User talk:Kwékwlos/Possible Japonic relationship with Indo-European

Some quick feedback

I too am fascinated by this kind of stuff. Here's some ideas.  :)

• abura

Deverbal from aburu "to roast". Compare also tsuka "grip; mound" from tsuku "to put one's hand to; to build up", among other nouns attested since olden times that end in -a and have close connections to nearly-homophonous verbs ending in -u.

• aru

I've read somewhere (can't recall at the moment) that the ari form attested in OJP is probably an innovative shift from even-older aru. Dunno if that has any impact on your reconstruction.

• pane

Generally listed in monolingual JA resources as an ancient compound of pa ("leaf, feather") + ne ("root"), probably originally in reference to the quill that is the root of a feather.

• paru (verb), piro (adjective stem)

You've got both arising from the same purported PIE root. I didn't see any attempt at explaining the vowel variation?

• pukeru (to steam)

Not attested until the 1600s. Given the semantics, this appears to be a derivation from root puku, "to plump up", cognate with 袋 (fukuro, “bag”), 膨れる (fukureru, “to swell”), etc.

• pukureru

See also the comments at User:Chuterix/Chuterix's_Proto-Japonic_Reconstruction_Dictionary_Project/Proto-Japonic_words_possibly_borrowed_from_other_Proto-Languages?#fukureru -- this is a derivation from a verbal root puk-.

• fureru ← puru

The sense of "to experience something" isn't borne out by Japanese sources. The oldest attested meaning is more like "to come into brief contact with".

• furu ← puru

Unclear how identical JA realizations for both 触る (furu, “to come into brief contact with”) and 降る (furu, “to fall”) would have such divergent PIE roots. There is also 振る (furu, “to shake; to sway; to swing”), which might be related to one or both of these other words.

• kage

Note that Old Japanese has several senses for this: "shadow", and also "light", as well as "image" and "outline". See also the KDJ entries here for "shadow", and here for "light; image; outline".

By way of derivatives from the "image; outline" sense, see also 鏡 (kagami, “mirror”), from this same kaga- + 見 (mi, “seeing”).

• kawa

Unclear why this has "crab" under the "PIE" column? Perhaps a copy-paste error?

• kuruma

JA sources derive this from kuru-, adverbial root related to koro- and having to do with "rolling; spinning; turning" + ma, possibly a shift from 輪 (wa, “ring; wheel”) or even 馬 (uma, “horse”) from the idea of "horse-drawn carriage".

FWIW, I have not seen any Japonic evidence of any form karuma.

• mago

The -go ending is clearly 子 (ko, “child”). Note too the existence of older form umago. See also the KDJ entry here.

• mazeru

This is from older mazu, and there are also the modern intransitive verbs 混じる (majiru) and 混ざる (mazaru, “to be or become mixed into something else”). Given that majiru conjugates based on stem majir-, that particular form might be a fusion of mazi- + 入る (iru, “to go in”).

• mozu

Given the probable correctness of the internal reconstruction as a contracted compound from 百 (momo, “one hundred; many, lots, varied”) + す (su, ancient suffixing element appearing in bird names), this is highly unlikely to be from any PIE term. Moreover, I cannot find any Ryukyuan cognates, making any Proto-Japonic reconstruction problematic: no such Ryukyuan entries in evidence at JLect, nor in Thorpe's 1983 paper.

• ni, nita

I can find no evidence for any OJP term nita meaning "earth, soil". Best I can find is nita as an adjective meaning "soft and mushy or viscous". KDJ entry.

Postscript -- the -ta in OJP nita might be an ancient and rare adjective-forming suffix indicating state. See also 新 (arata). If so, the root term then would be ni, attested in OJP as meaning earth, soil, more specifically ocher, and by extension, the color red. ‑‑ Eiríkr Útlendi │^{Tala við mig} 18:02, 29 January 2024 (UTC)

@Eirikr: This is also supported by the fact that we have nitasi in WOJ (Western Old Japanese) as well.

Speculatively, is this related to つ (tu, completive marker)? Chuterix (talk) 18:07, 29 January 2024 (UTC)

• "This is also supported by the fact that we have nitasi in WOJ (Western Old Japanese) as well."

Ya, that's a straightforward addition of adjective-forming suffix -shi to nita.

Interestingly, both forms nita and nitasi are attested in the Izumo Fudōki dated to 733 -- see also w:Fudoki#Manuscripts.

• "Speculatively, is this [adjective-forming suffix -ta] related to つ (tu, completive marker)?"

Not sure. Possibly?

If we parse OJP -tu less as "completive" and more as a stative, which for verbs could mean "the action has begun", then this would fit better for forming adjectives, since those are states and have no "completedness". As a verb suffix, this could also indicate a completive sense for any instantaneous-action verb, and an ongoing one for any repetitive / iterative / ongoing-action verb. And, this would fit with (I think it was?) Frellesvig's proposed bipartite ancient copular elements, one starting with n- and the other with t-.

Specifically as form ta, this might be a similar kind of "resultative" ending in -a, as we see with ancient verb-derived resultative nouns, like tuka or abura, etc.

‑‑ Eiríkr Útlendi │^{Tala við mig} 00:58, 30 January 2024 (UTC)

• na- "not"

In the na-X-sö construction, you mention that "X can be any verb." This isn't quite accurate -- that verb must be in the 連用形 (ren'yōkei), the continuative / stem / adverbial / nominal / infinitive form of the verb (different English-language authors use different terminology to describe this verb form). One way to parse this construction in English might be "without [VERB]-ing".

As a side note, that Twitter link demonstrates some confusion on the part of the poster. Negative auxiliary -zu is reconstructed from -anisu, which makes any introduction of negative prefix na- problematic. We do know pretty clearly that -zu is not from -na-su "not do", and we do know pretty clearly that -nu is part of this cluster, as this still surfaces as a conjugation form, specifically the 連体形 (rentaikei, “adnominal”), and there is also an attested -ne form for the 已然形 (izenkei, “realis”), such as in the "because this didn't happen" ending -neba. See also the ず entry at Kotobank.

• nii "new"

Just a row above, we have PIE ne → PII ná → Japonic na. But now in this row, we have PIE né → PII ná → Japonic ni. This seems phonologically confused.

• okiru

How do you account for the Japanese verb cluster okiru, okoru, okosu? If the stem of the underlying root is oki-, how do you account for the Old Japanese predicate form oku, or adnominal form okuru? And how does any root oki- compound in a way to produce endings -oru and -osu?

• orochi

JA sources generally list this as oro + chi "mystical power", c.f. 霊 and the chi in 力 (chikara, “power, strength”). I also see 尾ろ (oro, from older woro) as a lexeme meaning "tail", possibly "tailed". See also KDJ entry for 尾ろ, other KDJ entry for phrase containing 尾ろ. Compound woro-chi would thus mean "tailed mystical power".

• shimo

See Reconstruction talk:Proto-Japonic/simə -- the simə reconstruction is mistaken. Old Japanese texts spell this as ⟨simo₁⟩, which could only come from Proto-Japonic simo.

• tsuzumi

The Encyclopedia Nipponica article here at Kotobank suggests this might be a borrowing from an Indian term resembling dudubhi or dundubhi, much as listed in the table here but described as a borrowing rather than cognate.

• Yamato

The yama here is highly unlikely to be from PIE Yemo "man", given the complete lack of any other yama morphophonemic unit in Japanese with this meaning.

As a comparandum, consider also the term Hayato, another ancient polity in Japan.

The term 人 (hito, from ancient pito₂; "person") appears as suffixing element -to, making it more likely that Yamato is yama ("mountain"?) + -to ("person, people").

• Yomi

If we're going to posit that PIE Yemo means "person" in Yamato, but somehow yields "death" in Yomi, a lot more explanation is needed.

That's all I have time for at the moment. HTH! ‑‑ Eiríkr Útlendi │^{Tala við mig} 21:23, 15 February 2023 (UTC)

Added several more items above, maintaining the ordering from your table. ‑‑ Eiríkr Útlendi │^{Tala við mig} 00:33, 16 February 2023 (UTC)

Sound changes (and response)

Added sound changes with explanations based on my theory. May be expanded. Kwékwlos (talk) 02:14, 28 February 2023 (UTC)

abura: I propose that the verb was derived from the noun, rather than vice versa.

As a general trend for word formation, Japanese is much more verb-y. I'm not sure I can think of any cases where a verb derives unambiguously from a noun, without the use of an auxiliary like する (suru). ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

ari: Final -u must have been a generalized conclusive stem that replaced -i.

I've read elsewhere that conclusive form ari was a prehistoric (pre-writing) innovative shift from aru. There was a grammatical pattern in Old Japanese that used the continuative / gerundive conjugation of a verb to conclude a sentence, and given the high frequency of use for the copular verb, this could have provided the impetus for shifting the conjugation paradigm. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

pane: If a compound, then the form pa came from the loss of an unstressed nasal syllable.

Why? Japanese has other monosyllabic terms. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

paru/piro: paru is assumed to derived from the full-grade *pleth₂- > *pratH- > *par-, while piro is from the zero-grade *pl̥th₂- > *pr̥tH- > *pir-.

Verb paru isn't about "becoming flat", though -- it's about "becoming taut". C.f. derived verb 腫れる (hareru, “to become swollen”), including for pocks or boils or similar protrusions from the body that are decidedly not flat. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

pukeru: Ryukyuan reflexes support PJ *pukasu. Still, I strongly believe this is a newer word.

pukureru: If puk-, I assume this was earlier *p(ʰ)uHt- + *kar-/*kr̥- "to do", so they should be related.

fureru ← puru: This one is somewhat doubtful if from *bʰer-. Maybe some other PIE root would be valid, perhaps *per-/*pr̥- "to strike" (compare Proto-Slavic *perti (“to press, push”)).

furu ← puru: I don't see a likely semantic connection likely between these words, as IE prefers polysemy with "fall" and "fly". Still, even if they come to have the same form, they must have phonologically converged from distinct roots.

Why "must have"? Consider English flutter, which is semantically related to ideas of "falling" and "swaying", and as something flutters, it may "come into brief contact" with something. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

kage: The primary meaning is "shade, shadow", but phonologically it is complicated.

Who is saying that the primary meaning is "shade, shadow"? One of the oldest texts in Japanese, the Man'yōshū, uses the word to mean "light", as of the sun or moon. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

kawa: Fixed. Indeed a copy-paste error.

kuruma: See https://www.jlect.com/entry/1084/kuruma/ for karuma. kuru can be derived from *kʷel- and that wouldn't affect its cognancy.

Exploring any PIE connection via kuru, rather than kuruma, seems much less problematic. Some of the attested daughters of Proto-Indo-European *kʷel- include vowels shifted to /o/, aligning with Japonic root /koro/ of similar "rolling, turning" meaning. ‑‑ Eiríkr Útlendi │^{Tala við mig} 21:53, 13 March 2023 (UTC)

mago: Ryukyuan forms have < *-ga (with the exception of Yonaguni) and I suspect this was a remodeling of the final vowel to agree with ko "child", if the Yonaguni form is not an archaic retention.

As further evidence of a Japonic compound of ma + ko, there is the Old Japanese term mima, a compound of honorific prefix 御 (mi-) + the ma from mago. C.f. 御孫 (mima). ‑‑ Eiríkr Útlendi │^{Tala við mig} 21:41, 13 March 2023 (UTC)

mazeru: If we start with mazu, one would have to explain why common Ryukyuan cognates of modern -eru have forms like Okinawan -iin/-iyun. Still, the base verb should start from a reduplicated zero-grade form with an irregular *a.

We see other verbs with base forms -u that also have -iru derivatives, such as 生く (iku) → 生きる (ikiru). Speculatively, there might be a continuative / iterative / stative sense added by this conjugation shift. This would appear to fit for the difference between mazu / mazeru ("to mix", momentary / instantaneous) and majiru ("to be mixed", stative). ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

mozu: Even if there are no Ryukyuan cognates, it's still possible that the term is Proto-Japonic (given an early attestation in Japanese), but was lost in the Proto-Ryukyuan language. As for momo + su, this would require irregular syncope of the second syllable. In addition, the Korean sae should be connected to sagi (likely by loanword relationship) and not -su.

Granted about the unlikely Korean connection, I'll remove that from the モズ entry later. That said, there is a su element that appears in many ancient names for birds: uguisu ("nightingale"), karasu ("crow"), hototogisu ("little cuckoo"). Re: syncope, there are examples of historical Japanese where mi or mu + su contracts to nzu. And the prevailing consensus for OJP is that voiced consonants were pre-nasalized. While momo → mon is irregular, it is also not entirely unlikely, especially considering that the high pitch / accent was on the initial mo. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

ni, nita: Thorpe (1983) has the gloss "soggy ground".

Thank you for the lead, I'll look that up later. So far, this isn't quite corroborated by Japanese sources I've seen, FWIW. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

na-: I agree, but given attested Japonic evidence *na- would have been prefixed, instead of suffixed as in the modern languages. The vowel variance may have been caused by the process in which the negative marker was grammaticalized as a verb.

nii: The change of *a to *i is puzzling, but I think this would be caused by sporadic metaphony, as in "niwa".

okiru: I propose oki- < okwi < *əke < *əkə-y < *əkə-ri, oko- < *əkə-, oku < *ək-u < (?) əkə-u, and okuru < *ək-uru < (?) *əkə-uru, possibly implying a stem *əkə-. The sequence *-ršti became identical to *-ri, so it was absorbed, leading to *əkə-.

orochi: There are various theories. But this one would be a strong contender, given the mythological similarities to the Indo-Iranian and Indo-European versions.

shimo: I agree.

Cheers! ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

tsuzumi: If a borrowing, it would have to be via Buddhist meditation. Yet in Sanskrit this is found in the oldest texts (Rigveda) and is a widespread word, being found in Ryukyuan with regular reflexes, which gives some support to the Central Asian theory.

Buddhism came to mainland Japan in the 600s. Depending on the state of contact between the Ryukyuan islands and mainland China, Buddhism may have come around a similar date. That gives plenty of time for diffusion and localized phonetic innovations. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

Yamato: It is highly unlikely for pito to simplify to to in Japanese. If it really did, compare karyūdo "hunter" where the long vowel ū preserves a trace of the first syllable. Perhaps instead, we could propose something such as "place of Yama" being generalized into an ethnonym, with the literal etymological meaning subsequently lost. For yama "mountain", given the diversity of semantics, a semantic shift "place where Yama (underworld) resides" > "forest; grove, mountain" can be theorized.

I'm not sure why you view this as "unlikely". Monolingual JA sources list to as a reading for 人, derived as a contraction of hito (ancient pito). See the KDJ entry here at Kotobank, listing five compounds with this same to element. ‑‑ Eiríkr Útlendi │^{Tala við mig} 19:38, 13 March 2023 (UTC)

Yomi: *Yemos is a proper noun, not a common noun. Also, the vowel variation has usually been explained as apophony and may have preserved the earlier semantics given above. Kwékwlos (talk) 02:14, 28 February 2023 (UTC)

Some feedback

@Kwékwlos. lso pinging @Eirikr.

• kapara 3.1 'dry riverbed with sand' (in Ryukyuan it exists as a word meaning 'river'; pR *kawa A means 'water well'). Curious if 河原 is a contraction of *kapa-para 'water valley', or this is actually a phonetic spelling? -ra 'pluralizer or nominalizer' wouldn't make sense.

Also, what is Robbeets (2005)? Chuterix (talk) 01:18, 27 January 2024 (UTC)

We don't know for sure. Robbeets (2005) = Is Japanese Related to Korean, Tungusic, Mongolic and Turkic? The Altaic hypothesis is best mentioned as a contact area. Kwékwlos (talk) 01:29, 27 January 2024 (UTC)

@Chuterix — Re: kawara, various JA monolingual references explicitly source this from older かわはら, in turn from かは + はら: Old Japanese 川・河 (kapa, “river”) + 原 (para, “field, plain”). See also Daijisen at Weblio, or the Nikkoku at Sakura-Paris. Regular haplology would reduce kapa para to just kapara. ‑‑ Eiríkr Útlendi │^{Tala við mig} 01:52, 27 January 2024 (UTC)

The Nihon Kokugo Daijiten 2 on JapanKnowledge deleted the etymology, and lists no etymological theories. Forgot to check those other sources; thanks! Chuterix (talk) 02:04, 27 January 2024 (UTC)

@Kwékwlos, for that matter, Japanese 川 doesn't manifest as kawara in any source I've seen, and the meanings given by Japanese sources for kawa and kawara are distinct: "river" for kawa, and "river plain" for kawara. ‑‑ Eiríkr Útlendi │^{Tala við mig} 01:55, 27 January 2024 (UTC)