Reconstruction talk:Proto-Japonic/mi-

Bad pagename

@Mellohi! To lemmatize Proto-Japonic verbs with 終止形 suffix -u is very bad. There are vowel stems and consonant stems in the verbs of Japonic languages, and they have uttery different phonological histories and inflections. 終止形 of every Ryukyuans are not derived from with -u. Correct pagename shold be mì-.--荒巻モロゾフ (talk) 09:15, 30 April 2019 (UTC)

@荒巻モロゾフ: Both Japonic branches use the infinitive -i forms, correct? I could move them there instead, in addition to moving all Proto-Ryukyuan verbs to the infinitives as well. The reason why I didn't leave it at a root is in analogy with Indo-European languages, which also have "utterly different phonological histories and inflections" yet it and its descendants lemmatize at different forms each. mellohi! (僕の乖離) 13:39, 30 April 2019 (UTC)

終止形 [terminal form] of Okinawan /-(j)uɴ/ is derived from the cognate parts of 連用形 [infinitive form] + 居り (wori, to be) in the Mainland Japanese. On the other hand, verbs of Miyako have multiple origin of the 終止形 including cognate parts of 連体形 [Attributive form] of Old Japanese and the specific forms[1]. Therefore there aren't any unified protoform of 終止形 of Ryukyuan and if you want to make the articles of Proto-Ryukyuan verbs, it's good to make the title from the word stem. Old Japanese inflection can be analyzed like as following table (note that ablaut of /i/ into /u/ indicated by **ï).

	未然 **-(a)	連用 **-i	終止 **-(r)ï	連体 **-(ïr)o	已然 **-(ïr)ai	命令 **-iə
四段 **-φ-	/-a/ < *-a	/-i1/ < *-i	/-u/ < *-u < **-ï	/-u/ < *-o	/-e2/ < *-ai	/-e1/ < *-e < **-iə
上一段 **-i-	/-i1/ < *-i	/-i1/ < *-i	/-i1ru/ < *-iru < **-irï	/-i1ru/ < *-iro	/-i1re2/ < *-irai	/-i1(yo2)/ < *-i(jə) < **-iiə
下一段 **-ui-	/-i2/ < *-ui	/-i2/ < *-ui	/-u/ < *-u < **-uï	/-uru/ < *-uro < **-uïro	/-ure/ < *-urai < **-uïrai	/-i2(yo2)/ < *-ui(jə) < **-uiiə
下二段 **-ai-	/-e2/ < *-ai	/-e2/ < *-ai	/-u/ < *-u < **-aï	/-uru/ < *-uro < **-aïro	/-ure/ < *-urai < **-aïrai	/-e2(yo2)/ < *-ai(jə) < **-aiiə
カ変 **kjə-	/ko2/ < *kə < **kjə	/ki1/ < *ki < **kjəi	/ku/ < *ku < **kjəï	/kuru/ < *kuro < **kjəïro	/kure/ < *kurai < **kjəïrai	/ko2/ < *kə < **kjiə
サ変 **se-	/se/ < *se	/si/ < *si <**sei	/su/ < *su < **seï	/suru/ < *suro < **seïro	/sure/ < *surai < **seïrai	/se(yo2)/ < *se(jə) < **seiə
ナ変 **-nj-	/-na/ < *-na < *-nja	/ni/ < *-ni < **nji	/-nu/ < *-nu < **-njï	/-nuru/ < *-nuro < **-njïro	/-nure/ < *-nurai < **-njïrai	/-ne/ < *-ne < **-njiə

The vowel stems basically can be interpreted as the 未然形 [irrealis form] followed by the suffixes. Also in Okinawan language, 未然形 is the basis of the inflection, so the same stems are usable.--荒巻モロゾフ (talk) 09:04, 1 May 2019 (UTC)

If anyone has any idea where @荒巻モロゾフ sourced these reconstructions, I would be most interested in learning the title(s).

I also see apparent mistakes in the table above: the 下一段 (shimo ichidan) conjugation paradigm is shown with stems ending in /i₂/, but historically, we have instead stems ending in /e/. This paradigm does not appear until the Heian period, and solely for the earlier form of modern verb 蹴る (keru, “to kick”), and thus we have no Old Japanese 甲・乙 vowel distinctions on record. This seems to be a shift from older くう (kuu), in turn treated as 下二段 (shimo nidan) in the "w" row as kuwu, with stem form kuwe-, then fusing into kwe- and then ke-. At any rate, I am not aware of any /i₂/ stem for 下一段 (shimo ichidan) verbs as shown in the table above.

... Looking at this row more closely, I think @荒巻モロゾフ made a typo in the row label -- I think that's supposed to be 上二段 (kami nidan) instead. ‑‑ Eiríkr Útlendi │^{Tala við mig} 23:21, 25 March 2024 (UTC)

@Eirikr: I think this is Aramaki's original research. **/ï/ is a totally ad hoc reconstruction; it is not supported by any sources other than Old Japanese. (IIRC there's a special Eastern old japanese paradigm of ko- 'to come')

(Sadly,) many serious Japanese linguists focus merely on noun prosody. IIRC, linguists like Thomas Pellard and Yosuke Igarashi claim that the split of 2.4/5 nouns (tone class B and C mix) in Ryukyuan goes back to Proto-Japonic, but the precise reconstruction of the pJ accent is unknown. Many scholars believe that the Heian period Kyoto accent goes back to Proto-Japonic, with interpretations of a high dot tone as a high pitch and a low dot tone as a low pitch. However there are opponents to the standard theory: Samuel Ramsey and Elisabeth de Boer (rest in peace to de Boer) reverse the tones, and hypothesizes that there was dialectal interference in Ryukyuan that caused this split. Chuterix (talk) 23:35, 25 March 2024 (UTC)

Should it be *miri?

@Mellohi! I am sure that Proto-Japonic does not allow vowel clusters outside diphthongs, but would that be a good reconstruction for the infinitive? Kwékwlos (talk) 21:06, 28 May 2019 (UTC)

@Kwékwlos: We could move the verb articles to the roots (e.g. in this case *mi-) to avoid confusing ourselves any further. mellohi! (僕の乖離) 02:10, 29 May 2019 (UTC)

Related

This verb root is also likely closely related to noun *ma (“eye”) (listed, I think erroneously, at *may).

(Incidentally, there are currently a very many noun roots in Category:Proto-Japonic_lemmas that include a final -y, which itself likely represents a fused particle or other derivational suffix, given the plentiful examples of OJP and JA compounds including the forms without final -y.)

‑‑ Eiríkr Útlendi │^{Tala við mig} 17:49, 11 June 2019 (UTC)

To express /j/ as <y> is not suitable for Proto-Japonic. Final consonants are not recognized in Proto-Japonic. Difference between 被覆形 (embedded form) and 露出形 (exposed form) should be represented as a hiatus in the hyphenated (?) form (in the case of 目(め) (me), it become *mai 2.3 ). Some words include the vowels presumed to be generated from the continuation of the another vowels. (e.g. 嘆(なげ)き (nageki, OJ: <nage₂ki₁> ← *nanka II (長(なが) (naga)) + *iki 2.4a (息(いき) (iki)); <e₂> ← *ai), 数(かぞ)える (kazoeru) ← 数(かぞ)ふ (kazofu, OJ: <kazo₁pu (kazo₁pe₂-)> ← *kansu 2.4x (数(かず) (kazu)) + *ap-ai- II (和(あ)ふ (afu) → 和(あ)える (aeru)); <o₁> ← *ua) and けり (keri, OJ: <ke₁ri> ← *k((C)V)i- II (来(き) (ki)) + *ar-i II (有(あ)り (ari)); <e₁> ← *ia)--荒巻モロゾフ (talk) 15:49, 12 June 2019 (UTC)

FWIW, I agree with 荒巻モロゾフ on this -- the final -y notation seems ill-suited to Japonic. ‑‑ Eiríkr Útlendi │^{Tala við mig} 16:17, 12 June 2019 (UTC)

@荒巻モロゾフ, what is your opinion at lemmatizing at the 被(ひ)覆(ふく)形(けい) (hifukukei)? My understanding (perhaps flawed?) of the academic literature is that these are the core forms, and that the 露(ろ)出(しゅつ)形(けい) (roshutsukei) represent a fusion with a particle or some other suffixing element. If the 露(ろ)出(しゅつ)形(けい) (roshutsukei) are indeed derivational, wouldn't it make more sense to lemmatize at the root non-derived form? ‑‑ Eiríkr Útlendi │^{Tala við mig} 16:17, 12 June 2019 (UTC)

Some researchers suppose existence of Pre-proto-Japonic language (日本曾祖語). Some reconstructed words of Proto-Japonic don't follow the expected relatives (like *mentu 2.1 (水(みづ) (midu) → 水(みず) (mizu)) vs *mi(-na) 2.1?) (水(み) (mi), 水(みな) (mina)), *kə(-i) 1.3 (木(き) (ki), 木(こ) (ko)) vs *ku-nta-mənə 4.5 (LLLL) (果物(くだもの) (kudamono)), *kir- II (切(き)る (kiru)) vs *kensu 2.1 (傷(きず) (kizu)) etc.). If you want to pursue the origin of those words, lemmatizing at the root non-derived form might make some senses.--荒巻モロゾフ (talk) 18:14, 12 June 2019 (UTC)

AFAIK all western literature I've seen reconstruct the roshutsukei forms for Proto-Japonic. As for the -y notation, I'll concede on that one - only Vovin seems to use it consistently. mellohi! (僕の乖離) 16:32, 13 June 2019 (UTC)

Hmm, hmm, interesting. If we keep the 露出形 for Proto, I think we should add the 被覆形 -- while we can easily derive modern me from proposed Proto ma-i (or however we ultimately decide to format that) via simple vowel fusion, a common process in languages around the world, we cannot explain the compounding 被覆形 unless we also posit that the 被覆形 similarly existed in Proto. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:50, 14 June 2019 (UTC)