Wiktionary:Etymology scriptorium/2023/November

Schokolade
How is the ending -ade usually explained if the word was borrowed from French? ←₰-→ Lingo Bingo Dingo (talk)  12:04, 2 November 2023 (UTC)


 * Kluge says it's actually from, not from . —Mahāgaja · talk 16:30, 2 November 2023 (UTC)
 * Le Trésor puts the French form chocolate as late as 1640, so a French intermediary cannot immediately be excluded. Surprisingly many languages have a voiced dental here: Armenian&thinsp;·&thinsp;Azerbaijani&thinsp;·&thinsp;Belarusian&thinsp;·&thinsp;Bokmål&thinsp;·&thinsp;Breton&thinsp;·&thinsp;Bulgarian&thinsp;·&thinsp;Czech&thinsp;·&thinsp;Danish&thinsp;·&thinsp;Estonian&thinsp;·&thinsp;Scottish Gaelic&thinsp;·&thinsp;Georgian&thinsp;·&thinsp;German&thinsp;·&thinsp;Hebrew&thinsp;·&thinsp;Hungarian&thinsp;·&thinsp;Icelandic&thinsp;·&thinsp;Ido&thinsp;·&thinsp;Irish&thinsp;·&thinsp;Kazakh&thinsp;·&thinsp;Latvian&thinsp;·&thinsp;Lithuanian&thinsp;·&thinsp;Macedonian&thinsp;·&thinsp;Manx&thinsp;·&thinsp;Mongolian&thinsp;·&thinsp;Polish&thinsp;·&thinsp;Russian&thinsp;·&thinsp;Serbo-Croatian&thinsp;·&thinsp;Slovak&thinsp;·&thinsp;Slovene&thinsp;·&thinsp;Swedish&thinsp;·&thinsp;Tajik&thinsp;·&thinsp;Ukrainian&thinsp;·&thinsp;Uzbek&thinsp;·&thinsp;Welsh&thinsp;. What is going on here? Did they all directly or indirectly borrow the term from Dutch? --Lambiam 17:16, 2 November 2023 (UTC)
 * I suspect a lot of the Slavic and post-Soviet languages got it directly or indirectly from German. I'm a little surprised that all the Celtic languages have -d since they're all primarily in contact with English or French; though in Scottish Gaelic d is actually voiceless and the word is pronounced . On the other hand, the Celtic loanwords of also end in d, so it's possible they could be borrowed from  and just change the t to d for their own internal reasons. —Mahāgaja · talk 21:28, 2 November 2023 (UTC)
 * The Scandinavian words are also likely to have been loaned from German. Finnish, unlike Estonian, mostly lacks a voiced/ voiceless distinction, otherwise, it'd likely have a -d, as well. Wakuran (talk) 18:16, 5 November 2023 (UTC)

Nouakchott
English Wiktionary says this name comes from Arabic +, but this seems spurious (whence the ك?), and conflicts with the etymology given on Wikipedia ("originally derived from Berber: Nawākšūṭ, "place of the winds," with a citation, albeit non-academic). Can someone please resolve? Soutrinda (talk) 13:47, 4 November 2023 (UTC)
 * I solved it by removing it. The first word looked quite impossible, possibly confused from multiple words, the rest mismatched, as you noticed. The  got banned for nothing less than “adding nonsense/gibberish: long history of bad edits on etymologies of entries”, I can’t think of a better ad hominem. Fay Freak (talk) 04:19, 5 November 2023 (UTC)

Perhaps

 * Moved to . --Lambiam 14:50, 6 November 2023 (UTC)

Reconstruction:Proto-Germanic/būraz
Reconstruction:Proto-Germanic/būraz considers "dwelling" and "dweller" unrelated, with the etymology of the second one unexplained, whereas lumps them together, with the result that the Yiddish word for "farmer" is transcluded into ety 1 of Reconstruction:Proto-Germanic/būraz but plainly belongs in the other section with  and cognates. Should we split into two etymology sections (or fix the transclusion issue some other way)? - -sche (discuss) 13:20, 5 November 2023 (UTC)
 * I've split the OHG entry into two etymologies, as the modern German entry already was. —Mahāgaja · talk 15:28, 5 November 2023 (UTC)

gelding
Currently the third sense of uses a quote template which doesn't actually exist: RQ:Wycliffe NT. Presumably this is meant to be either RQ:Wycliffe NT Lichfield or RQ:Wycliffe NT Hunter, but these are both Middle English templates. That being said, the actual quotation looks like modern English, so maybe it's a more recent version - I don't know. Is anyone who edits Middle English able to sort it out? Theknightwho (talk) 01:50, 6 November 2023 (UTC)
 * : The quote was added by, and the template was added  by WF, who, in typical fashion, didn't check to see if his edit actually worked. I did my best to fix it up, but apparently the Lichfield site doesn't let you link directly to the image anymore- even if you go to the correct page on the site and click the links there, they don't work. As for the quote itself: Wycliffe is definitely ME, so Eq's text is a translation of that translation. Chuck Entz (talk) 04:45, 6 November 2023 (UTC)
 * I've created and moved the quote there. If there's a modern English citation for the sense 'eunuch', obviously it can be re-added. —Mahāgaja · talk 08:36, 6 November 2023 (UTC)

Ryukyuan words for dolphins
Can anyone produce a source for this etymology connecting Ryukyuan words for "dolphin" (Various forms of FITU) to Sinitic 海豚? Henry Wonh (talk) 00:17, 7 November 2023 (UTC)
 * It's from the JLect entry: https://www.jlect.com/entry/1377/fiitu/ However, the derivation of this term from Chinese is unlikely. In Shuri we expect */ke:tun/ < **kaiton if it were truly a Chinese etymon. Comparision of the dialects give proto ryukyuan with tone class C, evident by Shuri /fì:tù/ and Yonaguni /fìQtû/; note how the long vowel in Shuri and falling accent in Yonaguni proves this tone class. Vovin had once compared this to  but the semantics are unrealistic and the -zi part cannot be accounted for. No other word in the world languages meaning 'dolphin' sounds even remotely similar to pR *peto C, so this might be an original word, whether it's fron Japonic or formed within Ryukyuan. Chuterix (talk) 01:32, 7 November 2023 (UTC)
 * Re: Chinese, I'm not sure why you're suggesting a connection with some phonetic form ? That would presuppose that the term entered Ryukyuan via Japanese.  This might instead have been a direct borrowing, without going through Honshu and Kyushu.
 * Middle Chinese reconstructed phonetics for are .  Japanese speakers interpreted the  initial as a  sound, but there is the possibility that Ryukyuan speakers would have interpreted this instead as a  sound.
 * Consider also modern Min Nan pronunciation hái thûn, Mandarin hǎi tún, Cantonese hoi2 tyun4, Hakka hói thùn. For all of these, an earlier Ryukyuan reflection as feto, as mentioned in the JLect entry, looks to be within the realms of the reasonable. ‑‑ Eiríkr Útlendi │Tala við mig 18:48, 14 November 2023 (UTC)
 * @Eirikr: Fortition of /x/ to */p/ is very unrealistic. It is more likely that pR had *p rather than *f.
 * Not unless it was borrowed from a Chinese dialect that had *pV to(N). Chuterix (talk) 22:08, 10 March 2024 (UTC)
 * @Chuterix: Note that I explicitly say it's not fortition of to .  Instead I posit the possibility that Chinese  became Ryukyuan  on borrowing.  And if it is a borrowing from Middle Chinese, I don't think we have grounds for reconstructing any Proto-Ryukyuan form.
 * I am not aware of any Chinese lect where Middle Chinese reflects with an initial.
 * Looking at the list of Chinese synonyms at 海豚, a distant possibility might be 🇨🇬 — that gets us the initial, but the second syllable is all wrong. ‑‑ Eiríkr Útlendi │Tala við mig 00:38, 12 March 2024 (UTC)

*ke(n)g-
Listed as etymology for. Cognates? What is the basis for reconstruction at the PIE level? Saph668 (talk) 14:05, 8 November 2023 (UTC)
 * Claimed cognates outside of Germanic include and  as well as ; see Pokorny's entry on pages 537–38 of IEW. Pokorny doesn't suggest any connection to, but it is rather tempting. —Mahāgaja · talk 14:45, 8 November 2023 (UTC)
 * And, whence 🇨🇬 meaning “hook” in its descendants and some other stuff is derived, is different or what? Should probably be merged and link thence and thither. Fay Freak (talk) 16:26, 8 November 2023 (UTC)
 * *kogъtь is an unlikely derivation (*-tis on a noun could be explained as a Slavic innovation, in the o-grade despite the suffix being *(é)-ti-, though that could also be a result of suffixation at the Slavic layer; deletion of medial n), the Iranian terms are fine; the Germanic term is also an unlikely derivation: deletion of medial n, irregular ablauting grade seem to point to post-PIE substrate origin, probably from the same substrate as that of *kogъtь. : *ḱank- is too phonetically different to definitely be the same root, but it does seem to be derived from the same substrate as the other terms. Consistent treatment of the stem as a verb also likely points to the term being a verb originally, something like "to hook"; this doesn't affect the Germanic or Iranian derivations, which could work regardless of original lexical class. - saph 🍏 13:16, 14 January 2024 (UTC)

*enu > 犬
Can this etymology be sourced? Particulary the *e > i transition? Also the connection between and  as well as between  and  +  seems suspect, especially considering the differing conjugation patterns. The theory that it comes from at least makes a little more sense, if we suppose a medial form *jenu, like in the etymology of, but I'm not sure if the timeline fits. Additionally, wenu appears to specifically refer to puppies, sort of like the distinction between and ; this may actually be a cognate w- element, possibly *woinu > wenu? Horse Battery (talk) 04:44, 10 November 2023 (UTC)
 * I tried to argue with Eirikr that it is borrowed from Taiwanese e, same character, meaning puppy. The most common diminutive, though I don't know much Kanji, appears to be 子 MC tsiX > JP shi, su, tsu, zu and cetera, see the compounds. In this view one has to ask why JP nu (Ety 3, usage note) is a historic variant of the ending  JP tsu, and why "子 is also an obsolete variant form of the katakana ネ (ne)", man'yōgana 祢 kun かたしろ, and why  JP ne, shin obtains very similar readings, too. I can't. 46.189.28.112 15:00, 10 November 2023 (UTC)
 * I have no idea what you're trying to say, this isn't very coherent. Why are you bringing up 子 and つ?
 * Regardless, I now understand that the article was giving a list of proposed etymologies and wasn't trying to link with  and . I may try to edit it at some point to make that more clear. Now what I'm still looking for is:
 * a source for the reconstruction of *enu given in the article, particularly considering that every derivation starts with i- as well. Occam's razor suggests a reconstruction of *i- for these words.
 * a source for the proposed  > inu etymology, which appears to be separate from the *enu reconstruction. and  appear to have distinct meanings, so a simple sound shift doesn't seem like a compelling etymology.
 * Horse Battery (talk) 19:54, 10 November 2023 (UTC)
 * @Horse Battery:
 * The anon would appear to be banned user, who posted in this thread on my Talk page about 🇨🇬. This user has a habit of posting long, rambling, disjointed idea-dumps speculating about word derivations.  Unfortunately, much of what they say is difficult to follow, and many of their suggested connections do not seem to be connectable to known etymologies and sound shifts.
 * The possible relation with OJP wenu is mentioned in the wenu entry here in the NKD entry at Kotobank. Compare the NKD entry for inu.
 * See also the Gogen Yurai Jiten entry and the Nihon Jiten entry. We should probably add these as either references or "further reading" bullet points.
 * About the *enu reconstruction for Proto-Japonic, that is from . @Chuterix, can you explain your reasoning for the initial e- here?
 * ‑‑ Eiríkr Útlendi │Tala við mig 19:12, 14 November 2023 (UTC)
 * @Eirikr: *enu is reconstructed and not *inu, because we'd otherwise expect syncope of initial *i- in Ryukyuan dialects, as seen in Proto Ryukyuan.
 * Initial *e- vs *i- does not have a well distinction in Ryukyuan dialects preceding other consonants. Although palatalization triggers in Northern Ryukyuan dialects if preceding velar or alveolar, but it's more difficult if the mora of that syllable has -i too, because no matter what that will palatalize. Chuterix (talk) 19:25, 14 November 2023 (UTC)
 * Second, wenu might be 'puppy' from small.
 * Also IP addresses keep making faulty pJ reconstructions that don't have Ryukyuan cognates or adding Austronesian or Tungusic comparisons to them. They appear to be in Vietnam. Chuterix (talk) 19:28, 14 November 2023 (UTC)
 * I'm at school so I shall check JDB later Chuterix (talk) 19:28, 14 November 2023 (UTC)
 * @Chuterix, do you have any examples of problematic IP edits to Proto-Japonic entries? ‑‑ Eiríkr Útlendi │Tala við mig 19:56, 14 November 2023 (UTC)
 * I've already RFD'd all of them. One example is where that IP put "Proto-Japonic" *yərəntu there. I was banned so I got to do it late. The JLect entry here seems to now be a ghost word, clearly borrowed from Japanese. The only word that survives is yuruzinamuN 'various snacks'.
 * Anyways yorodu might a loanword from an ancient Korean language. Anyways I gtg. Matane. Chuterix (talk) 20:03, 14 November 2023 (UTC)
 * I don't see anything from you at Requests for deletion/CJK...? ‑‑ Eiríkr Útlendi │Tala við mig 20:28, 14 November 2023 (UTC)
 * I meant . Chuterix (talk) 20:36, 14 November 2023 (UTC)
 * Are we sure that the differences in Ryukyuan reflections of 🇨🇬 and 🇨🇬 are due to the initial vowel?
 * In addition to the differences in the following vowel, I see that modern Tokyo Japanese has different pitch patterns, with "rice" as pitch-pattern 1, and "dog" as pitch-pattern 2.
 * ‑‑ Eiríkr Útlendi │Tala við mig 20:06, 14 November 2023 (UTC)
 * What matters is Proto Ryukyuan tone classes. While belongs to 2.4 accent class,  belongs to 2.3 accent class. Both 2.3 and 2.4 which is considered to be the low register regularly corresponds to Proto Ryukyuan tone class B, which is represented by Shuri Okinawan tone 0 (means no accent) (called B in linguistic studies), both ?Nni 'rice plant' and ?iN 'dog' belong to accent class 0 (B). See also kikigengo.ninjal.ac.jp and see * enu and *ine. The syllable count is omitted while JapaneseAccentClass+RyukyuToneClass (e.g. 4B and 3B) is used. It's reconstructed and done by Yosuke Igarashi, from a spreadsheet. He does not give real cognate sets due to copyeight issues, according to personal communication with him. However it's clear that it's the reason for giving accent (for words with no Ryukyuan cogs, they're at least attested in Shuri, Yamatohama, Yoron, Nakijin, Asama, Hirarq, or Ishigaki AFAIK. This is how I figured how he got, although the cognates are sparse).
 * Thus what matters is comparing Ryukyuan. It's why we have pJ . I wonder if you were stumped when you first saw that *e was reconstructed.
 * In C tone words it's easily identified if Shuri has long vowels such as mù:tù 'origin' or leaves C class (falling accent) in Yonaguni. Chuterix (talk) 20:35, 14 November 2023 (UTC)
 * Also @Eirikr, this madness is why I deleted and rewrote the etymology from ; too speculative though it's not impossible the last syllable is cognate with . Chuterix (talk) 20:46, 14 November 2023 (UTC)

Reconstruction:Proto-Indo-European/peg-
Why would this be reconstructed as *peg- rather than *pek-? All the descendants have a voiceless obstruent after the consonant in question, which neutralizes the distinction between k and g anyway. And wouldn't make us expect  with a long ē if PIE had *g? Why can't this be ? (as an editor of the page though not the creator, who has no other contributions to Wiktionary besides creating this page). —Mahāgaja · talk 18:53, 10 November 2023 (UTC)


 * The consonant could also be an IE palato-velar as well. Though Latvian paksis may suggest *k. Kwékwlos (talk) 03:33, 11 November 2023 (UTC)
 * Lachmann's law is debated: a fair number of works on it suppose that it was restricted to participles, not a generally applicable sound change. However, even if *peg- is not definitely excluded by the Latin form, I see no obstacle to reconstructing *pek-. Furthermore, De Vaan views the attempted connection to pakṣá- with skepticism.--Urszag (talk) 07:02, 14 November 2023 (UTC)
 * I can't find a mention of *peg- anywhere. I'm not sure if anything should be reconstructed. There is very little to go by. —Caoimhin ceallach (talk)

tapir and jaguar
The etymology of these two in Portuguese has been bugging me for some time now. Ultimately, both are without a doubt from Old Tupi and, respectively. What gets me is the lack of the last vowel. This dropping is strange in Portuguese and it hasn't happened to any other word borrowed from Tupi. It gets even more suspicious when the doublets and  exist in Portuguese.

My initial guess was it came indireclty into Protuguese as a borrowing from French, as this language has this kind of vowel dropping for Tupi borrowings and the fact that the words and  were far more common in Brazil, even to this day. English etymology saying their was borrowed from French would support that (but they say  comes from Portuguese).

The problem is that all online dictionaries I've searched (Infopedia, Michaelis, Aulete etc...) say it's a direct borrowing from Tupi. I've looked into Machado's and Nascente's etymology dictionaries and Navarro's Tupi dictionary and they all say the same. So, in the end, I have this theory but can't find any reference to back it and allow me to edit these Etymology sections. Trooper57 (talk) 22:21, 10 November 2023 (UTC)


 * To the puzzling terms you can add the noun (front of a hut) and , which cannot have made a French detour. Also, we write that  is ; shouldn't the etymon be ?  --Lambiam 15:33, 12 November 2023 (UTC)
 * I guess they were borrowed from dialectal or later forms (ie. from after its usage as a written language was banned by the Portuguese authorities) of Tupi which lost the final -a. Rodrigo5260 (talk) 17:34, 27 November 2023 (UTC)

References to Proto-Indo-European *lewh₁-
Is this edit by the IP 69.120.69.23 correct? I can't tell, as it's pointing to for which we don't currently have an entry. The edit summary says "wrong root; see Latin luō § Etymology 2", but that entry refers to which also doesn't exist yet (the entry currently only has the sense "louse"). I think the same IP has made multiple changes of this nature. — Sgconlaw (talk) 13:33, 11 November 2023 (UTC)
 * It's correct root is (which laryngeal is unknown): . —Caoimhin ceallach (talk) 20:55, 16 November 2023 (UTC)

Russian
Russian for, which is a Maoist term that saw use during the of the 1960s & 70s to denounce those deemed to be following the capitalist way. Most languages have simply calqued the original Chinese, but this seems to be different. Theknightwho (talk) 19:36, 11 November 2023 (UTC)
 * Could it be a communist pun on капут (kaput), or would that be overthinking things? Wakuran (talk) 23:34, 11 November 2023 (UTC)
 * Unlikely; a punster would have coined капутист with a single ⟨п⟩. --Lambiam 15:00, 12 November 2023 (UTC)
 * I think it is  --Lambiam 15:08, 12 November 2023 (UTC)

Pildești
The ety ("Cheldea + -ești") seems wrong, maybe a copy-paste error... - -sche (discuss) 04:02, 13 November 2023 (UTC)
 * Romanian can shift /p/ > /k/ before front vowels, see a similar etymology in Pochidia.  Perhaps in this case, Cheldea is a shifted form of a proper name Peldea that later was restored to its original form.  Here are the diffs from that day .... it doesnt look like a copy-paste error to me, but it looks like the source relies on a lot of information that may be inaccessible on the Internet.   That would explain why neither Cheldea nor Peldea turn up much on a superficial Google search. — Soap — 19:31, 13 November 2023 (UTC)
 * The name$?$ Cheldescu appears in a single book scan here. The book was published in Vaslui, Romania, which is only about 90km from Pildesti.  Possibly just a coincidence, but ... hmm... it might just be that there was a very rare surname in this one area, and a village was named after a member of that family.  — Soap — 22:26, 13 November 2023 (UTC)


 * An explanation for this development is mentioned in this journal article (p. 97). Einstein2 (talk) 22:31, 16 November 2023 (UTC)


 * /p/ to /k/ is a common phonetical evolution in some Moldavian dialects of Romanian. Some locals see it as "wrong", so they attempt to correct it, even when that evolution didn't occur. For instance, piftea from chiftea is one such a hypercorrection.
 * Also, I added a bit of details and the reference to the article. Bogdan (talk) 10:23, 28 November 2023 (UTC)

Proto-Balto-Slavic and Indo-European dual endings
I just created to try to provide greater clarity about the origin of forms like Latin  (it was previously stated to derive from, which doesn't seem to be its direct ancestor). Along the way, I've run into a few difficulties that I hope someone could help me with. For clarity and concision, it seems best to refer to all solidly reconstructable forms at as high a level as possible, so I'd like to link to the Proto-Balto-Slavic form that is the common ancestor of Latvian, Lithuanian , and Proto-Slavic. The last of these says "From Proto-Balto-Slavic *abu" but I can't figure out why -oh₁ would become -u or why it doesn't match the outcome shown at. The citation to Derksen 2008 in the second entry seems to cite a third notation, "“*duoʔ; *duoiʔ". What exactly is the reconstructed thematic nom/acc/plr dual ending at the Proto-Balto-Slavic stage? @ZomBear, @RadomirZinovyev, would you be able to help with the creation of the Proto-Balto-Slavic entry?

Another anomaly that I encountered was the reference at to an ancestor ending in -ow. Is this worth keeping? I found some references to -ou as a PIE masculine direct dual ending (e.g. Shields 2004:26, The emergence of the dual category in Indo-European: "The thematic nominative-accusative dual suffix *-ōu (e.g., Skt. -āu)", and Adams' entry on antapi refers to "the old masculine *-ōu") but I can't figure out the status of that reconstruction is vs. -óh₁ (e.g. is *-ōu a secondary development from -óh₁, found only in some subset of IE languages?). Finally, I found that in contrast to the reconstruction of the feminine dual ending as -éh₂h₁e (found in Ringe 2006:49), there seems to be an alternative reconstruction ending in -éh₂-ih₁ ("Based on the concurrent Indo-Iranian and Slavic data, the direct case ending is more or less safely reconstructible as *-ah₂-ih₁", "The Proto-Slavic Genitive-Locative Dual: A Reappraisal of (South-)West Slavic and Indo-European Evidence," Yaroslav Gorbachov 2017, page 79). Are any of those alternative reconstructions worth including as variant forms in the table? Urszag (talk) 06:36, 14 November 2023 (UTC)


 * In Proto-Balto-Slavic a sound like óh₁ would render into "ō" while at the East-Baltic stage "ō"s become uo. Which in both Latvian and Lithuanian (inconsistantly) render into a simple "u" (primarily in conjugated forms). Hence, dwóh₁ > duwō / diwō (Both are possible dialectal renderings). In the case of Latvian the final "i" is hypothesized to originate as an analogy to other typical dual conjugations, thus "divi". As for h₂n̥tbʰóh₁, the only thing which seems strange to me is that the initial h₂ should not typically render as anything in Proto-Balto-Slavic or daughter languages. Considering that, from what I know (don't take me as an authority) h₂n̥tbʰóh₁ should render as inbō. While the Proto-Slavic "oba" points to a pre-Slavic "abā". That being said, if "h₂" did render as an "a" in this circumstance and cancelled out the typical "n̥ > in" (in favour of the intial "a" plus a "n") then it could be possible to attain the form "anbō". This would be a viable explanation for the Lithuanian form, and would suggest a Proto-Slavic form of "ǫba".
 * About the dual forms and conjugations I won't comment as I haven't studied this subject deeply enough to consider giving an meaningful opinion. RadomirZinovyev (talk) 14:00, 14 November 2023 (UTC)
 * Pinging @Thadh who might have context on this. Chernorizets (talk) 22:34, 15 November 2023 (UTC)
 * George Dunkel (LIPP vol. 2, p. 124 under *bʰó- 'both' ) has a quite different view. He doesn't consider, , and to be directly connected to  and . Rather, he derives them from , which is quite a bit more plausible than "from the front > both". He argues against the above etymology (pp. 126-7), which is originally from Jay Jasanoff (1976). The original PIE nom./acc./voc. dual ending is usually reconstructed  (Fortson 2010 p. 126, older 2004 p. 105 ), but in the thematic declension an extended  occurs. It's not clear why. (See LIPP p. 819 under *u 'there'.) ―Caoimhin ceallach (talk) 22:50, 16 November 2023 (UTC)

抓 (zhuā, "to grasp") as semantic compound of 找 ("to seek") and 爪 ("to claw")?
Is there evidence that 抓 (zhuā, "to grasp") is a semantic compound of 找 ("to seek") and 爪 ("to claw")? Or a combination of 我/手 and 爪 ("to claw") representing an object?

Could the Etymology Scriptorium be divided up on a per-language basis, or have a "URL-based search" function? So that a URL like "en.wiktionary.org/wiki/Wiktionary:Etymology_scriptorium/zh" would link to a search for posts on Hanzi? Alphabet-based languages would be harder because of the probability of intra-language discussions for particular words happening. Jimw338 (talk) 17:18, 14 November 2023 (UTC)
 * We have split Wiktionary discussion pages, usually starting by separating English from everything else. See WT:RFV for an example. You might take your request for language-specific search to WT:GP. DCDuring (talk) 16:33, 18 November 2023 (UTC)
 * Isn't this just/ mainly a combination of left radical 手 (something to do with hands) and right radical 爪 (something that sounds roughly like claw)? The sound correspondences don't always work that well in modern Mandarin, though. Wakuran (talk) 19:44, 14 November 2023 (UTC)

여러분
@Eirikr The last part could be. Chuterix (talk) 20:42, 14 November 2023 (UTC)
 * Yes, and 여러 means "several, various, many" :] Leasnam (talk) 14:12, 15 November 2023 (UTC)
 * Okay, I see the note on the entry page now... Leasnam (talk) 14:16, 15 November 2023 (UTC)

apixaban (Eliquis)
Currently the etymology for Apixaban stops at “-xaban” for “Xa clotting factor inhibitor”.

I think the “api-“ is a shortening of (pyr)a(zolo)py(ridine), the type of molecule that apixaban is. Barrmor (talk) 03:35, 17 November 2023 (UTC)

Wuischke
RFV of the etymology; tagged but not listed by an anon who wrote: for " in 1419". Because: Moved here from the main RFV page as it's pretty clear the user's issue is with the etymology, not with the term's existence. —Mahāgaja · talk 07:26, 17 November 2023 (UTC)
 * mentions it, but without a language indication (could also be mhg Middle High German or la Latin).
 * mentions it as "Ugest parva", is Latin and makes sense.
 * This is pure speculation, but could it be related to Old Norse vǫxtr, Old High German wahst (growth), or Proto-Germanic *waiþō with a -st suffix, or would that be phonetically unlikely? Also -st is mostly used for constructing verbal nouns, so a hypothetized *waiþistiz or similar might not make much sense. Wakuran (talk) 11:44, 17 November 2023 (UTC)
 * Considering both localities named Wuischke are in the traditionally Sorbian-speaking part of Germany, it's very likely that the word is of Slavic origin, not Germanic. The Upper Sorbian name for both places is, which looks like a diminutive in of something or other. I think "parcva" is a typo on the part of the hov.isgv.de website and it should be "parva" as the Google Books scan shows. I also notice that Wuischke (2) gives the oldest name as "Vgestchen" with the German diminutive suffix . So the fact that one of them is called "parva" (Latin for "small") and the other uses the German diminutive suffix makes me think it's all the more likely that the -k of  is the diminutive. That leaves the wujež- part. I don't know what it is, but I do think it's most likely to be Slavic. —Mahāgaja · talk 12:49, 17 November 2023 (UTC)

humour (English)
The etymology given of this English word is "...from Latin humor, correctly umor (“moisture”), from humō, correctly umō (“to be moist”). However, the etymology of the Latin word is from ūmeō (“to be moist”) + -or.   on the other hand means "to bury", while  means "head/title". Is this a mistake, or is this some feature that I'm not aware of? Duchuyfootball (talk) 04:20, 18 November 2023 (UTC)
 * Seems as if the h- might have stemmed from a Late Latin hypercorrection or folk etymology. Wakuran (talk) 10:51, 18 November 2023 (UTC)
 * The English etymology is incorrect. The correct etymon is . My bigger objection with it however is that it ignores the, on the face of it, surprising shift "moisture" > "wit". ―Caoimhin ceallach (talk) 11:10, 18 November 2023 (UTC)
 * The sense development is revealed by reading sense 5, sense 3, sense 2 and sense 1 in that order. —Mahāgaja · talk 12:11, 18 November 2023 (UTC)
 * That's true. Still think it deserves mentioning. —Caoimhin ceallach (talk) 15:22, 18 November 2023 (UTC)
 * So is someone going to fix it, or we just leave it at that? I'm not confident about my English expertise so I will pass.--Duchuyfootball (talk) 12:43, 18 November 2023 (UTC)
 * I've updated it. —Caoimhin ceallach (talk) 15:22, 18 November 2023 (UTC)

방아
The Chinese Edition of Wiktionary says, that the term 방아 was first attested in the following work: 《釋譜詳節 (석보상절)》 (1447)

Is that true? If so, what was the exact Middle Korean term? -- Apisite (talk) 07:42, 18 November 2023 (UTC)

intimus (Latin)
The etymology of is From the radical of inter + -imus., but there is no information about suffix "-imus". Can somebody look into this? Duchuyfootball (talk) 12:53, 18 November 2023 (UTC)


 * Like other superlatives on -imus, such as, this is likely inherited from Proto-Italic or even PIE, although some may (e.g. and ) may have been formed by analogy. --Lambiam 13:52, 18 November 2023 (UTC)
 * Thanks for this information. Can you by the way add this in the page? It's only showing the English suffix. Duchuyfootball (talk) 14:10, 18 November 2023 (UTC)
 * I don't know of scholarly sources that support my hypothesis that some Latin superlatives on -imus may have been formed by analogy to others for which the ending was inherited. For minimus, de Vaan writes that it is derived from *minos by means of a productive process, but does not make clear in which language(s) this process took place. --Lambiam 10:34, 19 November 2023 (UTC)

ubique (Latin)
Compound of ubī̆ (“where”) + -que (“each, ever”). Correct me if I'm wrong, but the definitions of -que do not include the sense of "each, ever", but rather "and", which would make ubique "and where". Duchuyfootball (talk) 15:02, 18 November 2023 (UTC)
 * It's a special sense, like the use of "-ever" in English to form words like "whoever, wherever". I added it as a separate POS since -que isn't a conjunction in words like this.--Urszag (talk) 15:53, 18 November 2023 (UTC)
 * That's correct. I'd add that this sense should get it's own etymology, as it is unlikely to have arisen in Latin. Indefinite/generalising/distributive *-kʷe occurs in Indo-Iranian, Greek, Germanic, and Armenian. especially has cognates in Sanskrit, Avestan, Gothic, and Armenian (LIPP vol. 2 pp. 442-6 "3. *-kʷe" ). —Caoimhin ceallach (talk) 16:19, 18 November 2023 (UTC)

bexx
RFV of the etymology. The Arabic word matches its meaning more; perhapse the Maltese word  is from Arabic  instead? --kc_kennylau (talk) 16:49, 18 November 2023 (UTC)
 * Aquilina compares it to, but imo it is likely of onomatopoeic origin. The descendant of in Maltese is  because the  is emphatic in the Arabic word. Besides,  <  is impossible. — Fenakhay ( حيطي · مساهماتي ) 16:59, 18 November 2023 (UTC)
 * Perhaps a conflation of the two roots then? Because I can't find a source for having the meaning "sprinkle". --kc_kennylau (talk) 17:10, 18 November 2023 (UTC)
 * means to “sprinkle” or similar literally, I have used to imagine, also kind of.
 * It does not need to be inherited at all though, as Fenakhay has remarked. In general dialects do not even descend from Classical Arabic, this is the kind of words that don’t, too colloquial and due to their imitative nature “not real words” enough for being thoroughly recorded in literature. Like I have recorded all kinds of frass for, including only real examples from the internet broken down, but of course only about half is written about or literary. It is more difficult for me to read through Arabic texts to find such stuff if it exists—some sequences always exist if the corpus is large enough. Fay Freak (talk) 17:34, 18 November 2023 (UTC)

excappāre (Vulgar Latin)
explains this word as "get out of one's cape, leave a pursuer with just one's cape". I find this kind of obscure with little connection to the present meaning, especially the "leave a pursuer with just one's cape" part. Can anyone give a better explanation for this? Duchuyfootball (talk) 02:39, 19 November 2023 (UTC)
 * Etymonline gives the same etymology. That a skilled escape artist could sorta drop his cape quickly and get out of sight doesn't strike me as too unlikely. Both semantic shifts and figurative expressions are very common in language evolution. Wakuran (talk) 02:50, 19 November 2023 (UTC)
 * Could it be from ? --kc_kennylau (talk) 13:21, 19 November 2023 (UTC)
 * I fail to see any believable semantic evolution. The occasional connections seem to have occurred long after te daughter terms of *excappāre evolved. Wakuran (talk) 21:53, 19 November 2023 (UTC)
 * Thanks for this explanation. It's just that English is not my native language so I find it hard to understand what "leave a pursuer with just one's cape" means. Duchuyfootball (talk) 15:44, 19 November 2023 (UTC)
 * In medieval Italy, a "cape" wasn't the long, flowing garment we associate with royalty and superheroes; a more appropriate modern term would be a shawl, I believe, if not cloak. It would cover the shoulders, meaning that if a thief was apprehended from behind, grabbed by the shoulders or upper arm, they could simply take the action (-re) of squirming their way out (ex-) of the cape (cappa) -- and out of trouble.
 * This is the etymology I've always been taught, though I do note that cape -- deriving from PIE, meaning "head", also the ancestor of cap -- has no etymological relationship to capture, captor, captive, or their Latin ancestor, , coming from PIE . I don't know enough to say whether the popular etymology could be wrong, but the coincidence in "escape" meaning "to evade capture" quite possibly helped the term become the standard choice. Qwertygiy (talk) 17:56, 19 November 2023 (UTC)

거북(geobug) Korean
I think the etymology of the word “거북“ could be the Turkish word “Kabuk” because the pronunciaton and the meaning of the two word is so similar. But I don’t know any Korean. So, can someone knows the language enlighten me? Józio Kovács (talk) 11:50, 19 November 2023 (UTC)
 * Seems very unlikely to me, particularly with the Altaic hypothesis largely being discredited. Probably just a coincidence. Wakuran (talk) 13:00, 19 November 2023 (UTC)
 * Yes it is discredited but denying that these languages exchanged vocabulary because of the geographical circumstances like the Balkan Sprachund did would be inappropriate. One can still find Mongolic, Tungusic, Koreanic and Turkic vocabulary smashed into each other in those languages. Maybe it’s not too much but still you can see the trace and impact of it Józio Kovács (talk) 16:13, 24 November 2023 (UTC)

bitch eating crackers
The earliest instance I've seen is a 2013 tweet from a parody account named Will_Ferrell (which is _not_ w:Will Ferrell), here(nitter link):. It's earlier than any other mention I can find, and it seems to have really taken off around then. Is this enough to cite the coinage as being this particular tweet? grendel|khan 18:15, 19 November 2023 (UTC)


 * Scratch that; I found this January 2012 comment on Reddit saying the line. And this July 2011 post on Blogspot. Or this Livejournal comment from November 2011. Nothing I can see from 2010, at least. Maybe the true origin really is lost in the mists of time. Darn. grendel|khan 18:22, 19 November 2023 (UTC)
 * BTW, it is not likely that this is a true adjective. In English grammar virtually any noun phrase can be used attributively, to modify a noun. It would be silly for us to duplicate every sense of the noun, perhaps slightly reworded, in an adjective section. To test whether there is some definition that would qualify as a true adjective, see the tests at WT:English adjectives. DCDuring (talk) 22:19, 19 November 2023 (UTC)
 * The LiveJournal image can be traced back to this Tumblr post, and that exact phrasing ("bitch, eating those fucking crackers") can also be found on that exact same day here. Before that, on July 9, 2011, a joke compilation blog includes it. (edit: corrected date) Confermusearename (talk) 03:39, 11 December 2023 (UTC)
 * The first Twitter attestations occur on November 1st, which coincides with the Tumblr post and AmIRite post. It seems that's when it went viral. confermusearename (talk) 04:00, 11 December 2023 (UTC)

Proto-Tungusic numerals
This edit by caused some pages to have module errors as it changed the reconstructed forms of the Proto-Tungusic numerals. I also noticed that some numerals (either before or after the edit) do not line up with the reconstructed words in Appendix:Proto-Altaic reconstructions. Where did Ardahan Karabağ come across these new reconstructed forms, and can we come to one standard? --kc_kennylau (talk) 21:17, 19 November 2023 (UTC)


 * I will fix the all numerals as soon as possible. "New" reconstructed forms are reconstructed by Johannes Benzing (1955). The Proto-Tungusic reconstructions in the "Proto-Altaic" section are baseless and an EDAL comparanda. Benzing's methods are way better for Proto-Tungusic. Ardahan Karabağ (talk) 04:58, 20 November 2023 (UTC)

畫虎𡳞
RFV of the etymology. — justin(r)leung { (t...) 01:45, 20 November 2023 (UTC)
 * . — justin(r)leung { (t...) 06:23, 26 November 2023 (UTC)
 * If you search on Google about 唬爛 you will see articles about the folk etymology Mahogany115 (talk) 08:54, 30 November 2023 (UTC)

Looking for a PIE form similar to *smeylos
I'm working on a Bulgarian entry, and the etymological dictionary eventually derives it from 🇨🇬. I'm unable to find the PIE root with that meaning - is about laughing, and I don't immediately see anything else we have that's close. Google was no help either. Any suggestions for PIE roots or adjectives meaning "sharp" that look something like the form I've posted?

Thanks,

Chernorizets (talk) 08:30, 20 November 2023 (UTC)


 * Is it possible there's a connection with ? Either a loan from Greek or a shared borrowing from a non-IE substrate.  The PIE root we give there, *smī- , is nonstandard because of the /ī/, so a different form of it might be available. — Soap — 08:42, 20 November 2023 (UTC)
 * There are lots of PIE roots that have been proposed over the years that we don't have entries for. Pokorny's dictionary proposes a root *smēi- *sməi- *smī̆- (in our notation, ) meaning 'cut, carve'. He doesn't mention any Slavic descendants, but does mention and  as possible descendants. The "*smī-" mentioned in our etymology of  would be the zero grade *smiH-. —Mahāgaja · talk 08:57, 20 November 2023 (UTC)
 * @Mahagaja thanks! Indeed, the Baltic cognates given by the dictionary are 🇨🇬, and ; 🇨🇬, . Dialectal Proto-Slavic  gives rise to the name of a pointy-leafed plant in a few Slavic languages, see . Chernorizets (talk) 09:10, 20 November 2023 (UTC)
 * @Mahagaja FWIW, the work by Trubachev which links the Slavic and Baltic terms via a PIE root can be found here: https://inslav.ru/publication/serbo-luzhickiy-lingvisticheskiy-sbornik-m-1963 (page 166 of the PDF). He writes it as, , implying that the "L" is part of the (extended?) root. Chernorizets (talk) 10:16, 20 November 2023 (UTC)

Reconstruction:Proto-West Germanic/hindiba%CA%80i
In addition to the already listed descendants, SAOB mentions the (from context apparently exclusively Southern Swedish) forms ,,, ,. Where should these be placed in the descendant tree at Reconstruction:Proto-West Germanic/hindiba%CA%80i? I know that Southern Swedish has a strong historical Danish influence but in this case it's not clear whether we should treat the terms as descending via a Danish intermediary or directly from (presumably, if that's where Danish also got the term) the Middle Low German form. Helrasincke (talk) 09:46, 20 November 2023 (UTC)


 * If there's no evidence for these words in Danish, it's probably best to say they went straight from MLG into Swedish. —Mahāgaja · talk 10:52, 20 November 2023 (UTC)


 * Traditional Scanian seems to have been largely a Danish dialect, so to me, it would make sense to derive these words from Danish. Does Scanian have a separate language code, though, or should we mark the words as dialectal within a paranthesis? Wakuran (talk) 11:19, 20 November 2023 (UTC)
 * Scanian has the code . —Mahāgaja · talk 11:44, 20 November 2023 (UTC)
 * @Mahagaja To clarify, the term does also exist in Danish. I suppose it's really a question of whether we consider the Scanian term to have been borrowed via Danish or straight from MLG. Depending of whether we are assuming an areal spread or a borrowing in the context of extensive contact during the Hanseatic era (which seems to be how most MLG loans into Scandinavian are explained), it seems possible that aside from a diffusion chain North Germany > Danish Islands > Scandinavian peninsula (plausible), both adopted the term independently around the same time of intensive contact with MLG (possible), or even that the term may even have entered Danish from MLG via Scanian (probably least likely but not impossible). There is also the possibility that Danish adopted the term from which Scanian then borrowed it (surely unlikely since judging from this map Scania seems to have had at least as much opportunity for contacts with MLG as the [standard] Danish language area had). To my knowledge it would be very difficult to prove definitively either way, though we can probably fall back somewhat on the probabilities. Helrasincke (talk) 23:06, 25 November 2023 (UTC)
 * A further question is whether the term perhaps already existed in North Germanic. Icelandic at least seems to have inherited both constituent elements and . Helrasincke (talk) 23:18, 25 November 2023 (UTC)
 * The word is first mentioned in Icelandic in the 18. century. I can't find out since when raspberries have grown there. Íslensk orðsifjabók  says it's borrowed from Danish. —Caoimhin ceallach (talk) 00:35, 26 November 2023 (UTC)
 * Yes, I suspected they may not grow natively in Iceland (and Plants of the World Online marks it as introduced), but that doesn't exclude the possibility that (continental) Old Norse didn't also have an unrecorded word for it. I'm hoping someone with more expertise can chime in. Helrasincke (talk) 13:49, 26 November 2023 (UTC)

(English)
I notice there is no etymology given for this word. Could it be from detect + ? The latter is defined as a suffix for adjectives, though. Duchuyfootball (talk) 14:24, 20 November 2023 (UTC)
 * Well, yeah. The noun is derived from the adjective, though. Cf. Etymonline. The international sense "sleuth" seems to be a borrowing from English, generally. Wakuran (talk) 14:30, 20 November 2023 (UTC)

Germanic ge- prefix
The given etymology for the ge- prefix in Dutch, German, etc., says that it comes from the PIE *ḱóm, probably first proposed in Lockwood (1968). However, this has never made sense to me as it blatantly violates Grimm's Law (the proposed origin should lead to Proto-Germanic **ha). I was recently working on a project, and came across the fact that PIE stative verbs were made using reduplication of the initial consonant with an epenthetic *e added between. Since many PIE words started with either *gʰ or *ǵʰ, I wonder if one of those reduplicated prefixes could have instead led to the ge- prefix, in a way that fits with Grimm's Law and is still semantically sound. Smeyers31 (talk) 20:15, 20 November 2023 (UTC)


 * Derivation from follows Grimm's law followed by voicing (similar to Verner's law) because the ge- syllable is unstressed. Voicing of initial consonants in unstressed syllables is very common; it happened in the very same word (but several centuries later) in  from . It's also why the th in English unstressed words like the, this, that is  rather than  (as it is in stressed content words like thick and thin). —Mahāgaja · talk 21:38, 20 November 2023 (UTC)
 * It's a creative solution, but not very likely. If that were the case you'd expect schwimmen, **geschwamm, **schwommen, i.e. you'd still have to explain why the preterite doesn't have ge-, but the past participle does. —Caoimhin ceallach (talk) 22:35, 20 November 2023 (UTC)
 * Also, Germanic does have a reflex of the reduplicating verbs, namely the CAT:Proto-Germanic class 7 strong verbs. —Mahāgaja · talk 07:37, 21 November 2023 (UTC)
 * I noticed a slight mistake in your explanation (or perhaps we're saying two different things). Either way .... it's not because ge- was unstressed, but because the syllable before it (in a preceding word) often was.  And it seems that Verner's law usually didnt apply in initial position.  So I would say that this qualifies as an irregular change needing an explanation.  However I still agree with you that it most likely could be explained by the existence of the preceding words, rather than being generalized verbal reduplicating prefixes, which as you say are directly attested in verbs like leolc and did not generalize to just one single consonant.   — Soap — 20:24, 27 November 2023 (UTC)

Connection between rake/h₃reǵ- and rachis/rhákhis?
On the page for rachis, we see: From New Latin rachis, from Ancient Greek ῥάχις (rhákhis, “spine, ridge”).
 * Following the link for ῥάχις: According to Beekes of Pre-Greek origin or related to Lithuanian ražis (“stubble”). Compare ῥαχός (rhakhós, “thorn hedge”).
 * Following the link for ῥαχός: Clearly related to ῥᾰ́χις (rhákhis, “spine, backbone”). Since ϝρᾰχ-/ϝρᾱχ- cannot be derived from an Indo-European root, the word might instead be of Pre-Greek origin.

Meanwhile, on the page for rake: From Middle English rake [and other forms], from Old English raca, racu, ræce (“tool with a row of pointed teeth, rake”), from Proto-Germanic *rakō, *rekô (“tool with a row of pointed teeth, rake”), from Proto-Indo-European *h₃reǵ- (“to straighten, right oneself”).

These seem related, but I don't see any links or explicit connections between these pages. Could there be a connection here (perhaps one that I'm missing), or am I just looking at a coincidence? (Total etymology noob. Fell down this rabbit hole while exploring the etymology of Polyrhachis, the spiny ants.) --Xarm Endris (talk) 01:13, 22 November 2023 (UTC)
 * I think the expected Greek reflex of a word beginning in a laryngeal plus /r/ would be vowel-initial. There are some people who don't believe PIE even allowed words beginning with /r/ at all, .... we list some  but they all begin with reh₁-, so it doesnt seem to have been a common word shape.  The argument for h3 seems to rest on the Greek reflex ὀρέγω, which we pair with Latin reg- "rule".  As for the vowel mismatch, I dont know.  I would lean towards saying it probably is a coincidence, but wouldnt rule it out completely.  — Soap — 05:58, 22 November 2023 (UTC)
 * In addition, χ can only come from PIE *gʰ, ǵʰ, or *gʷʰ. —Caoimhin ceallach (talk) 22:43, 23 November 2023 (UTC)

PIE root needed: *bʰeHw-
This root is implicated in the etymology of a number of words: 🇨🇬, 🇨🇬, 🇨🇬 and its descendants, 🇨🇬 and its ancestor chain, likely 🇨🇬, and a few others, and more in other Slavic languages.

Could someone please confirm that this is the correct shape of the root? There are not a ton of hits on Google, and a lot of them may be references to Wiktionary. If confirmed, it seems to me at least like it would be worthy of inclusion in Wiktionary. Chernorizets (talk) 03:56, 22 November 2023 (UTC)


 * We even have Category:English_terms_derived_from_the_Proto-Indo-European_root_*bʰeHw-. Chernorizets (talk) 04:00, 22 November 2023 (UTC)
 * None of those etymologies seem to point to a particular root with any kind of certainty, apart from the Bulgarian (but I can't vouch for the quality of the Bulgarian etymologies), which points to (*bʰweH- in LIV). Altogether, no grounds for *bʰeHw-. —Caoimhin ceallach (talk) 19:59, 22 November 2023 (UTC)

super
Has no one explained the anomalous initial s- in Reconstruction:Proto-Italic/super? It escaped my notice all this time because it has an obvious Greek cognate, but someone reminded me just recently that all Greek words beginning with an upsilon take aspiration, and moreover that this word is also cognate to English over and probably up, which imply that the PIE original began with a vowel (or a laryngeal). Where did the s- come from? I know about s-mobile, but if that explains this, it would be the only known example of s-mobile before a vowel or laryngeal. Surely something this obvious has been talked about before, but I wouldn't know where to look. Has this been explained? if not, can we at least add something pointing out the anomaly? Thanks, — Soap — 16:26, 22 November 2023 (UTC)
 * Isn't the *u just a syllabic *w? Chuck Entz (talk) 20:27, 22 November 2023 (UTC)
 * I'm not sure if I understand you, but an initial *wp-cluster in *wpér feels like quite a mouthful. Wakuran (talk) 15:01, 23 November 2023 (UTC)
 * I think his point is that in PIE, u and w are allophones of a single phoneme, so that what surfaces as *upér is underlyingly . However, I notice that we don't have any PIE entries beginning *(s)w- or *(s)y- even before a vowel, which might mean that s mobile doesn't occur before *w and *y even when they're nonsyllabic. (Or maybe we just don't have entries for that yet.) —Mahāgaja · talk 15:13, 23 November 2023 (UTC)
 * R:ine:Sihler:1995 (§ 406.7, p. 441) says the s- of, , and is of obscure origin. I don't know what etymology of  he's thinking of, though; ours doesn't seem to have an issue with the s-. I suppose he's connecting it with  and/or . —Mahāgaja · talk 07:38, 23 November 2023 (UTC)

mycorrhiza
RFV of the etymology.

An IP asked the simple question: "where does the extra 'r' come from?". It looks like what you would get if the first part ended in a consonant which assimilated to the "r". The problem is that none of the candidates in English, Latin or Ancient Greek have an "o" sound followed by a consonant. Most parallel constructions in English only have a single "r". I can only guess that the example of glycyrrhiza (from, compound of  + ). is involved somehow. Chuck Entz (talk) 00:43, 23 November 2023 (UTC)
 * So far as I know, /r/ is always doubled after a vowel in Greek compounds. hence -rrhea, -rrhaphy, etc. Any compounds without the doubled r are likely modern creations by those who have forgotten or decided to move past that rule.  Indeed I just remembered Polyrhachis from a few sections above.  We list other words with poly- and a single r, but they seem to be all modern coinages. — Soap — 05:34, 23 November 2023 (UTC)
 * I think eurhythmic is also a modern coinage. I've never been sure whether  is just a misspelling or a blend of Europe + rhythmic. —Mahāgaja · talk 07:13, 23 November 2023 (UTC)
 * Actually, I just learned that the Ancient Greek word is without doubling of the r and without the rough breathing (but compare Epic ). I wonder whether Dave and Annie were familiar enough with Ancient Greek that they got the spelling "right" (etymologically speaking) as opposed to the usual English spelling. —Mahāgaja · talk 15:35, 23 November 2023 (UTC)

Why is the middle bone of a finger called 'middle' and not 'medial'?
I was learning about the bones in the body, and something stuck out to me as sort of strange. The book discussed that the fingers have three bones. The one farthest away from the hand is distal, the one closest to the hand is proximal, but the one in the middle is... middle.

I'm still in the early stages of studying up on anatomy and medical terminology, but based on my readings, I have only seen 'medial' everywhere. It's the first time I've seen 'middle'. Just to make sure my book wasn't wrong, I looked it up elsewhere, and it seems it is indeed the 'middle' bone.

I don't know why this is bothering me so much! But does anyone know why the middle bone of a finger is called 'middle' and not 'medial' with the same ending as the names of the other two bones? I'm wondering if I've been misunderstanding what 'medial' means this whole time or something. There has to be an explanation for what seems like an anomaly! Etymologicalifornia (talk) 04:55, 23 November 2023 (UTC)


 * In anatomy, medial has a very specific meaning, "closer to the of the body or the midline of an organ". The middle finger bone isn't closer to the median plane of the body, so it can't be called medial without confusion. I suppose they could have coined a new word like inbetweenal, but they didn't. —Mahāgaja · talk 07:44, 23 November 2023 (UTC)
 * Incidentally, according to File:Scheme human hand bones-en.svg, the bones between the distal phalanges and the proximal phalanges are called the intermediate phalanges (which the thumb is missing). The next bones in from the proximal phalanges are the metacarpals, which are in the "main body" of the hand as opposed to the fingers. —Mahāgaja · talk 15:27, 23 November 2023 (UTC)

Could Hungarian word “puha” be cognate with the Finnish and Estonian word for “holy”
Since Hungarian is an Uralic language and related with Finnish and Estonian it made me curious about this possibility. In Hungarian “puha” means soft. And in Estonian and Finnish there are “püha” and “pyhä” both mean holy. The words are too similar and the meanings are not so similar but still not that of a long shot. So, can this be a thing or is it just some coincidence? Józio Kovács (talk) 16:30, 24 November 2023 (UTC)
 * The phonology doesn't match; The reconstructed Proto-Uralic form (based on other cognates across Uralic) is . This would yield a Hungarian form like ? Something of that sort. Thadh (talk) 16:58, 24 November 2023 (UTC)
 * In two distantly related languages (which Hungarian and Finnish are) as a rule of thumb similar words are unlikely to be related (except if it's through borrowing) and words which are related typically sound nothing like each other. There are exceptions, like and, but they're rare. —Caoimhin ceallach (talk) 20:24, 24 November 2023 (UTC)
 * A random example: is related to, while  is (probably) related to . Chuck Entz (talk) 21:50, 24 November 2023 (UTC)
 * Apparently Sanskrit atman (breath) and German atmen (breathe) are believed to be cognates, although it might be disputed. Wakuran (talk) 21:46, 25 November 2023 (UTC)
 * The way I'd put it, is that and  are cognate. The ns in  and  are unrelated. Also none of the descendents of  would be recognisable for a German speaker, for instance, meaning "you"! —Caoimhin ceallach (talk) 00:17, 26 November 2023 (UTC)

Loanwords in Chinese: More refs
I noticed that some modern words have missing etymologies. The following websites have digital versions of some dictionaries for reference.

请教白话文词源相关材料 - 资源求助 - FreeMdict Forum

汉语外来语词典 (岑麒祥) (文字版) (1月26日更新) - 汉汉 - FreeMdict Forum 汩汩银泉 (talk) 20:02, 25 November 2023 (UTC)

Calendar's etymology: from Latin calendarium (“account book”), from kalendae (“the first day of the month”), from calō. So "kalendae" is from "calo". But it isn't the case for kalendae's etymology. Kalendae's etymology: ...from Proto-Indo-European root *kelh₁- (“to call, summon”); compare calō. There's a conflict. Duchuyfootball (talk) 02:36, 26 November 2023 (UTC)
 * Is it, really? Although calō isn't related to call, the relationship seems equal to the one for call and calling. Wakuran (talk) 03:11, 26 November 2023 (UTC)
 * What I want to point out is calendar's etymology states that kalendae is from calo, while kalendae's etymology doesn't.--Duchuyfootball (talk) 05:32, 26 November 2023 (UTC)

The etymology of this word falls short in comparison to other numerals. Can someone please add to it? Duchuyfootball (talk) 05:33, 26 November 2023 (UTC)
 * The main etymology seems fairly comparable. It's just that there are fewer cognates listed. Wakuran (talk) 12:28, 26 November 2023 (UTC)
 * Check now :) Leasnam (talk) 04:20, 28 November 2023 (UTC)
 * Great work! Thank you! Duchuyfootball (talk) 13:37, 6 December 2023 (UTC)

寢
RFV of the etymology. — justin(r)leung { (t...) 06:10, 26 November 2023 (UTC)
 * , maybe as well. — justin(r)leung { (t...) 06:24, 26 November 2023 (UTC)
 * Published sources? --Mar vin kaiser (talk) 07:15, 26 November 2023 (UTC)
 * @Mar vin kaiser see the edit history of the references i used. i didnt create any entries beyond what was in any sources so far, even if one dude last time was expecting me to do so. Mlgc1998 (talk) 07:29, 26 November 2023 (UTC)
 * Again, as I said in the edit summaries, none of the sources you cite explicitly link the two words. You also mentioned native speakers possibly “confusing” the two words. Native speakers are generally not the place to go for etymology, as they may know words but don’t necessarily know their history. The misleading thing is that you cite sources that don’t say what we say they’re saying. — justin(r)leung { (t...) 08:38, 26 November 2023 (UTC)
 * What if it's just a coincidence that the sounds are close? We can't claim it to be so unless it's in a source. Maybe if the etymology is super obvious (for example Taiwanese Hokkien terms that sound exactly like Japanese words and the meaning is the same) then that's acceptable. But the etymology you added, the source doesn't say anything about "chhím", and it's hard to claim an ellipsis just from the similarities of the sounds. That's how folk etymology starts like thinking "cockroach" coming from "cock" and "roach". --Mar vin kaiser (talk) 10:43, 26 November 2023 (UTC)
 * @Mar vin kaiser It is in multiple sources, besides the 3 I've already put. Check all the sources listed in ChhoeTaigi and other sites. I've read so much sources there and I don't see you guys doing anything besides taking everything I say on face value that I have to constantly explain everything whilst multiple people here simply click undo and behave dismissively. The Barclay supplement explicitly writes the etymology there and the japanese dictionaries also equate those entries to each other. I don't see how this is not "super obvious". you say that a Taiwanese Hokkien term "sounds exactly like" a Japanese word, and "the meaning is the same". Isn't that what this is? the dictionaries are using the exact same sample sentences, definitions, and recording the same pronunciation and tones. we always only get etymologies from dictionaries and related documents with usage. I mean where else? for the state of this language. Have you individually understood what the japanese definition, mandarin definition, and english definitions they've all written? cuz I sure as hell spent hours translating and understanding what each word in definitions and same sentences that were put. Now if this is considered just a coincidence, how is that any different to the Taiwanese Hokkien term with the Japanese word? Idk tho if this is just cuz it is in the Chinese entries that etymologies are done away like this, because in Tagalog entries, this case would be considered more than obvious already. My friend from davao was telling me he hears "chhím-á" from his mom and another pastor from cotabato was saying that's how he says it but I didn't make a dedicated entry for it cuz only 2002⁺ 台華線頂辭典 was bold enough to make an entry on theirs. Also, @Justinrleung I said native speakers were confusing the ch~chh, meaning the pronunciation, not the etymology. of course, they're not the ones talking about the etymology (the sources are writing that), they simply pronounce it that way, which is some form of unclear allophonic sound they themselves are doing, which I think is maybe cuz they are confused themselves if it should be ch or chh, which the 1931 臺日新辭書 records both an entry for chím-á and chhím-á in the chhím entry, written in japanese of course. Mlgc1998 (talk) 22:41, 26 November 2023 (UTC)
 * I just wanted to start with an apology that I missed the "cf. chhím" in Barclay. Thank you for being more careful in your last edit. The other sources do not make any etymological claims at all, so I don't think they should be cited for the purposes of etymology. The etymology as it stands looks good to me. I think it would still be too big of a leap to say chhím comes from chit-má/chím-má/chím-á. — justin(r)leung { (t...) 01:41, 27 November 2023 (UTC)
 * Thanks for making a more careful edit. Borrowings are easier to prove than derivations with the same language, requires more evidence basically. I think with the evidence you've gathered so far, it's okay to say that its possibly etymologically related to, but I agree with Justin that it may require more evidence. I think the difference here with what we do in Tagalog is that I think a lot of the derivations we do in Tagalog are more recent, so it require less proof to give an etymological link. Meanwhile, the more "far back" a derivation is, the harder to prove the link. The danger is quite high in making false etymological claims, similar to how past etymologists have been proven wrong because they were hasty in claiming a certain etymological link. --Mar vin kaiser (talk) 06:12, 28 November 2023 (UTC)
 * @Justinrleung @Mar vin kaiser Ok, Thank You, guys!! Careful approach is needed in voluntary unpaid work for fine output. I apologize as well for any hasty interpretations. Mlgc1998 (talk) 00:33, 29 November 2023 (UTC)

νωρίς
There is no etymology given. Duchuyfootball (talk) 05:30, 27 November 2023 (UTC)
 * Added. —Mahāgaja · talk 07:21, 27 November 2023 (UTC)
 * Thank you! Duchuyfootball (talk) 03:51, 28 November 2023 (UTC)

toxicum, toxicus
There's significant overlap in the descendants. Aren't the modern adjectives from toxicus? Ultimateria (talk) 06:18, 27 November 2023 (UTC)
 * I would expect toxicum to only list noun descendants; adjectives would more appropriately go at toxicus. The noun does seem to have had a fair number of descendants: we have noun entries for Friulian, Italian, Portuguese, Romanian, Romansch, Spanish, Venetian. In the case of languages that don't have a noun entry yet, it's hard to tell if that actually means that a noun doesn't/didn't exist.--Urszag (talk) 06:50, 27 November 2023 (UTC)

szókincs, szótár
Through Google Books ancientness research I came to the conclusion that szókincs was a probably a calque of German Wortschatz, itself from Dutch woordenschat. The peculiar combination of word and treasure points to it too.

In a similar fashion I speculate that szótár is derived from Latin dictionarium. @Panda10 disputes this, instead asserting that it is a calque of German Wörterbuch.

Any third opinions on these matters? Synotia (talk) 10:16, 27 November 2023 (UTC)
 * German also has the (dated) term Wortvorrat, which is an even better fit, although Wortvorrat might still be a semi-calque of dictionarium. Wakuran (talk) 12:05, 27 November 2023 (UTC)
 * They all might be calques of rather than  since the first parts mean 'word' rather than 'speaking'. —Mahāgaja · talk 21:13, 27 November 2023 (UTC)
 * On the other hand, doesn't mean 'treasure'. Is  calqued from something else? —Mahāgaja · talk 21:18, 27 November 2023 (UTC)
 * or . “Treasure of blabla” is a common book title. I don’t know exactly, have not paid particular attention to book titles, which are marketing lies, but from antiquity it is common to give books names accentuating their copiosity, as was easier with fewer fields of study available. . .  is an “ocean”. Only superlatives. Fay Freak (talk) 21:56, 27 November 2023 (UTC)
 * If szótár were calqued from Wörterbuch it would be szókönyv. Indeed vocābulārium seems like a better match. —Caoimhin ceallach (talk) 23:01, 27 November 2023 (UTC)
 * As the Scandinavian languages have words like ordförråd/ ordforråd that seem to be directly derived from German Wortvorrat (or a related variety), it seemed likely that it was the immediate source for Hungarian, as well. Wakuran (talk) 12:21, 28 November 2023 (UTC)
 * Can mean "dictionary" or "glossary" though? A cursory search didn't turn up anything. —Caoimhin ceallach (talk) 19:43, 28 November 2023 (UTC)
 * DWB ( / ) only seems to list the sense "vocabulary". I wouldn't know which sense in Hungarian that would be the oldest. Wakuran (talk) 01:27, 29 November 2023 (UTC)
 * Have you guys found anything contradicting my theory? Because Szókincs coming from Wortschatz seems very plausible considering the earliest trace I found being a German literature review by Hungarians of German descent. Synotia (talk) 08:36, 29 November 2023 (UTC)
 * They may all be calqued from . There is possibly an, but I could not find uses of this term dating back to classical times. Note, though, that the use of  for a collection of words is found in the Greek title of Henri Estienne's bilingual  / Thesaurus graecae linguae, published in 1572.  --Lambiam 13:00, 29 November 2023 (UTC)

About szókincs: You mentioned that you conducted a Google Books ancientness research. I assume that the comment "First attested in a 1837 article about German literature by Ferenc Toldy and Mihaly Vörösmarty" is the result of that research. Would you share the link to the source? It would be helpful to make a decision, especially about the year. It is not enough to provide links to the two individual's Wikipedia page. The Új magyar etimológiai szótár (New Etymological Dictionary of Hungarian) by Károly Gerstner doesn't contain this word specifically, even though it lists multiple compound words starting with szó. However, at the end of the article szó-, there is a comment: ''A 2. csoport szavai tudatos szóalkotással keletkeztek a nyelvújítás korában, főleg német mintára; vö. ném. Wörterbuch, Wortschatz, Wortspiel stb.'' (The words of the 2nd group were created by conscious word creation during the language reform, mainly on the German model; cf. German Wörterbuch, Wortschatz, Wortspiel, etc.). Panda10 (talk) 18:00, 30 November 2023 (UTC)


 * The fact that could secondarily mean "vocabulary" also speaks for a Latin origin. —Caoimhin ceallach (talk) 19:44, 30 November 2023 (UTC)

No etymology given. Duchuyfootball (talk) 03:51, 28 November 2023 (UTC)
 * added. Leasnam (talk) 04:34, 28 November 2023 (UTC)
 * Thank you! Duchuyfootball (talk) 13:39, 6 December 2023 (UTC)

faex
Claims of *eh₂ > ae and *dʰr > f are dubious, at best. De Vaan lists no etymology. Saph668 (talk) 03:55, 28 November 2023 (UTC)
 * I agree those sound changes are unusual, but I also think we can do better than the experts in this sort of situation. The PIE root we give is actually dʰrā́ks, with no laryngeal, and we suggest it's a substrate borrowing.  I'm not sure what to think about that, .... do voiced aspirates regularly appear in substrate borrowings? .... but if we can agree that it's not a true PIE root, it might explain an irregular inheritance in one branch and therefore still be something we can point to as the original word.  We do also have fraces, which might be a doublet of the same word, and looks more regular.  — Soap — 08:19, 28 November 2023 (UTC)
 * The breathy plosives don't appear in any other substrate loan to my knowledge. And sure, you could say that the substrate status could explain it, but that goes against all precedents for regular sound changes in Italic. It's probably a Mediterranean or Tyrrhenian loan unrelated to fracēs… but I think it would be best to keep it in the descendants and mark it as uncertain, as it is now, and tone down the certainty on the etymology listed for faex. Also cf Reconstruction talk:Proto-Indo-European/dʰrā́ks for discussion of the laryngeal. The whole root is dubious, really. Saph668 (talk) 21:56, 28 November 2023 (UTC)
 * Updated the etymology at faex. Take a look there. Saph668 (talk) 01:13, 29 November 2023 (UTC)

वार (Hindi)
The original etymology given can be seen at Urdu وار, but at some point user عُثمان changed this to Punjabi ਵਾਰ, mentioning the 'comparative dictionary of Indo-Aryan languages' (so here). Said user is now apparently blocked, so I'm not sure if his opinions should be taken seriously. Exarchus (talk) 09:34, 28 November 2023 (UTC)

самостоятельный (Russian)
Calque of German selbstständig, like Dutch zelfstandig? Synotia (talk) 07:02, 29 November 2023 (UTC)


 * Probably. See also 🇨🇬. Tollef Salemann (talk) 07:23, 29 November 2023 (UTC)
 * @Synotia, @Tollef Salemann: Can be a parallel development. A question is "who borrowed from who". If we don't know (no source) better not to make up. I find analogous phrases in German as in Slavic and phrases but have no idea who started them, e.g. =  meaning "unharmed, OK", as in "are you OK?". Anatoli T. (обсудить/вклад) 07:33, 29 November 2023 (UTC)
 * The German term is possibly a calque of . The word is attested in German (in the form ) at least as early as 1539. A 16th-century borrowing from Russian to German appears unlikely. --Lambiam 12:16, 29 November 2023 (UTC)
 * Take a look at Estonian iseseisev Synotia (talk) 13:46, 29 November 2023 (UTC)

ליאָדעס
Since so many dictionaries attest the plural form while only one has the singular form, is it possible that the singular form was actually a back-formation from the plural? We do have precedent of Yiddish taking a Belarusian plural-only word and adding to form the plural word within Yiddish, i.e. . Not even the JNW has. Insaneguy1083 (talk) 17:42, 29 November 2023 (UTC)
 * Are the Slavic forms plural, though? Aren't they rather mass nouns? Wakuran (talk) 23:31, 29 November 2023 (UTC)
 * yes, but no. Regardless, they are inflected like plural nouns, and are subsequently marked as plural-only nouns in Yiddish. The -(e)s ending appended looks to me like a calque of the  ending, making the full term a partial calque. Insaneguy1083 (talk) 03:50, 30 November 2023 (UTC)
 * Could the -e-ending have been reinterpreted as a diminutive, like English -ie, Dutch -je, or is that unlikely? Wakuran (talk) 11:55, 30 November 2023 (UTC)
 * isn't a diminutive ending in Yiddish, so no. Insaneguy1083 (talk) 18:24, 1 December 2023 (UTC)

English debacle (verb) and French débâcler
I'm beginning with the English entry first, but the conclusion will ultimately affect the French entry as well.

Some issues I'm seeing with the etymology, which coincidentally matches a lot of the other etymologies, like etymonline.com.
 * 1). Old French doesn't exist. The desbacler mentioned in the etymology is actually the Middle French term meaning "to unmoor [boats], removing them from a dock to make way for arriving loaded boats". I may have been the one who added this as Old French, if so, it's my fault. It needs to be removed.
 * 2). There is at least one modern transitive use of French débâcler in 1589 meaning "to open [by removing a crosspiece]" (literally to remove the stick from, un-brace), a sense which I do not find in the modern usage (please correct me if this is wrong). What I do see as the usual modern sense appears to be intransitive and refers to "breaking up of ice on waterways" literally "unfreezing, becoming unfrozen".

Deconstructing the word to its base, early Modern French ; and current French, whose etymology is admittedly uncertain, there are yet other issues. There is one Late Old French attestation (1292) meaning to "secure or close a door with a wooden crosspiece". There is no trace of the word in The Dictionary of Middle French (1330-1500). It appears again in one isolated attestation in 1596 (Ils baclent et estoupent sa bouche - "They block and stop up (choke) his mouth") with the meaning of "close, block", and is assumed to have survived throughout Middle French indirectly in, mentioned above.

Then there is the Flemish intransitive verb, a frequentative of (cf. ).

When reviewing the above, I'm led to infer that there may be at minimum 2 separate French etymologies for débâcler: a transitive one that literally means to "un-brace, unblock" and has clear connection to the word for "wooden staff, stick" and whose base has now morphed into a modern usage meaning "to botch"; and an intransitive one that doesn't seem to share the same origin, but which instead aligns more closely with the Flemish meaning "to crust over with ice", but only in the reverse sense of "to become un-crusted over with ice", hinting at an intermediary Early Modern  that has since been lost.

The etymologies for English and French certainly need some rework, but I wanted to make sure I have everything tied down before I started anything. Any thoughts/concerns (?), please share ! Leasnam (talk) 16:06, 30 November 2023 (UTC)