+12+14 show aspiration/frication of *p *t; 14 could be just about even derived from 10.
3 clearly has "too many" vowels and probably unpacks them into diphhongs thru to some Irish-esque palatalization scheme. I've tentatively reconstructed the source of VhV as a breathy voiced vowel #a̤.

[Bessunire]

I tried my hand at transliterating the katakana. I'm not sure how well 6, 12, and 15 turned out, and I couldn't figure out 7 at all, since the font makes it pretty much impossible to tell the difference between -n and so. I added 6, 12, and 15 to the bottom of the chart, with the same formatting, followed by proposed changes to the proto-language reconstruction.


4. p a zzinë åb lånn a zzë du tre r i gë rammi b lë r emmë rror -----

5. p e ndin ab lonn a nd ti dre t i c s rammi byl s ormy rror
2. p e nnin ab lonn a n to rre d i go rammi b s immu rro
9. p e nnine ap lon a nse to re d i ge ram i v le shi m sler ---

3. p ehantine ab on ahanto tiudreotuiucuis ram iub oe s osmeosror ---

12. pf e nzine apf onn a nze tu zre z i ke ram i b le s osme sler
10. f ē ntīni āp lōn ā nti θaotrē t ī ka s rām ī b lē s īsma sror
14. f i tin af lun a t θu θre θ i k s ram i u l s e m llo

7. p e nuchini apu romu namu chi ttore ch i kashi ranumi pu ri sh i mma rruru
13. p e sene epo runa mese to tsure s e ke ranm i pe s e ma ru
15. p i chini epe rona chi tsu teru echikushi rami b eji shi mma rruha


#p a̤ ntine #ap lon a̤ nte #teutre t i kV s #ram i b le #s esmV sror

#p a̤ ntine #apV lon a̤ mVte #teutVre t i kV ʃ #ranm i b le #s esmVsror

It's hard to tell if a vowel actually exists or if it is because of the CV nature of most Japanese syllables, but I believe that the vowels after p in the 2nd word and between t and r in the 3rd do exist, since they are different qualities in each katakana language, despite being missing in the others. I think the i at the end of the 3rd word in 7 and 15 does not actually exist, but indicates a pronunciation of /ʃ/. Another possibility is *s > ʃ in 7 and 15.

It seems that 7, 13, and 15 have all undergone palatalization of *t before *i, with 13 at a more advanced stage than the other two. It's harder to tell, however, whether or not *teu > tsu in 13 and 15, since there is no katakana for tu. It's also unclear whether or not *l > r in these languages.

A possible explanation for the weird shape of the 2nd word in 6 is *na̤mVt > *na̤mVm > *na̤nVm > *ma̤nVm. 7 Also seems to have *nC > *nuC, as well as devoicing of voiced stops. In 15 there seems to be *mVt > *nVt.

15 also presents the problem that I have no idea what to do with the final syllables of ramibeji and shimmarruha. The middle of tsuteruechikush(i) is problematic as well.

[mèþru]

I assume you meant 13 instead of 12 for some of the times you said 12.

Korean phonetic / romanisation:
1. pe.di.nʌ ʌb.ɾe.ne.dʌ tu.di.dik̚ ne.mim.nʌ sis.nʌ / bedineo eoblenedeo dudidig lemibleo sislul
11. pa.a.den ʌ.bje.e.not̚ to.dwe.de.gʌt̚ no.o.mi.bʌ sen.zo.o / baaden eobyeenood dodwedegeos loomibeo sensoo

Number 7 appears to be Ainu, but ㇴ is never used except in the end of a word in Ainu so I'd just leave it until commented by the OP.

Since no one put up #8:
penkin nronamen tsukekin ramibii himaru

[Bessunire]

I assume you meant 13 instead of 12 for some of the times you said 12.

Whoops. My numbers are off. 6 and 12 should be 7 and 13. Thanks for the catch!

[mèþru]

#6:
pidene ablonade ditredegez ramible zemlor

[Kawdek]

Here's my attempt at rectifying the katakana samples by making a few significant assumptions, including some large kana being used for non-syllabic consonants (like e-dan for 15 generally), and that チ and ツ represent an etymological *ti and *tu (except in 7 since ト゚ most likely represents /tu/ so チ may still be /tɕi/, which I've chosen to write tsi to preserve the relation with /t/):
7. pentsini apromnamtsi tutretsikas ranmipri sismasrur
8. penkin nronamen tukekin ramibī himaru
13. pesene eprunnamese totsreseke ranmibe semaru
15. pitini epronamti tutrietikeš ramibdi sišmašruha

This is how I believe they all align:

Code: Select all

1	p e   tinë	ëp re ne    të	t u t  i t i  k		re mi p rë	s i    srur
2	p e  nnin	ab lonna  n	t o r re d i  go 	rammi b		s immu rro
3	p ehantine	ab  o nahan to	t iud reot uiucuis	ra miub  oe	s osmeosror
4	p a  zzinë	åb lånna  z zë	d u t re r i  gë 	rammi b lë	r emmë rror
5	p e  ndin	ab lonna  n d	t i d re t i  c  s 	rammi byl	s ormy rror
6	p i   dene	ab lo na    de	d i t re d e  ge z 	ra mi b le	z e m   lor
7	p e  ntini	ap romna  m ti	t u t re t i  ka s	ranmi p ri	s isma srur
8	p e  nkin	 n ro na  men	t u k  e k i  n 	ra mi b  ī	h i ma  ru
9	p e  nnine	ap lo na  n se	t o   re d i  ge 	ra mi v le	shi m  sler
10	phē  ntīni	āp lō nā  n ti	thaot rē t ī  ka s	rā mī b lē	s īsma sror
11	p ā   ten	ëp yē nō    t	t o t we t e  kë s 	nō mi p  ë	s e n  s ō
12	pfe  nzine	apf onna  n ze	t u z re z i  ke	ra mi b le	s osme sler
13	p e   sene	ep runna  mese	t o tsre s e  ke 	ranmi b  e	s e ma  ru
14	phi   tin	aphly na    t	thy thre thi  k  s 	ra mi  yl	s e m  llo
15	p i   tini	ep ro na  m ti	t u t riet i  ke š	ra mi b di	s išma šruha

This is my first attempt at a reconstruction so I might be misguided with my presuppositions, but my proposal for the rough reconstruction would be:
*pēHntine aphlōmnāHmte teuthretikas rānmible sesmVsrur