Wiktionary:Etymology scriptorium/2017/January

discussion rooms: Tea room • Etym. scr. • Info desk • Beer parlour • Grease pit ← December 2016 · January 2017 · February 2017 → · (current)

---

puree, purée

These entries should be merged, but they have different etymologies- which is preferred? DTLHS (talk) 23:33, 31 December 2016 (UTC)

It looks to me like it was the noun that was borrowed, not the verb, so I think French purée would make more sense. If the noun was just nominal use of a participle, I suppose we could link to the infinitive, but it goes way, way back as a noun, and etymonline even expresses doubt as to whether it really came from the verb at all. Chuck Entz (talk) 02:13, 1 January 2017 (UTC)

Turkish bayram and its cognates

Could someone versed in Iranistics help out with this? Clauson gives Persian baḏrām and Nişanyan derives it from Middle Persian paδrām or Sogdian patrām with an Indo-Aryan etymology which I'm unable to judge.

However there are some problems. For one, Gharib glosses the Sogdian word (ptrʾm) as "calm, peace" which doesn't correspond that well semantically with "feast, celebration" of Turkic cognates and connects it with "MP padrām". The second problem is that I'm unable to find the Middle Persian word in dictionaries, neither Durkin-Meisterernst nor Mackenzie seem to have it.

Interestingly there is another Iranic root that is similar phonologically and has a closer semantic, "*badra" with a cognate in Sanskrit भद्र (bhadra, “blessed, happy”), but other than Avestan, all languages have changed -dr- to -hr- and diverged in semantics to "share, portion".

There is an alternative etymology more popular in Russia, that relates the word with Mongolian баяр (bajar) and its cognates. this would require the ð of Karakhanid بَذْرَمْ (baḏram) to be secondary, which I found doubtful at first, but considering that Khakas material suggests *y instead of *ð and that Karakhanid form could have easily been corrupted by the Persian word I'm starting to see some of its merit. Crom daba (talk) 03:44, 2 January 2017 (UTC)

See this page from Eren, Hasan (1999) “bayram”, in Türk Dilinin Etimolojik Sözlüğü [Etymological Dictionary of the Turkish Language] (in Turkish), Ankara: Bizim Büro Basım Evi.

I don't speak Turkish, but judging from a Google translation, the origin is disputed. --Vahag (talk) 09:44, 2 January 2017 (UTC)

Might you have it typed out (so I can google translate it), I don't have a Turkish keyboard and pasting that together would last forever. Crom daba (talk) 19:31, 2 January 2017 (UTC)

Here you go. OCR mistakes are possible.

bayram ‘dinî veya ulusal bakımdan kutlanan gün veya günler’ ~ Tkm bayram. - TatK bayram. Tatarlar beyrem biçimini de kullanırlar. - Nog bayram. - KKlp bayram. - Kzk meyram. - Krg mayram. - Özb bayram. - Şart mayram. - Alt, Tel, Kuğ, Şor, Sag, Koy, Kaça, Küer payram. - Bar, Küer peyrem. Eski Türkçeden başlayarak kullanılır. Orta Türkçede baöram olarak geçer. Kâşgarlı Mahmud’a göre, Oğuzlar bayram biçimini kullanırlar. Bu örnekte Oğuzlar -δ- sesini kurala uyarak -y- ’ye çevirmişlerdir. Kâşgarlı Mahmut, bu anlamda öz Türkçe bir söz olduğunu seslendirmiştir. Kökeni karışıktır. Doerfer’e göre (TMEN 823), Kâşgarlı Mahmud’a borçlu olduğumuz baöram biçimi bayram’ın Türklüğüne tanıktır. Buna karşılık Clauson (ED 308 a), İran kökenli bir alıntı olduğunu kesin olarak dile getirmiş, Farsça badrâm ‘a delightsome place’ biçimini vermiştir. Onun Farsça bayram ‘a Muhammadan solemn festival’ biçimini göz ardı ettiği anlaşılıyor. Sevortyan’a göre (ESTJa 1978, 35-36), Orta Türkçe baöram biçimi, Farsçadan geldiği yolundaki görüşü çürütmüştür. Türkçeden belli başlı komşu dillere de geçmiştir. Doerfer (TMEN 823) bu dillerdeki alıntıları saymıştır. Onun topladığı verilere Arapça (Suriye) beyrâm ‘fete’ biçimi de eklenebilir. Tacikçe bayram, mayram biçimi Özbekçeden geçmiştir (Doerfer : TLT 34). Tuvalar bayır biçimini ‘bayram, tören’ olarak kullanırlar. İlk bakışta bu biçim ile Türkçe bayram arasında açık bir benzerlik olduğu göz ardı edilemez. Ancak bayram ile bayır arasında etimolojik bir bağdan söz edilemez. Räsänen’e göre (V 57 a), Sevortyan’ın bayram'ı ‘şölen’ anlamına gelen bay biçiminden yola çıkarak *bayır ’dan getirmesi ve bayrak ile bayram arasında olası bir bağdan söz etmesi yanlıştır. Tuvaca bayır büyük bir olasılıkla Moğolca bir alıntıdır : Moğ bayar ‘Freude, Fröhlichkeit’ Bk. Rassadin : FLTJa 158. Ancak Rassadin (Zaimstvovanija) Tuvaca bayır’ı vermemiştir. Räsänen : LTS 230 ; V 54 b; Clauson : ED 308 a ; Sevortyan : ESTJa 1978, 35-36; Bazin : Lazard Arm 31.

Sevortyan in ESTJa denies the Iranian etymology and argues for a native origin.

The Middle Persian word that you couldn't find is Manichaean Middle Persian pdrʾm (padrām, “in peace”). See Boyce, Mary (1977) “pdrʾm”, in A word-list of Manichaean Middle Persian and Parthian (Acta Iranica 9a, Série 3 – “Textes et mémoires”, vol. 2-supplément)‎^[1], with a reverse index by Ronald Zwanziger, Leiden, Tehran-Liège: E.J. Brill, Bibliothèque Pahlavi, page 68. --Vahag (talk) 09:50, 3 January 2017 (UTC)

/n/ in Indo-European infinitives

Germanic, Greek, and Persian/Kurdish have the consonant /n/ in infinitives. (Maybe other languages, too, but I'm only aware of these.) Just for my information: Is there cognacy between (some of) these endings, or is it coincidental? Thanks a lot. Kolmiel (talk) 17:23, 4 January 2017 (UTC)

Accroding to Ringe, Germanic infinitives come from a remodeling of the verbal adjectives in *-nós as pre-PG *-o-nom > PG -aną which is affixed to the present stem. This does not seem to be a formal match for Greek's infinitives in -ειν (-ein), -σαι (-sai), or -ναι (-nai), but I don't know their origin of the top of my head without consulting Sihler. —JohnC5 17:36, 4 January 2017 (UTC)

Sihler says the -en element of the Greek infinitive is "an endingless n-stem loc.", see page 608 for details. Crom daba (talk) 18:22, 4 January 2017 (UTC)

If that is true, then those two are unrelated. Aren't Persian infinitives something like -dan / -tan? —JohnC5 18:49, 4 January 2017 (UTC)

Yes. Although this -d-/-t- is present in all forms of the perfect stem, from which the infinitive is formed by adding -an. Kolmiel (talk) 18:56, 4 January 2017 (UTC)

Historical grammar of the ancient Persian language by Edwin Lee Johnson derives the Persian infinitive from "the datives of n-stems" Crom daba (talk) 19:13, 4 January 2017 (UTC)

That would mean a certain relation between the Greek and Persian infinitives? Kolmiel (talk) 19:18, 4 January 2017 (UTC)

They would be related inasmuch as they are the same nominal ending, but they represent unrelated innovations since each are separately affixed to dissimilar, post-PIE verbal stems. —JohnC5 17:54, 5 January 2017 (UTC)

Yes, okay. Thanks again. Kolmiel (talk) 21:19, 5 January 2017 (UTC)

родительный

The etymology says that this, the Russian word for “genitive”, is a calque of Ancient Greek γενική (genikḗ), but the Greek word is an adjective derived from the word for “kind”, while the Russian word is an adjective from the word for “parent”, роди́тель (rodítelʹ). (It could also come from the verb “to beget”, роди́ть (rodítʹ), using a slightly different morphological analysis.) The Greek noun from which the adjective comes is related to the verb “to be born”, γίγνομαι (gígnomai), so there is a semantic relationship, but it is somewhat distant. It would be nice to have verification of the Russian etymology. I suppose it is likely that the word was intended as a calque, but it isn't a precise one. If it comes from the verb, it would be an exact calque of the Latin word genetīvus. — Eru·tuon 06:35, 5 January 2017 (UTC)

I think this is a translation but not a calque. Also compare Latin genetivus. Crom daba (talk) 11:32, 5 January 2017 (UTC)

I agree. It seems more likely that Russian case names were translated from Latin or a West European language. -- Anatoli T. ^{(обсудить}/^вклад) 11:45, 5 January 2017 (UTC)

Huh, I thought the definition of a calque was that it's a translation. How can something be a translation and not a calque? — Eru·tuon 00:13, 6 January 2017 (UTC)

A calque is a translation of individual parts. I think this is a calque, but from Latin rather than Greek. Compare Slovene rodilnik, based on the same verbal base. Slovene would not be expected to have significant Greek influence. —CodeCat 00:18, 6 January 2017 (UTC)

Well, I thought that a single-morpheme word could be a calque, such as words for “mouse” in various languages being used for the cursor-controlling computer tool due to the influence of English. Wikipedia indicates that this type of situation is called semantic calque. A new word is not created, but an additional meaning for an already existing word is imported from another language. Fortunately this definition is unnecessary in this case. I think you, @CodeCat and @Atitarev, are probably correct, and I changed the etymology. — Eru·tuon 01:34, 7 January 2017 (UTC)

I'd like to point out that russian "род" has the same stem and means roughly "kind". 147.32.219.28 16:40, 1 February 2017 (UTC)

一日

RFV of the etymology. ばかFumiko￥^talk 17:59, 5 January 2017 (UTC)

• Freely confirmable at Kotobank.

I have updated and expanded the entry. ‑‑ Eiríkr Útlendi │^{Tala við mig} 02:55, 6 January 2017 (UTC)

弥生

RFV of the etymology. ばかFumiko￥^talk 18:13, 5 January 2017 (UTC)

• Freely confirmable at Kotobank. ‑‑ Eiríkr Útlendi │^{Tala við mig} 02:55, 6 January 2017 (UTC)

RFV of etymology of Nordic smá

The articles on English small and Old Norse smár (> Modern Scandinavian små) claim that they are cognates, both being derived from Proto-Germanic *smalaz. However, both Kroonen and Orel reconstruct different PG proto-forms - *smala- and *smēha-, and none of the other reference works accessible to me assert common origin for both. Also, Old Norse smalr, Scandinavian smal ('thin, narrow') exist separately from Old Norse smár / Scandinavian små, and I'm not aware of any North Germanic sound change that would have elided the *l in *smalaz. I suspect the editor just went on intuition here.--95.42.29.86 18:43, 5 January 2017 (UTC)

Fixed. Leasnam (talk) 19:37, 5 January 2017 (UTC)

Fried Liver Attack

Could someone with knowledge of Italian verify this etymology? Does the Italian name really exit? Does it antedate the English terms (presumably so as Google books has no cites prior 20thC).

Partial translation of Italian Giuoco Fegatello (“fried liver game”).

Also, does the metaphor mean anything other than the opening "makes it hot" for Black?Sonofcawdrey (talk) 21:08, 6 January 2017 (UTC)

Fegatello Attack

Could someone with knowledge of Italian verify this etymology, likewise?Sonofcawdrey (talk) 21:09, 6 January 2017 (UTC)

swatch

Could this word be related to the following Dutch words: zwatsen (“to move forth and back”) (http://www.vlaamswoordenboek.be/definities/term/zwatsen), compare related zwachtel (“bandage”) (http://www.woorden.org/woord/zwachtel)? And English swathe (“bandage”). --Eliot (talk) 06:00, 7 January 2017 (UTC)

cattus

The etymology for Latin cattus says it is Afro-Asiatic and gives Berber, Nubian, and Egyptian as evidence. This is all well and good but Wikipedia says that Nubian is Eastern Sudanic and not Afro-Asiatic. I know nothing about the Nubian languages, so wanted to bring this up for anyone competent to resolve this contradiction. Jan sewi (talk) 22:03, 7 January 2017 (UTC)

烕

RFV of the glyph origin/etymology. Sounds like a folk etymology. —suzukaze (t・c) 22:23, 7 January 2017 (UTC)

I rewrote the glyph origin. The original explanation is unlikely although I found a similar explanation in 汉字源流字典 (maybe unreliable). — justin(r)leung _{{ (t...) | c=› }} 23:22, 7 January 2017 (UTC)

便

RFV of the glyph origin/etymology. Sounds contrived. —suzukaze (t・c) 00:17, 8 January 2017 (UTC)

It's from Shuowen: 安也。人有不便，更之。从人、更。(unless I misinterpreted it) Wyang (talk) 00:22, 8 January 2017 (UTC)

Hm, OK. —suzukaze (t・c) 03:11, 8 January 2017 (UTC)

@Wyang, Suzukaze-c: It may not be as simple. This suggests that 便 may be the original character for 鞭. — justin(r)leung _{{ (t...) | c=› }} 03:23, 8 January 2017 (UTC)

Thanks, I added this in. Wyang (talk) 03:46, 8 January 2017 (UTC)

@Wyang: I've corrected it to match the explanation in the book, but I'm not sure what they mean by "以初文冕为声". — justin(r)leung _{{ (t...) | c=› }} 04:00, 8 January 2017 (UTC)

Thanks for finding 𠓥! I think the top may be a person, and the bottom from 攴. I'm not sure about 以初文冕为声 either. Wyang (talk) 04:06, 8 January 2017 (UTC)

Proto-Italic verb-inflection table templates

I believe @CodeCat is managing these? I've noticed that they give the ending of the 2nd person plural passive – the ancestor of Latin -minī (by the way, this sense is not mentioned there) – as *-m(e?)n(ai?). However, there's a proposal (mentioned, for example, in Meiser 1998: 219) to derive this ending from the passive participle from which fēmina and possibly alumnus and columna are also thought to descend. If this suggestion is accepted as plausible, the ending could be given as -manoi?. --Florian Blaschke (talk) 13:05, 8 January 2017 (UTC)

Really inflection tables for Proto-Italic seem like an iffy idea altogether, given that the only branch we know well is Latin and its descendants. It's going to end up as "Latin, but with some earlier traits reconstructed whenever we have Sabellic or Faliscan evidence on what happened exactly". This will work decently for general phonology (there's no reason to oppose Proto-Italic reconstructions per se), but whenever Latin has innovated with respect to PIE, & other Italic evidence is missing, there will be little chances for figuring out if a thing happened in Proto-Italic or only in early Latin.

Wider-reaching morphological reconstruction is probably still doable to an extent, by examining parallels and whatnot — but how much of what we have is actually backed up by literature, and how much has been just eyeballed together? I notice that even with phono, we e.g. do not have *xʷ (*formos pro *xʷormos), *θ (*faznom pro *θaznom), even though they're likely reconstructible for Proto-Italic proper. (*θ > f is pan-Italic but postdates *ts > θ in Sabellic, *θ > *ð > d in Latin; *xʷ > f in Sabellic is similarly probably simultaneous with *kʷ > p, and in Latin it's again later than medial voicing to g). So if CodeCat or someone else wants to put together a summary about how much morphology exactly we can reconstruct for PIt., perhaps either Wikipedia or WT:About Proto-Italic would be a better location.

(Similar issues might come up with other less-studied intermediate proto-languages as well — Proto-Celtic, Proto-Hellenic, Proto-Indo-Iranian, etc.)

— And as long as we're at Proto-Italic: we also seems to be missing *əm and *ən (from PIE *m̥, *n̥; to Latin em, en ~ Sabellic 1st syllable am, an), as in *kentom pro *kəntom; maybe also e.g. *nowem pro *nowəm (though in non-initial syllables Sabellic too shows eN). --Tropylium (talk) 14:42, 9 January 2017 (UTC)

गविष्टि

I gather that गविष्टि (gaviṣṭi) is the Sanskrit word mentioned by the linguist in the movie Arrival. She says it means “desire for more cows”, and this page says something similar. Can someone with Sanskrit knowledge provide an etymology? — Eru·tuon 00:39, 9 January 2017 (UTC)

@Erutuon: I've updated the entry. To the best of my understanding, the analysis presented in the movie is not really true, viz. the semantics are “eagerness for cows” → “eagerness, desire” → “eagerness for battle” → “battle.” That's not to say that the original meaning would not be obvious to speakers, but it's hard to say whether the intention (“eagerness for cows” = “battle”) existed synchronically. —JohnC5 02:07, 9 January 2017 (UTC)

critique

The OECD's etymology is more specific than ours, claiming, "mid 17th century (as a noun): from French, based on Greek kritikē tekhnē ‘critical art’." Should we include this? ---> Tooironic (talk) 06:21, 9 January 2017 (UTC)

What source says hai is tōon for 海?

@TAKASUGI Shinji, Eirikr What source says hai is tōon for 海 as noted in the kanji entry? ばかFumiko￥^talk 01:57, 10 January 2017 (UTC)

• I'm not sure at the moment. I just went through the history and found that User:Lo Ximiendo added that reading in October 2014. The JA Wikt page at ja:海 shows the hai reading as 唐音 (tōon, “Tang reading”), added by JA Wikt user ja:User:Tây yuán sang in this edit in March 2013.

I don't have access to the Kanjigen or similar specialist kanji dictionaries; perhaps one of those would list that reading? It is definitely used, albeit only in limited contexts, as in the Japanese reading for 上海 (Shanhai), which even the Microsoft IME recognizes (I type shanhai in the Latin alphabet, the IME generates しゃんはい as the kana rendering, and offers to convert that to 上海). ‑‑ Eiríkr Útlendi │^{Tala við mig} 02:16, 10 January 2017 (UTC)

Kanjigen doesn't appear to list any toon reading of 海. Shanhai can be considered as just another direct borrowing from Mandarin, like paozu, Gonbao (see Category:Japanese terms borrowed from Mandarin for more), hence its logical irregularity rather than its supposed, unattested rarity as an actual on reading. You can just type テレビ, press space, and てれび will appear in the prompted list, so the fact that IME renders input as hiragana shouldn't be a factor. ばかFumiko￥^talk 02:44, 10 January 2017 (UTC)

• Apologies for any confusion -- I was instead offering evidence that the hai reading is accepted. Whether it is tōon or otherwise, I have no data. ‑‑ Eiríkr Útlendi │^{Tala við mig} 02:51, 10 January 2017 (UTC)

唐音 on Chinese Wikipedia has 海 ハイ. I don’t know whether it is a reliable article or not. — TAKASUGI Shinji (talk) 03:57, 10 January 2017 (UTC)

恐龍

RFV of the etymology. ばかFumiko￥^talk 14:45, 10 January 2017 (UTC)

Chinese Wikipedia has a reference. — TAKASUGI Shinji (talk) 15:46, 10 January 2017 (UTC)

공룡

RFV of the etymology. ばかFumiko￥^talk 14:47, 10 January 2017 (UTC)

Middle English open syllable lengthening and /ʃ/

Middle English experienced Open syllable lengthening, where vowels in stressed syllables were lengthened when the syllable ended in a vowel. I'm curious what the effect was when the next syllable began with /ʃ/. Did open syllable lengthening apply here, or was it blocked? The latter would indicate that it was actually a geminate, /ʃː/. —CodeCat 01:01, 11 January 2017 (UTC)

I cannot think of any examples where a vowel lengthened. However, I can think of several that remained short (asche, waschen, flusshen) and a few that were originally long and became short (adweschen, maschen, ruschen, wischen) Leasnam (talk) 14:06, 11 January 2017 (UTC)

Interesting. That would suggest that it was still long in Middle English, and thus that we should treat it as long for Old English too. I wonder what can be said about length of word-initial sc- in Old English. —CodeCat 15:22, 11 January 2017 (UTC)

For what it's worth, the same would apply to Middle High German -sch-, which also hinders open-syllable lengthening. The shift from /sk/ to /ʃ(ʃ)/ is considered to have taken place in late Old High German. How do we know that it was earlier in English? Kolmiel (talk) 16:03, 11 January 2017 (UTC)

I wonder if compound words could provide evidence for initial sc-. — Eru·tuon 19:22, 11 January 2017 (UTC)

I don't think it would give much evidence for anything. It could be allophonically short initially, long medially, so that compounding would just trigger the long allophone. —CodeCat 19:25, 11 January 2017 (UTC)

Again, -- this may be a stupid question -- how do we know that scip was already pronouned /ʃip/ at all, rather than /ʃtʃip/ or /ʃcip/? Kolmiel (talk) 20:55, 13 January 2017 (UTC)

Actually, that's a very interesting question. I'm not aware of any direct evidence for the pronunciation of palatalized sc, or of palatalized c in the Old English era. Old English didn't have contact with languages such as Persian that had sh or ch sounds that might provide evidence if English had adopted loanwords that contained those sounds. All we know is that palatalized sc at some point became a palato-alveolar fricative, and c a palato-alveolar affricate, at the very latest by the Middle English era. Presumably they transitioned from velar to palatalized (front-) velar and then somehow to palato-alveolar, but at what point the sounds reached each stage is uncertain. Perhaps there is evidence I'm not aware of, though. — Eru·tuon 02:15, 14 January 2017 (UTC)

There is evidence at least for g, since it is used in words with original /j/. Of course, palatalization of c and sc did not necessarily happen at the same time as g. But I'm sure there is more internal evidence if you search hard enough. --WikiTiki89 15:46, 17 January 2017 (UTC)

JFTR, in OHG, /sk/ is thought to have changed to /sx/ as a first step, just like in Dutch, where this stage is preserved, and the evident origin of the sch spelling in both languages – more precisely, in early OHG (8th century) the stage /sk/ appears to have been preserved, but from about 800 on spellings like sch and sg occur (besides sc and sk), and become more and more common, until they become increasingly dominant from the 11th century on. The change from /sx/ to /ʃː/ may have only happened in the MHG period, however.

The fullest treatment of the problem for OE in a handbook I've found is in Brunner. The best evidence seems to be alliteration – it indicates that Ælfric already has a pure sibilant. By the time Norse loanwords entered English (compare this list), the pure sibilant stage must already have been reached even before back vowels. Only Norse words with skr- are borrowed into English with shr-. --Florian Blaschke (talk) 11:14, 1 September 2023 (UTC)

Topolect

The etymology at topolect says, "Coined by sinologist Victor Mair", and refers to a 1991 paper by Mair. That paper says in relevant part, "[John DeFrancis] ingeniously proposes 'regionalect'. As an alternative, I would suggest 'topolect', which aside from being fully Greek in its derivation has the added advantage of being neutral with regard to the size of the place". There is no suggestion that the word is a fresh coinage, though it's possible that Mair derived it from Greek roots.

There is scattered attestation of topolect since the 1960s (e.g. this calque of örtliche Sprachform). The word becomes somewhat more common after 1991. It seems to appear mainly in studies of Jewish "lects" in the 1970s and 1980s, and Chinese topolects since the 1990s, but were talking about very small numbers.

Is Mair's calque of 方言 (fāngyán) really a coinage? Seems more like a refinement of usage. Cnilep (talk) 08:48, 11 January 2017 (UTC)

Spanish and Portuguese acontecer

User:ProudPrimate has added many bogus etymologies over the years. I'm trying to clean some of them up. He/she claims that acontecer comes from Latin contingere, which seems doubtful to me. Any ideas for its actual source? Looks like a- + -cont- from somewhere, + -ecer < Latin -escere. Benwing2 (talk) 03:49, 12 January 2017 (UTC)

That must be from the RAE: for acontecer they say "De contecer", and for contecer, "Del lat. contingĕre, en lat. vulg. contingescĕre." DTLHS (talk) 03:55, 12 January 2017 (UTC)

The etymology is correct, but it would benefit from more information, like that the words derive from the inchoative form contingescere (attested?) and have a prothetic a- (very common in Iberian Romance verbs). — Ungoliant ^(falai) 11:20, 12 January 2017 (UTC)

La lengua de Cervantes mentions a word contir with variants acuntir and cuntir and derives them from computō (through *computēscō?). It further says that accontingēscō gave aconteñecer. KarikaSlayer (talk) 01:54, 14 January 2017 (UTC)

Yes, that was my concern. The loss of expected -eñ- is irregular and I can't think of any other words where it occurs. Benwing2 (talk) 14:20, 14 January 2017 (UTC)

According to my etymological dictionary (Dicionário etimológico da língua portuguesa, Nascentes) there was an intermediate form *contigescere in VL. — Ungoliant ^(falai) 16:05, 14 January 2017 (UTC)

That would imply an expected aconteecer in Old Portuguese and maybe Old Spanish (I think at least; possibly the two intertonic vowels would compress early). Do either occur? Benwing2 (talk) 23:13, 14 January 2017 (UTC)

ginseng

Does this really come from Mandarin? Both Wiktionary and the OED claim this, but "gin" is very different from "ren". Perhaps it's from Min Nan? ---> Tooironic (talk) 09:00, 12 January 2017 (UTC)

The first use seems to be in 1654, when Gin-ſem appeared in Martino Martini (1614 – 1661)'s DE BELLO TARTARICO HISTORIA (English version: Bellum Tartaricum, or the conquest of the great and most renowned Empire of China by the invasion of the Tartars [i.e. Manchus]). The Latin passage was:

Hi ergo quotannis per vicinas Leaotung terras, Sinas, ingrediebantur, ad commercia tanquam subditi vel amici admissi, bellum autem non cogitabant, serè ad egestatem redacti. Adserebant autem Ginsem, radicem, illam à Sinis adeo aestimatam, ...

And the corresponding English version was:

And those Tartars every year, either as Subjects or Friends, came into China by the Province of Leaotung to traffick with the Inhabitants; For, being brought to poverty and misery, they thought no more of making war against China. The Merchandise they brought were several, as the root cal'd Ginsem, so much esteemed amongst the Chinese, ...

Judging from the context in which ginsem was found, and the use of other transcriptions in the book, such as Leaotung (遼東), Niuche (女直) on the same pages as ginsem, and the map used in the book, my impression was that the transcriptions relatively consistently reflected the pronunciations in Middle Mandarin, or Mandarin (late imperial lingua franca) spoken at the end of the Ming Dynasty.

The transcription gin-sem (人蔘) bears apparent similarity to the transcription system invented by Matteo Ricci (1552 – 1610), in the book Xizi Qiji (downloadable here); In Ricci's system, this word would be rendered as gîn-sēn, which, according to Coblin (A Diachronic Study of Míng Guānhuà Phonology, 2000), represented phonetically /ʐin sɛŋ/ (discussion on 人 gîn), which is quite close to modern Beijing Mandarin /ʐən ʂən/. Wyang (talk) 11:36, 12 January 2017 (UTC)

Gothic *𐌷𐌰𐌻𐌾𐍉𐍂𐌿𐌽𐌰 (*haljōrūna, "witch")

This term is attested in Latinized form in Jordanes' Getica (c. 550) as haliurunnas (acc.pl), where it is described as a Gothic word for "witch" (magas mulieres). The word is clearly Germanic and has at least two cognates or near-cognates in OHG hellirūna f (“necromancy, sorcery”, a gloss of necromantia) and OE helrūna m (“demon”, Beowulf 163), and the constituent elements of the compound seem to be clear enough (𐌷𐌰𐌻𐌾𐌰 (halja, “netherworld, hell”) + 𐍂𐌿𐌽𐌰 (rūna, “secret, mystery”)).

What I don't understand is how there seemingly isn't any agent suffix; as shown above it's just a bare compound which one would expect (as it does in OHG) to mean something like necromancy or sorcery, not witch. The OE term isn't just hel + rūn either, it has some kind of suffix going on. Why would the reconstruction *𐌷𐌰𐌻𐌾𐍉𐍂𐌿𐌽𐌰 (*haljōrūna) then, as Köbler assures me, mean witch? Am I mistaken in thinking that it would need to have some kind of suffix attached in order to have that meaning? — Kleio (t · c) 20:07, 16 January 2017 (UTC)

In Old English, -a is a kind of agentive suffix, isn't it, as in dēma? But I guess that wouldn't be -a in Gothic too. Maybe the Gothic word is a bahuvrihi? —Aɴɢʀ (talk) 20:32, 16 January 2017 (UTC)

Yes, I figured the OE -a is probably an agent suffix but I wasn't entirely sure. I'd expect something like an ōn-stem as in 𐌰𐍂𐌱𐌾𐍉 (arbjō, “heiress”), personally, if it were to have a suffix. (I think -𐌾𐍉 (-jō) in that word may be the feminine equivalent of the masculine agent suffix -𐌾𐌰 (-ja), though I'm not 100% on that.) Your suggestion may explain it, although I can't recall other Gothic compounds that behave like that, with a second element in the compound that refers to an abstract concept now suddenly referring to a person when compounded. — Kleio (t · c) 20:46, 16 January 2017 (UTC)

As far as I can tell, the Gothic term is an ō-stem, the OHG term is an ōn-stem and the Old English term is an an-stem. --Florian Blaschke (talk) 21:47, 16 January 2017 (UTC)

That's not really the question here, though: it's the semantics of the Gothic word, which denotes a person (witch), which makes no sense to me. The problem is that the second element of the Gothic compound is just unmodified 𐍂𐌿𐌽𐌰 (rūna, “secret, mystery”); one would at least expect a suffix changing the end of the word to denote an agent. As it is now reconstructed, the word would be expected to mean something like "sorcery" just like its OHG counterpart, which it matches in terms of the etymological derivation of its constituent elements. — Kleio (t · c) 22:29, 16 January 2017 (UTC)

Since the word isn't attested except in a latinized form was altered to conform to Latin morphology, how do we know the form wasn't *𐌷𐌰𐌻𐌾𐍉𐍂𐌿𐌽𐌾𐍉 (*haljōrunjō) anyway? —Aɴɢʀ (talk) 23:35, 16 January 2017 (UTC)

Curiously, though, Old English has hellrūn, helrūn (“sorceress; woman with a spirit of divination”), which is a strong feminine o-stem. This matches the form and meaning of the postulated Gothic word and is likewise puzzling as one would expect it to mean "sorcery" or "necromancy" Leasnam (talk) 05:49, 17 January 2017 (UTC)

And similarly, helrȳneġu, which should mean something like "sorcery" also means "sorceress" instead. Strange. Leasnam (talk) 06:23, 17 January 2017 (UTC)

That's really interesting, I hadn't found those yet. That does seem to corroborate Angr's initial suggestion, and makes the Gothic reconstruction more likely to be correct, though it's still weird. @Angr: I personally would expect some remnant of the j to be reflected in a Latinization, something like *haliurun[n]ia or the like, if the Gothic term indeed had the suffix -jō. But it should be mentioned that the reconstruction I gave in my initial post is just what I found in Köbler and some very brief Google Books searches, perhaps others have different ideas. Definitely curious about this word now though, it's an interesting one and I'll look into it more when I get some spare time. I hadn't expected it to have so many apparent cognates. — Kleio (t · c) 15:32, 17 January 2017 (UTC)

@KIeio: actually, the nn of the Latin made me wonder if this word came from some variety of Gothic that had undergone the gemination of consonants before /j/ followed by deletion of /j/ otherwise known only from the West Germanic languages (e.g. *sunjō > synn). —Aɴɢʀ (talk) 16:11, 17 January 2017 (UTC)

@Angr, but would that not have also geminated the "l" in *haljō as well ( > *hallirunna, *hallurunna) as it does in Old English and Old High German ? Leasnam (talk) 17:52, 17 January 2017 (UTC)

@Leasnam: Good point. —Aɴɢʀ (talk) 19:21, 17 January 2017 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @Angr, Leasnam I've adjusted haliurunna's etymology section to reflect this discussion and some others I found after having a look around, notably an alternate view espoused by Scardigli (1973) and Lehmann (1986). Note also my addition of a reference to 𐍃𐍄𐌰𐌿𐌰 (staua), which is a word that can mean both "trial, judgment" and "judge" -- though the declensions differ. Wonder what y'all think. — Kleio (t · c) 12:11, 23 January 2017 (UTC)

The derivation rinnan (v.) -> runna (agent noun) doesn't seem irregular by the way; consider for example nuta ("catcher", masculine an-stem) from niutan ("to obtain, attain"). — Kleio (t · c) 13:56, 23 January 2017 (UTC)

Nicely explained ! Leasnam (talk) 03:44, 24 January 2017 (UTC)

奶昔

Similar to 紐約, I feel that perhaps this borrowing may have come through a southern dialect that retains the "k" sound in "shake". Any thoughts? ---> Tooironic (talk) 04:23, 18 January 2017 (UTC)

I agree. It's most likely through Cantonese. — justin(r)leung _{{ (t...) | c=› }} 05:06, 18 January 2017 (UTC)

Thanks Justin. ---> Tooironic (talk) 12:20, 23 January 2017 (UTC)

hauriō and *h₂ews-, and hauriant

hauriō is stated as coming from *h₂ews-, with the comment "spurious h", that is presumably meaning hauriō < *auriō < *h₂ews-; but the PIE root has no mention of this, and there doesn't seem to me to be much semantic resemblance either. Can it be confirmed or discredited?

This is a different *h₂ews-, a verb root reconstructed as meaning 'to scoop'; LIV gives reflexes from Greek (αὔω (aúō, “to carry (fire)”)), Old Norse (ausa (“to scoop”)) and Palaic. --Tropylium (talk) 13:27, 20 January 2017 (UTC)

While we're on this Latin word, can anyone justify it as the origin of the heraldic term hauriant, meaning (of fish) "rising vertically"? The OED s.v. haurient (which I've turned into an alternative form rather than the primary, as hauriant matches sejant, couchant etc., and is used by Fox-Davies), says it's because the fish is supposed to be raised above the surface to draw in air. How do they know? The heraldic term is likely to be proximately from French, not Latin (another reason for preferring -ant), so there could be an intermediate of some other meaning. The English term is first attested in 1572, as hariant, with no clue as to why it means that. The OED entry hasn't been revised since 1898. --Hiztegilari (talk) 18:30, 19 January 2017 (UTC)

Etymology for prefixed Modern Greek words

If we take a look at the word ανεργία (unemployment) the etymology says that it comes from the Ancient Greek word ἄνεργος. According to Wiktionary:Etymology#Affixation_and_compounds shouldn't the etymology show that it comes from αν- and έργο (i.e. without work)? If yes, there is also the word άνεργος (unemployed), should it have the same etymology as ανεργία or should one of the two words reference the other? That is, does άνεργος comes from ανεργία or the inverse? Thanks! Orgyn (talk) 13:00, 20 January 2017 (UTC)

I can't answer for Greek specifically, but in general the derivation would seem most important in the language in which it happened, in this case Ancient Greek. As far as I know, we do think it good to have it in the derivative language as well, even though this means more work in editing and may lead to contradictions... What's important (to myself, anyway) is to distinguish between forms that are from X and forms that are just analyzable as X. So, for example, Sunday isn't "from" sun + day, but only "analyzable as" such, because the compound existed already in Old English and probably even late Proto-Germanic (cf. *sunnōniz dagaz). Kolmiel (talk) 14:42, 20 January 2017 (UTC)

Indeed, we have lots of examples of this in English, e.g. unlike, which is diachronically from Old English unġelīċ and synchronically from un- + like. There's nothing wrong with listing both; indeed, it's preferable to do so in order to get the categorization right, since unlike ought to be in both CAT:English terms inherited from Old English and CAT:English words prefixed with un-. —Aɴɢʀ (talk) 16:02, 20 January 2017 (UTC)

That's exactly why I asked my question! (categorization) Putting both etymologies would solve this. What would be the word to use to "connect" both etymologies? unlike, atrophy and astrology all show different ways of doing it. Concerning my second question, how is etymology handled between derived terms? Orgyn (talk) 19:15, 20 January 2017 (UTC)

What I've often seen and used myself is "From Old English unġelīċ. Equivalent to un- +‎ like." --WikiTiki89 19:22, 20 January 2017 (UTC)

The wording I usually use is "synchronically analyzable as un- +‎ like" or just "synchronically un- +‎ like". I don't think every single etymology like this has to have the exact same wording. —Aɴɢʀ (talk) 19:24, 20 January 2017 (UTC)

I use the "Equivalent to" wording. I dislike "synchronically", because that's not understood by the average user. Moreover, I don't think we should be showing synchronic etymologies to begin with. Etymology is diachronic by definition. I would only say "Equivalent to un- + like" if the word was formed from the ancestor of un- and the ancestor of like, not if the word is currently analysable as such but was not actually formed that way. In my view, the idea of giving equivalent etymology is that one is directed to the entries of these parts to look up their individual etymologies, so that they are not repeated. —CodeCat 19:30, 20 January 2017 (UTC)

If a user doesn't know what a word means, he can look it up. We are a dictionary, after all. And I do think we should be including synchronic etymologies at least for transparent root + affix words, especially when the affix is still productive, because when a speaker uses a word like goodness, for example, we have no way of knowing if he's learned that word as a whole, the same way he learned good (in which case the etymology is only diachronic), or whether he's combining good and -ness himself "on the fly" (in which case the etymology is synchronic, as the speaker has coined the word afresh). So while goodness does go back to gōdnes and very probably *gōdanassuz, it is also being constantly re-coined in Modern English as good + -ness. It isn't simply equivalent to good + -ness, it is good + -ness. —Aɴɢʀ (talk) 22:11, 20 January 2017 (UTC)

Thanks guys for your answers! Orgyn (talk) 09:21, 21 January 2017 (UTC)

I agree with CodeCat, that not only would "synchronically" be less likely to be understood, but it also is not the right concept. --WikiTiki89 14:43, 23 January 2017 (UTC)

Then what is? "Equivalent to" isn't the right concept either. —Aɴɢʀ (talk) 14:59, 23 January 2017 (UTC)

Why isn't it? --WikiTiki89 15:12, 23 January 2017 (UTC)

Because unlike isn't merely "equivalent to" un- + like; it *is* un- + like. —Aɴɢʀ (talk) 16:02, 23 January 2017 (UTC)

But it is Old English unlīc, ungelīc and/or PGmc *ungalīkaz which are described as "equivalent to un- + like", not Modern English unlike. Leasnam (talk) 03:48, 26 January 2017 (UTC)

What it's supposed to mean is that it is equivalent to a derivation from un- + like, even though it is really inherited as a whole. --WikiTiki89 16:30, 23 January 2017 (UTC)

I know what it's supposed to mean, but it isn't merely equivalent to a derivation from un- + like. It is simultaneously *both* an inheritance from Old English *and* a concatenation of two Modern English morphemes. —Aɴɢʀ (talk) 16:38, 23 January 2017 (UTC)

Well in this particular case you're sort of right, because the morphemes were kept separate, otherwise we might have ended up with unylike or something, but in other cases there is no reason to say that. --WikiTiki89 18:52, 23 January 2017 (UTC)

Can you think of an example where there is no reason to say that? —Aɴɢʀ (talk) 09:25, 26 January 2017 (UTC)

I've seen "surface analysis" used, that seems to work. — Kleio (t · c) 16:25, 23 January 2017 (UTC)

Yes, that would work too. —Aɴɢʀ (talk) 16:38, 23 January 2017 (UTC)

I disagree. The surface analysis can contradict the true etymology. As I said, etymology is by definition diachronic, so merely analysing the contemporary surface form is not good enough. Apparent origins can be misleading. —CodeCat 17:14, 23 January 2017 (UTC)

We shouldn't have only the synchronic analysis any more than we should have only the diachronic one. But saying "etymology is by definition diachronic" is flatly untrue, since words are coined simultaneously with their usage all the time (cf. wug test). A word like "goodness" is not only inherited from Proto-Germanic/Old English/Middle English; it is also built by contemporary English speakers by combining "good" and "-ness". —Aɴɢʀ (talk) 09:25, 26 January 2017 (UTC)

But if that's the case, we should see serious creations such as goodity, goodation, goodment, which we do not Leasnam (talk) 15:18, 26 January 2017 (UTC)

We do, however, see Category:English back-formations (for example). People do analyze the apparent components of words all the time, and sometimes create new words based on those analyses. — Kleio (t · c) 21:50, 26 January 2017 (UTC)

PIE surface vs. morphophonology

I noticed that we seem to have around "PIE roots" used for one descendant only, e.g. *Hyekʷ-. I was about to protest this as silly, but then realized that this probably has benefits for our root shape categories.

But this also got me thinking that this may have other benefits yet as well. A common complaint fielded at PIE reconstructions is that they often inconsistently mix surface (lexical) and deep (morpho-) phonology. However, if we maintain roots as a fully different category from words, then we will have the option to actually claim the best of both worlds: we could transcribe roots morphophonologically, actual word-forms surface-phonologically.

We already have e.g.:

• w:Stang's law: *dyēm and not *／dyéwm
• w:Szemerényi's law: *ph₂tḗr and not *／ph₂térs
• Unrounding: *h₁léngʰus and not *／h₁léngʰʷus
• *mn-deletion: *gʷʰormos and not *／gʷʰormnos (cf. e.g. formus)
• *ḱwṓ and not *／ḱwṓn
• *ḱḗr and not *／ḱérd

I would suggest indicating also at least the following:

• Laryngeal coloring: *h₂ánti rather than *h₂énti, *ǵʰáns rather than *ǵʰh₂éns, perhaps *h₃óngʷn̥ rather than *h₃éngʷn̥ (though *h₃e and *o do have partly different reflexes, e.g. in Indo-Iranian with respect to Brugmann's law; we could use notation such as *eₒ or *e₃ for these cases)
• w:Siebs' law (*sBʰ → *sPʰ), though we don't seem to have any lemma examples so far.
• Devoicing: *yuktós rather than *yugtós, perhaps also: *nizdós rather than *nisdós?
• s-insertion: *bʰudʰstós rather than *bʰudʰtós, *wéytstōr rather than *wéydtōr, perhaps also: *tsḱéyti rather than *tḱéyti?

Various newer laryngeal-deletion and schwa-insertion rules would likely be just as good to indicate, though these would take a bit more time to dig up. --Tropylium (talk) 19:38, 20 January 2017 (UTC)

Beside the argument that we're simply writing things the way linguists do, there's also these points:

• Szemerényi's law was morphologised and no longer purely phonological. We find analogical long vowels in words such as *dyḗws, *pṓds. There is also no trace of it in e.g. proterokinetic genitive singulars or athematic 2nd person singulars, where it might have originally occurred.
• The same probably applies for Stang's law.
• Unrounding is indicated because it results in phonemic change: the non-labial consonant is a phoneme.
• Siebs' law, by contrast, does not change phonemes, it is allophonic.
• Devoicing is post-PIE, and absent in IIr. Voicing assimilation of *s doesn't create a new phoneme, so again it's not indicated.
• s-insertion may also be post-PIE, if IIr is an indication. Either way, it's allophonic.
• ōn > ō occurs exclusively in Szemeryényi's law environments, so the same reasoning applies.
• *ḱḗr is a unique example with no parallels.

—CodeCat 20:02, 20 January 2017 (UTC)

(e/c) We already do:

• w:Pinault's law: *h₂éryeti and not *／*h₂érh₃yeti,
• I tend to do schwa-insertion as seen generally through out *tep-, though I'm still on the fence about it.

As for the others you've suggested:

• As to voicing assimilation, Ringe points out that since these assimilations are so common, it's hard to know whether they occurred in PIE or in the daughters. We can't even tell if w:Bartholomae's law was in PIE.
• s-insertion is fine with theoretically, but, if we also accept assimilation, we start getting into the hairy business of *bʰudʰtós > *bʰudʰstós > *bʰudʰzdós? > *bʰudzdʰós??. This is a bit too far for my taste.
• I am deeply opposed to showing vowel coloration, as it muddies the morphology considerably. I'm aware that Stang's and Szemerényi's laws are not obvious either, but being able to distinguish between *ō, *oh₃, and *eh₃ is crucial for our reconstructions.

I would also add that we do not use:

• w:Sievers' law: We use *h₂ḱh₂owsyéti instead of *h₂ḱh₂owsiéti
• Lindeman's law?: *dyḗws instead of *diḗws
• w:Thorn cluster: *tḱéyti instead of *ḱtéyti or *ḱstéyti?

Overall, I'd say we should cleave to the linguistic standard, which we currently do, though you will always find counterexamples (Fortson does s-insertion, I believe). —JohnC5 20:25, 20 January 2017 (UTC)

Allophones can probably get hairy, yes. I'm mostly interested in increasing the non-PIE-initiated accessibility of the reconstructions (= decreasing the amount of "gratuitous" features in the reconstruction), but leaving that level of transcription off is reasonable in general.

We aren't able to distinguish *eh₃(C), *oh₃(C) and even other *oH(C), though: they are reflected the same everywhere. I would not go as far as writing out the compensatory lengthening though, as it is often enough taken as post-PIE (e.g. by some formulations of Dybo's law and Balto-Slavic accentology). But there is still the intermediate option to use *ah₂ and *oh₃. Note also that, as rare as it may have otherwise been, we are already using *a anyway.

As for s-insertion: *s is a phoneme, so the change clearly isn't allophonic. AFAIK Indic *TT is normally considered a back-mutation, since Iranian shows *sT and since Sanskrit has also the sandhi rule *-s > ∅ / T_#, which would be expected to turn *-TsT- back to *-TT- in any case. --Tropylium (talk) 21:30, 20 January 2017 (UTC)

Old Norse greiða

Would someone have more details on this word? Anything to do with *raidaz? --Barytonesis (talk) 17:57, 23 January 2017 (UTC)

Yes. greiða is from Pgmc *garaidijaną, from *garaidaz, a prefixed form of *raidaz (“ready”) Leasnam (talk) 03:55, 24 January 2017 (UTC)

Thanks! --Barytonesis (talk) 18:31, 26 January 2017 (UTC)

gôrre

Does this word actually descend from Late Latin gonorrhea, like Sicilian camurria does? Or is it a clipping of a borrowed word? Is it even descended from "gonorrhea" at all? Cirxe (talk) 03:52, 25 January 2017 (UTC)

On "bonkers"

I have found a good source of etymology here, but am unsure of standards on formatting here. If someone would be so kind as to demonstrate it to me and/or post it, that would be lovely. — JWhitt433 (talk) 06:38, 30 January 2017 (UTC)

bod

I am dazzled by how long this verb has gone without explaining its suppletion. I was thinking, "what's up with wyf, rwyt, oeddwn, etc."? So, Geiriadur Prifysgol Cymru claims that:

"[daw’r rhan fwyaf o’r ff. pres. ac amhff. myn. o’r gwr. IE. *es- ‘bod’ (e.e. ys (H. Gym. is), H. Lyd. is, H. Wydd. is: < IE. *és-ti"

"ansicr yw trdd. y mae, Crn. C. (y(m))ma, Llyd. C. ema; ynglŷn â’r be. bod, gw. bod1] "

"[?< *maʒe(h)es < *mages-est, sef cfl. lleol yr e. *magos (> ma1, -fa)+*est, 3 un. pres. myn. y f. ‘bod’; dtb. o’r defnydd hwn yw mae2, y mae, cf. Crn. C. ma, ym(m)a, Llyd. C. ema, Llyd. Diw. emañ] "

Anti-Gamz Dust (There's Hillcrest!) 19:00, 30 January 2017 (UTC)

travel

The etymology currently given lists English < Middle English < Middle Scots < Old French. Isn't (Middle) Scots a descendant of Middle English? KarikaSlayer (talk) 18:29, 31 January 2017 (UTC)

Middle Scots seems to start by 1400 and Middle English seems to end by 1500, so technically it might be possible. I wonder how useful the term "Middle Scots" is under these circumstances. Also since wikipedia says that earlier Middle Scots is virtually indistinguishable from northern Middle English. Kolmiel (talk) 07:00, 1 February 2017 (UTC)

Does a Middle Scots attestation (either for the word in general or the "altered" form with -l-) being earlier than the Middle English one really warrant loaning from there? You'd expect the ME word to have been around about just as long, maybe simply going unwritten for a bit longer. --Tropylium (talk) 22:37, 1 February 2017 (UTC)

Of course, it doesn't necessarily warrant that. But maybe there's some circumstantial evidence for a borrowing from the north. Kolmiel (talk) 09:51, 2 February 2017 (UTC)

RFV LA:nihilominus

From nihil +‎ ominusus or nihil +‎ omen?. Seems to me pretty straightforward in the form, but the meaning? Sobreira ►〓 (parlez) 11:43, 1 February 2017 (UTC)

Pretty sure it's nihilō +‎ minus. —Aɴɢʀ (talk) 12:00, 1 February 2017 (UTC)

Shame to me... :( Sobreira ►〓 (parlez) 17:16, 15 February 2017 (UTC)