Wiktionary:Etymology scriptorium/2019/April

This etymology (which claims it's a derivation of 隠 (on)) faces so many phonological and morphological problems. What kind of suffix even is this -i or -ni or whatever? Also, an -i suffixing onto a moraic nasal to form a ni syllable in this word does not seem historically likely, as we have things like 恋愛 (ren'ai) instead of re-na-i. Ryukyuan cognates of this word are also very widespread, e.g. we have the Miyako cognate un (note the loss of -i in Miyako vs. Kunigami unii).

I highly suspect that this is in fact a native word. mellohi! (僕の乖離) 01:44, 1 April 2019 (UTC)

FWIW, a number of these works note this theory (ctrl+f 隠). —Suzukaze-c◇◇ 17:17, 5 April 2019 (UTC)

• (after edit conflict) @mellohi! -- Shogakukan's KDJ explicitly lists this derivation:

（「隠（おん）」が変化したもので、隠れて人の目に見えないものの意という）

So does Daijirin:

〔姿が見えない意の「隠」の字音「おん」の転という〕

As does Gogen Allguide:

「おに」と読まれるようになったのは平安時代以降で、『和名抄』には、姿の見えないものを意味する漢語「隠（おん）」が転じて、「おに」と読まれるようになったとある。

Sources that mention timing list this as arising during the Heian period, when other such terms also arise from Chinese borrowings that develop an /-i/ on the end as a kind of phonetic excrescence. Some examples:

• 文 (fumi, “writing”), ancient /pumi/, apparently from earlier borrowed form /puɴ/
• 銭 (zeni, “money”), a shift from /seɴ/ (looks like we're currently missing this etym / sense / etc. at 銭)
• 縁 (eni, “luck, fate”), a shift from /eɴ/ (also missing this currently)
• 訓 (kuni, “meaning; native Japanese reading”), a shift from /kuɴ/ (also missing this currently)

So this kind of excrescence is not unknown elsewhere in the lexicon. One key is that final nasal /ɴ/ had no corresponding kana for some time, and thus it was previously written with the kana for む (mu) (and, indeed, the modern kana ん (n) arose from a hentaigana form of this same む). The excrescence may have developed partly due to the lack of an unambiguous spelling. The spelling for nasal /ɴ/ apparently stabilized around the 1100s; see the JA WP article section at w:ja:ん#表記 for more on that topic. ‑‑ Eiríkr Útlendi │^{Tala við mig} 17:29, 5 April 2019 (UTC)

The entry says that it's from German Stab, which is entirely plausible. But there's the dialectal variant šćap, which is said to be from Proto-Slavic *stjьapь (which looks a bit strange to me, but then I know very little about Proto-Slavic). So is one of these etymologies wrong? Are they really totally unrelated? In that case, they shouldn't be called "alternative forms". Or is štap an alteration of šćap under German influence? Thank you. 178.4.151.28 22:13, 1 April 2019 (UTC)

When the lemma šćap was created by an anonymous IP on March 2, 2013, the etymology was given as “From German (deprecated template usage) Stab”. On March 20, 2013, however, an IP in the same range changed the etymon to “stjьapь”.  --Lambiam 08:29, 2 April 2019 (UTC)

The latter IP made the same change to štap, but there the change was quickly reverted.  --Lambiam 12:53, 2 April 2019 (UTC)

Removed this Proto-Slavic invention and made šćap alternative form of štap, which it is. This does not exist in remoter Slavic languages to reconstruct anything and both is the same obvious borrowing from German. Fay Freak (talk) 20:47, 2 April 2019 (UTC)

Etymology at Tigger:

From the name of the tiger friend of Winnie the Pooh, an alteration of tiger, which is a loanword derived from Latin tigris, itself borrowed from Ancient Greek τῐ́γρῐς (tígris), which is derived from Ancient Greek Τῐ́γρης (Tígrēs), Τῐ́γρῐς (Tígris, “the river Tigris”), ultimately from Sumerian 𒀀𒇉𒈦𒄘𒃼 (^ÍDIdigna, ^ÍDIdigina, “the river Tigris”, literally “fast as an arrow”).

This seems way overkill. I would stop at "an alteration of tiger". However, there are parts of the etymology of "tiger" that are not actually present at tiger, and I do not have the knowledge to merge them with any confidence. If anyone with the requisite knowledge has the time and inclination, perhaps they could merge the etymology content to tiger and delete it from Tigger. Mihia (talk) 22:04, 4 April 2019 (UTC)

I removed it before even seeing your post, but I am now concerned about our entries which are consistent with this Sumerian etymology. —Μετάknowledge^{discuss/deeds} 01:49, 5 April 2019 (UTC)

OK, thanks. Sadly I cannot help with Sumerian. Mihia (talk) 17:08, 5 April 2019 (UTC)

I would like it if there were better guidelines about when to remove duplicated information. I tend to remove excessive etymological information, such as cognates when they are already shown as descendants on another entry, but there will always be some people who like all the information to be on all the pages. Having a proper guideline to fall back to would be good. —Rua (mew) 17:33, 5 April 2019 (UTC)

Repeating the etymology of X separately for each word derived from X seems to me like a maintenance headache and a recipe for the information to get out of sync -- unless, of course, it can be maintained in one place and then transcluded. Also, I think that for "light" derivations, such as "Tigger" from "tiger", it looks a bit silly to trace back to some ancient language. Mihia (talk) 19:22, 5 April 2019 (UTC)

According to the etymology of *sočiti, this is a causative-iterative from PIE *-éyeti. However, that would mean that the Balto-Slavic sequence *-eitei (from the infinitive) would be reflected in Lithuanian as -yti and Latvian -īt. The Balto-Slavic diphthong ei, however, normally appears in these languages as ei or ie (sometimes word-finally as i). Lithianian y only reflects PIE *iH as far as I know. So how is the unusual Baltic reflex of this verb formation explained? —Rua (mew) 19:23, 5 April 2019 (UTC)

I remember wondering about the same thing a while ago (but don't remember which entry that was.)

FWIW, Derksen gives both *sok-ei/i - normalizing the transcription (e.g., he doesn't indicate long vowels in non-initial syllables) and spelling them out that's Proto-Balto-Slavic *sakeitei ~ *sakītei

1. There is a pretty common trend among authors to try to "mash together" homonymic PIE roots with really unsatisfying semantic leaps (personally I dislike this greatly and I don't understand what's the benefit) he mentions this in the etymology for Lithuanian sekti II "to tell", now that's a matter of taste if you want to view PIE "to follow" and "to say" as the same root (I don't.)
2. I think it's pretty uncontroversial that Baltic and Slavic chose different means to make their iteratives? These movement verbs which are otherwise strikingly similar tend to have iteratives derived different ways: nesti : nositi, vesti : voziti vs nest : nēsāt, vest : vadāt and so on. Also nothing in Baltic is derivable from a PBS *-eitei (as you already mentioned), so a Slavic-only iterative making device? (Side note: at least Latvian sacīt (“to say”) has a quite perfective feel (i.e., the opposite of iterative) pretty sure that will also be true for Lithuanian sakyti)
3. When you look at the meanings of many of the Slavic *sočiti verbs it becomes quite clear they derive from PIE "to follow". The semantic shift "to follow/seek" > "take legal action" is very satisfactory in my opinion and has many nice parallels, cf., Russian иск : искать (isk : iskatʹ, “lawsuit : to search”), English pro-sec-utor, etc.

To sum up: *sočiti and sakyti are false friends (unless you mash together PIE "follow" and "say") and even in that case I don't see a reason why one couldn't be derived from *sakeitei and the other from *sakītei (thus not breaking any basic sound rules.) Neitrāls vārds (talk) 05:24, 9 May 2019 (UTC)

The etymology on this entry says it descends from Proto-Bantu *m̀bɪ̀já (“pot”). This is plausible, but it has some problems. First I should note that Bantu Lexical Reconstructions 3 includes the reconstruction *-bɪ̀gá "pot" (class 5/6, 7/8, or 9/10). It includes a variation *-bɪ̀já. However, whether the Zulu term descends from *-bɪ̀gá or *-bɪ̀já, the expected reflex would be -tshá IPA^(key): /tʃá/ or -já IPA^(key): /dʒá/ (if class 9). This is because both *g and *j delete in that environment. The resulting vowel sequence *ɪa causes palatalization. Guthrie states that the Zulu reflex of *-bɪ̀já is isí-tsha. The segments correspond, but the tone should be high on the root. ím-bizá has the right tone, but it cannot be the reflex unless the PB root had the form *-bɪ̀gía or *-bɪ̀jía.

Note also, the Tswana cognate pitsa (“pot”), in which the first vowel is a reflex of *i, and the consonant /ts/ seems to be a reflex of *gi or *ji. Perhaps a preceding *i could have the same effect, leading to a form *-biga/*-bija (ignoring tone) for both languages. However, I'm not certain if a root internal *i would result in the same consonant correspondence in Tswana. Further, I would expect spirantization in Zulu before that high vowel, so that *bi would result in /vi/.

With all these issues, isitsha seems to me to be the less problematic reflex (note also that according to Guthrie, the Tswana word se-ja-na is the reflex of *-bɪ̀gá). The improper tonal correspondence seems rather minor. However, if isitsha is the actual reflex, then the question is, where does imbiza come from?

Smashhoof (talk) 20:41, 5 April 2019 (UTC)

Russian лях (ljax), Polish Lach, Turkish Leh (and Ottoman Turkish), Armenian լեհ (leh) - a bit of circular etymology. There are also Hungarian(?), Persian and Tajik cognates. Vasmer says "лях" is from Old East Slavic/Old Russian. The singular must be ляхъ (ljaxŭ). @Vahagn Petrosyan might be interested. --Anatoli T. ^{(обсудить}/^вклад) 22:49, 5 April 2019 (UTC)

I don’t see a cycle. The Russian entry lacks an etymology, but the Russian Wiktionary says it’s from Polish Lach. Adding that, we get the linear descent string Armenian < Middle Armenian < (Ottoman) Turkish < Russian < Polish < Old East Slavic < Proto-Slavic.  --Lambiam 07:32, 6 April 2019 (UTC)

Perhaps you're right about the linear descent. How should the Russian entry use the reconstructed Old East Slavic *ляхъ (*ljaxŭ) when it's directly inheriting from it, which template? --Anatoli T. ^{(обсудить}/^вклад) 09:39, 6 April 2019 (UTC)

I believe the circularity is now gone. --Vahag (talk) 10:33, 6 April 2019 (UTC)

@Vahagn Petrosyan: Good job, Vahagn, thank you! --Anatoli T. ^{(обсудить}/^вклад) 10:39, 6 April 2019 (UTC)

The Old Church Slavonic has also the name Лѧхꙑ pl (Lęxy) for Poland. --Anatoli T. ^{(обсудить}/^вклад) 11:14, 6 April 2019 (UTC)

I assume it is used in the Russian recension (izvod) of Church Slavonic. Then it must be a borrowing from Old East Slavic. --Vahag (talk) 11:37, 6 April 2019 (UTC)

Means "fox" in Vietnamese. Is there a reconstructed Proto-Vietic word for the animal?

This page shows that the red fox might be found in Vietnam but its presence there has not been confirmed. (Don't be too harsh on me: I have previously been criticised here for relying solely on the present-day distribution of animal species to question the etymology of certain species names.)

Point of note: The Vietnamese Wikipedia article for the leopard cat says that the Muong name for the animal is "cáo khua". Maybe give me some hint as to the etymology of the word? --Corsicanwarrah (talk) 16:23, 7 April 2019 (UTC)

I have never come across the word for "fox" in any paper or recording on any minor Vietic language. The only cognate I have found is Mường cảo (“fox”) as recorded in Nguyễn Văn Khang's Từ điển Việt-Mường (the tone might vary, depend on the lect). But since Vietnamese and all Muong lects share the vast majority of vocabulary, and I can't eliminate the possibility of borrowing, this cognate is probably not of significance. PhanAnh123 (talk) 16:32, 9 April 2019 (UTC)

Foxes feature in some Vietnamese folktales and proverbs, and there is a Vietnamese water-puppet story featuring an animal which is recognisably a fox, so I'd put my chances on foxes being a traditional presence in Vietnamese culture and that the word for 'fox' is native. Trouble is, what did it derive from? --Corsicanwarrah (talk) 09:32, 20 April 2019 (UTC)

RFV of the etymology, specifically "calque from English". Can't it also be a calque from Chinese 廣東話? — justin(r)leung _{{ (t...) | c=› }} 01:27, 9 April 2019 (UTC)

I don't think the English etymology is valid and could be replaced with either Chinese or native. 語(ご) (go) can be added to any place or country name. It would be a stretch even for カントン語(ご) (kantongo) (the English etymology). --Anatoli T. ^{(обсудить}/^вклад) 01:35, 9 April 2019 (UTC)

@Fumiko Take. —Suzukaze-c◇◇ 01:43, 9 April 2019 (UTC)

The English sense "(South Africa, Australia) A reservoir." (the reservoir or body of water created by a dam is a separate sense) and the Afrikaans sense "pond, basin" correspond quite well in meaning. Did one of them pass it to the other as a semantic loan? The oldest mention I found for Afrikaans is from 1844. ←₰-→ Lingo ^Bingo _Dingo (talk) 13:16, 9 April 2019 (UTC)

This root has a lot of descendants, just looking at lists of cognates of each entry that lists it as a root. However, there's not a page for it. Is there a reason for this, and, if not, should I go ahead and create a page for the entry? GabeMoore (talk) 15:48, 10 April 2019 (UTC)

I've created it now. —Rua (mew) 17:57, 14 April 2019 (UTC)

“guta” = vowel < Old Irish “gutte” = vowel

Not a particularly pressing matter, I suppose, but there is no indication here whatsoever in the etymology section as to why exactly the relatively uncommon Old English /sk/ in āscian, which had become /ʃ/ in this word for many people in the Middle English period, is retained as /sk/ in Modern English, yet we note why seek has a /k/ in it now in its etymology section. What is good for the goose is good for the gander.

I know that asken, askien, and aske are all attested in Middle English, but I question how noteworthy that is in this case, considering that one of the Middle English forms of Ash Wednesday was "Askwednesday", even though Old English æsċe already used /ʃ/ in this word. I suppose that I would guess that those forms represent Norse influence or something of that sort, but I'm just guessing there.

Is it possible that the /sk/ was retained in part due to the commonness of the Middle English etymon of the (now nonstandard) alternative form to ask- aks, seen as axen and the like, in its time? I'm wondering if a combination of metathesis of axen (which we know is of metathetical origin in and of itself, back in Old English) in some speakers, the true natural retention of the Old English consonant cluster in this word, and an undoing of a shift to /ʃ/ seen in some versions of Middle English asken due, perhaps, to Norse influence (although I was under the impression that the axen form was particularly common in Scotland and the like, and although Norse influence hit a number of areas in Britain even at that point in English's development, we tend to see contemporary /sk/, /k/, /g/ in places where Proto-Germanic /sk/, /k/, and /ɣ(.)j/ had developed into /ʃ/, /tʃ/ and /dd͡ʒ/ mostly in dialects of Scots, so perhaps Norse influence is not a factor in this particular case) led to the ultimate retention of /sk/ in this word.

Does anyone have any insight on this? Tharthan (talk) 20:13, 12 April 2019 (UTC)

The Online Etymology Dictionary states that the form in English was influenced by a Scandinavian cognate, such as Danish æske.  --Lambiam 23:55, 12 April 2019 (UTC)

IIRC, /sk/ became /ʃ/ only in the palatalizing positions, that is, before front vowels. Like other -ian verbs this verb originally ended in *-ōną (extended with a *-ja- with later contraction in Old English), which meant that it didn't partake in palatalization. Crom daba (talk) 12:18, 13 April 2019 (UTC)

Old English sc was palatised in all positions (e.g. *skūrō > sċūr > schour (“shower”); *skōhaz > sċōh > scho (“shoe”); *skamō > sċamu > schame (“shame”); etc.). I don't think ask is really that unique as far as the sk is concerned...indeed many such parallels can be observed: mask (“mesh netting”) from max, dusk from dox, or ask (“newt”) from āþexe, where Old English x metathesises to Middle English sk then to English sk. Middle English was an environment where multiple variations were eventually culled down to a few, then one..."ask" just happened to be the one that won out among all the others. And as stated above, the Danish word may have been a contributing factor. Leasnam (talk) 23:57, 13 April 2019 (UTC)

Interesting. So you think that any cases that underwent metathesis were "restored" with the original sk later on? —Rua (mew) 17:47, 14 April 2019 (UTC)

Well, yes...English dusk is a perfect example of this: it metathesised once from PGmc sk > OE ks; then again later on from OE ks > ME sk. As far as ask is concerned, and I may be wrong, but I've always considered ask to be a descendant of āxian (i.e, āxian > ǎksien > askien > asken) rather than from āsċian. Leasnam (talk) 04:37, 15 April 2019 (UTC)

In fact, ask has metathesised more times than dusk, because in modern AAVE it's gone from sk back to ks. So to count them: PGmc sk to OE ks (=1); OE ks to ME sk (=2); then ModEng sk to ks...=3 times ! Leasnam (talk) 04:44, 15 April 2019 (UTC)

Doesn't the AAVE form continue a dialectal ME and EME form that has remained in use in various English dialects? As far as I know there is no general soundlawful AAVE metathesis of EME /sk/.

In any case, I note that the examples given above of OE sc being palatalized "in all positions" are all word-initial. It seems however that word-medially sc + back vowel has survived. Besides ask there are also e.g. dusk, tusk, mask (“mesh”). This is also the distribution given by Ringe & Taylor in The Development of Old English (2014). --Tropylium (talk) 10:07, 15 April 2019 (UTC)

Interesting. Dusk and mesh were also given by Leasnam as examples above, and dusk may be from (quoth the Middle English Dictionary) "OE dox, with metathesis; cp. OS dosan, OHG tusin. ME forms also point to OE variants *deohs (later *deocs) & *duhs", with the verb dusk from OE doxian (again per the MED), although Etymonline suggests this derivation is less certain, and Bosworth/Toller mentions both "dohx" and also "dosc(?)" as if they are alt forms, the latter of which would suggest an -sk form might've simply survived and won out. Your tusk and mesh do seem like good counter-examples, as far as metathesis goes, if they are as our entries say from tūsc and masc. - -sche (discuss) 12:53, 15 April 2019 (UTC)

OE "tusk" was also tux, in fact it was tux 14 out of 16 times it's attested, tusc is only attested twice; so like ask and dusk above, the sk in ME is potentially due to metathesis and not a survival of an OE word where sc=sk. Leasnam (talk) 14:05, 15 April 2019 (UTC)

Frequency seems irrelevant. If tusc is regardless attested, surely we ought to KISS and assume that Modern or Middle English reflexes continue directly any corresponding OE variants.

Really I get the impression that this is the only time that an actual soundlawful metathesis of these clusters happened anywhere in English, and every case of what is called "metathesis" later on is just interdialectal diffusion of the already extant variants (note that Middle English does not "re-metathesize" /ks/ to /sk/ in cases like mix, six, axe, flax, wax). We don't call it a new separate sound change either if American English e.g. standardizes ass rather than arse, it's just borrowing from some already nonrhotic variety into the rhotic ones. --Tropylium (talk) 16:31, 16 April 2019 (UTC)

For what it's worth (which is, admittedly, very little if anything), I would have to agree with Tropylium on the idea that "aks" is a descendant of āxian, and not a coincidental recurrence of the /sk/ → /ks/ metathesis in that word. I say this because, indeed, it seems to have entered African American Vernacular English through a dialect of Older Southern United States English (I have heard that one can still hear a couple or a few dialects in the South today that still have this pronunciation). This pronunciation has also historically been present in some dialects of British English as well, and since (as we know) there are descendants of āxian in Middle English, I don't find it unreasonable to believe that aks ultimately descends from āxian.

Regarding /sk/ surviving in any meaningful sense naturally in Old English (except maybe in some small, unrecorded dialect. But such a suggestion could be made for far too many things as an excuse to not say definitively that something happened or didn't happen), I highly doubt it, and I agree with Leasnam that words like dusk and tusk are probably due to undone metathesis (or metathesis of a metathesised form, if you prefer). mask (meaning mesh) is seemingly from Old English max, and furthermore there could have potentially been some influence as well from descendants of Middle French masque, since it seems that "net mask" may have been a potential use of the Old English word. Frankly, as far as I can tell, the only real potential example of a word that one could use to make an argument for a rare Old English survival of /sk/, is masker, ultimately from the Old English root word of Old English malscrung, but that could perhaps also just be due to the Norse influence that impacted other words being extended by analogy to this one. Tharthan (talk) 03:25, 16 April 2019 (UTC)

• قرمز

The etymology misses Middle Persian (Pahlavi) karmir. The word looks like a Middle Persian loanword rather than from Persian kermest. So it should be from Persian, from Middle Persian, from Iranian, from Proto-Indo-Iranian. --188.158.100.4 13:31, 14 April 2019 (UTC)

Another case of inconsistent etymologies.

The etymology at try is:

From Middle English trien (“to try a legal case”), from Anglo-Norman trier (“to try a case”), Old French trier (“to choose, pick out or separate from others, sift, cull”), of uncertain origin. Believed to be a metathetic variation of Old French tirer (“to pull out, snatch”), from Gothic 𐍄𐌹𐍂𐌰𐌽 (tiran, “to tear away, remove”), from Proto-Germanic *teraną (“to tear, tear apart”), from Proto-Indo-European *der- (“to tear, tear apart”), see tear. Related to Occitan triar (“to pick out, choose from among others”).

...Yet the etymology given for Modern French trier is:

From Middle French trier, from Old French trier (“to choose, pick out or separate from others, sift, cull”), from Gallo-Romance *triare (“to pick out”), a variant of Late Latin trītō, trītāre, from Latin trītus, the past participle of terō. The word sense originates from granum terere, to beat the corn from the chaff, or trier le grain in modern French, hence the meaning. Italian tritare keeps both senses of the word - to grind and to sort - confirming a common Romance origin. For loss of medial "t" see abbaye.

Old French tirer (“to pull out, snatch”), is a false cognate of Germanic origin.

Related to Occitan triar (“to pick out, choose from among others”), Catalan triar (“to pick, choose”).

...

I don't see much support for the Germanic origin that we allege is the "believed" etymology after a swift look-up on the Web, but I do see etymologies that support all or part (sometimes different parts of the etymology given here for Modern French trier) of the etymology given here for Modern French trier. Tharthan (talk) 22:06, 14 April 2019 (UTC)

French wiktionary states that it is disputed, and lists both ] Leasnam (talk) 04:27, 15 April 2019 (UTC)

(I was originally going to bring this to the Tea Room, but as I am asking somewhat more about etymology here, I decided to bring it here instead. It's one of those grey line cases, I suppose.)

The definition here almost makes it seem as if the people using this term meant to say übercool, but either a. made a typographical error, or b. confused the ur- prefix with the more common über.

Because, I mean... how does one say "very cool" with a word that literally means "proto-cool"? Is this some jocular coinage based upon the same idea as that wretched "original gangsta" coinage (I refer to the usage of that term as something akin to a retronym for the idea of something that is "old-school")?

"Dude, that's not cool- That's the ancestor of cool!"

Something like that? *chuckles*

Or perhaps:

"You think that's cool? Listen, that was the predecessor of cool, man. Get it straight!"

It sounds goofy, but that has never stopped anyone before.

Any suggestions? Does the textual/literary record provide any clues? The third example that we give, from The New York Times in 2014, seems to me to be the only real (potentially) cognisant usage of the term. In that particular instance, it seems as if it could have meant "ancient cool" or something like that. Still odd, but plausible at least. Tharthan (talk) 03:19, 18 April 2019 (UTC)

If this is an issue of confusion, then the Germans are confused too: see urcool. Now you may think they loaned this from English, but the adjective uralt is itself what it says; it is already attested in Old High German. So in German ur- can be used as an intensifier. The prefix über-, on the other hand, is not normally used that way. The adjective übergroß, for example, does not simply mean “very large” but rather “too large” – which is decidedly uncool.  --Lambiam 05:50, 18 April 2019 (UTC)

Are we then missing a sense at English ur-, meaning "extremely/very" ? Leasnam (talk) 14:54, 18 April 2019 (UTC)

If it can be shown to be productive in the intensifier sense, we should include it. But perhaps ur-cool is ur-alone in this sense – it might even be a borrowing from German.  --Lambiam 19:21, 18 April 2019 (UTC)

I see. I will note, though, that uralt could be interpreted as "ancient old", which makes sense. Is it possible that uralt may have led to ur- being used as an intensifier?

Also, is what you're saying that über- is generally analogous to English's "overly"? If so, then would it be fair to say that English über(-) and German über- are false friends (not particularly uncommon with loanwords)? Perhaps we ought to note that in the English entry for the word. Tharthan (talk) 21:57, 18 April 2019 (UTC)

1. I doubt that ur- in uralt originally meant “ancient”. There is a difference in the meaning of this prefix for nouns and for adjectives. For nouns (next to being an indicator of distance in kinship) it has the meaning of “original” or “primeval”. An example is Urfassung, meaning the orginal version (of a play etc.). Another example is Urknall for “Big Bang”. Since the prefix turns a noun into a noun, it is usually capitalized in this role, but not of course when the noun is turned into an adjective, like Ursprung → ursprünglich. For adjectives, it does not have this meaning. Instead, it is an intensifier, meaning “very” (as in urplötzlich) or “thoroughly”, “to the core” (as in urgesund).
2. If English people and French people are false friends, or English parking and French parking, then so are English uber- and German über-. But do we need to note this? There is an abundance of words that are written the same but have different meanings in different languages, like English wetter, German Wetter and wetter, and West Frisian wetter.
3. Indeed, “overly” is a reasonable sense for one of the meanings of German über-, but only when applied to adjectives. For nouns one could use "over-” (Überbevölkerung = “overpopulation”) or “excess” (Überangebot = “excess supply”, “surplus”). Of course, that is only one of its meanings; another quite common one is indicating a higher position, literally or metaphorically, as in überschatten = “to overshadow”, or motion across a gap, as in übersetzen = “to translate”.  --Lambiam 20:57, 19 April 2019 (UTC)

Thanks for the elaboration. In regard to wetter, Wetter, etc.- I was specifically talking about false friends that were not false cognates.

Why is it that you feel the possibility to be particularly unlikely? antik is a French loanword, Old High German firni’s modern descendant has a much narrower meaning than it even had in Old High German (which was not necessarily "ancient" even then), Old High German gamal does not seem have a modern descendant meaning ancient. The development of the meaning "original, proto-" for the prefix does not seem to be entirely unique to German (I was actually going to note oorsprong and oorzaak, but thinking about it, those are actually quite questionable in regard to something like this. The very fact that Modern Dutch oer- exists brings that claim into serious question), although one could probably convincingly argue that that is still ultimately due to High German influence.

Tharthan (talk) 03:25, 20 April 2019 (UTC)

The reason I find it unlikely that ur- in uralt originally meant something like “primeval” is that this is the only true adjective (not an adjectivalization of another word) in which that sense fits. I think it more likely that the ur-commonality was a sense “through and through”, “to the core”, which was widened to ”genuine”, “unadulterated”, “pristine”, “original”, “primal”.  --Lambiam 09:19, 20 April 2019 (UTC)

One could theorize a) it to come from Greek, with the aspiration and all and/or b) a connection to carui, which is listed in the etymology of caraway, on accounts of sounding similar and having similar meaning. Both isn’t sourced (I haven’t researched thoroughly though) but it would just be a theory

Anatol Rath 19.4.2019 18:37 UTC

One may theorize anything, but to make this even mildly plausible one should be able to identify a specific possible Greek origin. Descendence from Ancient Greek καρώ (karṓ) is unlikely, in light of the form. There is a Greek word χαρά (khará), but that means “joy”.  --Lambiam 21:10, 19 April 2019 (UTC)

The word is found in Caesar’s Commentarii de Bello Civili, book 3, chapter 48. In the context it is clearly a local term, not supposed to be Latin. The setting, if I am not mistaken, is Illyria, so it may have been a word in some Illyrian language. Unfortunately, what is known about the Illyrian languages is extremely fragmentary. Also, the people who called it thus may have been Valerius’ men, who presumably were mercenaries from all over the Mediterranean. It is a mystery to me where the idea of this possibly being wild cabbage comes from; the text states explicitly “sort of root” (genus radicis), which is about the only part of cabbage that is not eaten; moreover, cabbage was well-known to the Romans, among whom Cato was a firm believer in the medicinal value of this vegetable.  --Lambiam 21:42, 19 April 2019 (UTC)

The footnote on this page gives an overview of theories of the meaning of chara, among which Crambe ta(r)tarica and root of caraway, but also some others. This seems a better reference than Gottwein. None of this helps to clarify the etymology, though.  --Lambiam 21:59, 19 April 2019 (UTC)

It is obviously the eddo or taro, Colocasia spp., for this is what Portuguese and Spanish cará mean, on which also see Yemeni Arabic كُرّ (kurr, “eddo”). The nineteenth-century classical philologists could of course not think of this plant, eating culture being not that global yet as in the Roman Empire. 😜

Regarding the “wild cabbage”, Crambe tatarica (Wikispecies and Wikipedia entry missing) or Tartar bread plant is indeed described to have fleshy and sweet roots; the idea of it being chara is apparently from Thiébaut de Berneaud (Wikispecies and Wikipedia entry missing, only French Wikipedia entry). Fay Freak (talk) 10:10, 20 April 2019 (UTC)

But all the species of Colocasia are native to southeastern Asia, nothing farther west than Nepal and Bengladesh. How did they come into Latin? DCDuring (talk) 13:08, 20 April 2019 (UTC)

It is said that taro has been cultivated in Egypt, Palestine, Cyprus, Greece, Italy, and Herodotus, Theophrastus, Pliny the Elder, Dioscorides, Apicius, all referred to it in their writings, but it was referred to under many different names. Fay Freak (talk) 16:32, 20 April 2019 (UTC)

Apicius’ colocasium, which is a Latin borrowing of Greek κολοκάσιον (kolokásion), is usually identified with Nelumbo nucifera, more commonly known as the (sacred) lotus. The rhizome of Nelumbo nucifera can be eaten raw or prepared in a variety of ways similar to potatoes (boiling, frying), or be used to make flour. What evidence is there that this identification is incorrect, and that the word instead refers to taro (colocasia)? Also, this root vegetable was well-known in Roman cuisine, so isn’t it a bit strange that neither Caesar nor, apparently, any of the people around him could recognize it and name it themselves?  --Lambiam 09:02, 21 April 2019 (UTC)

(edit conflict) Don't forget that the Romans had contact with India. Also, Egypt had some trade with India before that. There are those who claim Latin colocasia was taro rather than lotus. The main problem I have with the taro theory is that Asian plants tended to spread west from the eastern Mediterranean- you'd think that the Romans would have found out about it before it reached Spain. Spain as a separate route for Asian things to reach Europe seems like more of a medieval Christianity/Islam phenomenon. Chuck Entz (talk) 17:18, 20 April 2019 (UTC)

Are there good sources for Spanish and Portuguese etymologies of cará, including dates of attestation? DCDuring (talk) 18:14, 20 April 2019 (UTC)

If a root veggie named khara (or something like that) grew – apparently in the wild – in Illyria around 50 BC, it is conceivable that its cultivation reached the Iberian peninsula only centuries later. But if the name did not change much on this trek, one would expect it to have left some evidence on its trail. Moreover, it seems that Portuguese cará originally referred to the yam, so it may be an indigenous (Brazilian) word. It is first found in the writings of Joseph of Anchieta. Later taro was also called cará in Portuguese, and the Spanish word may have been a borrowing from Portuguese.  --Lambiam 09:02, 21 April 2019 (UTC)

The yam was reportedly brought from Africa to Brazil by slaves and cultivated by them, so it appears possible that cará is from some African language.  --Lambiam 15:52, 22 April 2019 (UTC)

The Portuguese Wikipedia states (citing: Ferreira, A. B. H. (1986). Novo Dicionário da Língua Portuguesa. Segunda edição. Nova Fronteira, p. 346) that the word comes from Tupi ka’rá.  --Lambiam 16:04, 22 April 2019 (UTC)

Presumably then, the Tupi term referred to Xanthosoma, a New World genus, probably more specifically Xanthosoma sagittifolium (blue taro, purple taro), but was applied to the very similar Old World Colocasia and others. DCDuring (talk) 17:55, 22 April 2019 (UTC)

The Portuguese taioba, taiá however is already alleged to be the Tupí name for Xanthosoma (and is also used for Colocasia in Portuguese). Fay Freak (talk) 19:05, 22 April 2019 (UTC)

Again according to the Portuguese Wikipedia, in the southeast of Brazil (São Paulo, Rio de Janeiro), where Anchieta worked (he co-founded both cities), the name cará is used for Dioscorea species (in particular Dioscorea trifida), while Xanthosoma species are called inhame.  --Lambiam 22:30, 22 April 2019 (UTC)

The Germplasm Resources Information Network of the USDA classifies Dioscorea trifida, or “Amerindian yam”, as native to Southern America, including Brazil.  --Lambiam 22:49, 22 April 2019 (UTC)

I was trying to account for its application to Old World taro, but I suppose it could come to be applied to many starchy roots/corms/tubers however it was used originally used in Brazil. DCDuring (talk) 02:52, 23 April 2019 (UTC)

(edit conflict) There are local Dioscorea species throughout the tropics, but words like yam, inhame and ñame originally referred specifically to African Dioscorea yams. Unfortunately, words for common root vegetables are shuffled and mixed and matched in bewildering ways, and the species themselves travel all over the place as well: potato now refers to Solanum tuberosum, but it traces its origin from batatas, which was originally Ipomoea batatas (now qualified as sweet potato). Most Americans will tell you that yam refers to a type of Ipomoea batatas, because they know nothing about Dioscorea yams. In my experience there tends to be one iconic root crop in a given language whose name is qualified by various adjectives to produce the names of many other root crops, and the identity of that root crop varies widely. Trying to figure out the history of common names for root crops can be very confusing. Chuck Entz (talk) 03:00, 23 April 2019 (UTC)

I think that model applies to many vernacular names of organisms, though it may me more dramatically applicable to starchy roots/tubers, etc. Words for corn, millet, hawk, eagle, shrimp have similar characteristics. DCDuring (talk) 03:26, 23 April 2019 (UTC)

@Benwing2 At the moment, the Balto-Slavic page has the accent on the root, which is the form Derksen reconstructs in his Baltic and Slavic etymological dictionaries, though with a question mark. It is also the form you'd expect to arise from Hirt's law if the accent was on the second syllable beforehand, which is expected for an oxytone nominative singular. The circumflex in Slavic *sy̑nъ would then be a result of Meillet's law, which de-acutes a root accent in mobile nouns like this one. However, the Lithuanian form sūnùs has final accent, which cannot be reconciled with Derksen's reconstruction. The only right-shifting law that could be applicable is De Saussure's law, which shifts the accent from a non-acute onto an acute one. But that can't apply here because the root is acuted.

Jasanoff instead takes the position that the final accent is original, and is a result of an "analogical repair" that happened whenever the oxytone forms of a mobile paradigm shifted onto the root through Hirt's law. The result would be a mobile paradigm that nonetheless has no mobile accent, which was disallowed. The analogical repair could go one of two ways: either restore the mobile accent by undoing Hirt's law, or convert the paradigm to a fixed-accent one. He names *sūˀnús and *gīˀwás as two of the words that ended up undoing Hirt's law, with the nominative singular reverting to end-accent, the expected accent for mobile paradigms and the one found in Lithuanian. The explanation for the root accent in Slavic is then that it's actually the accusative form, which had become identical to the nominative except for the accent, and thus ousted the original nominative. We know that such a replacement of the nominative by the accusative happened in other Slavic o- and u-stems, given that the inherited o-stem ending ought to be *-o < *-as.

The question, then, is whose reconstruction we give more weight to (i.e. have as the canonical page name). I find Jasanoff's explanation plausible, whereas Derksen just adds a question mark and offers no explanation. —Rua (mew) 00:12, 20 April 2019 (UTC)

Why does Germanic have a short vowel in this word? Avestan seems irregular in this respect too. Crom daba (talk) 19:24, 22 April 2019 (UTC)

Likely emotion motivated alteration of the forms. It is rather silly to expect that all went according to rules. Even in the attested languages under the Proto-Germanic you see arbitrary variants. Fay Freak (talk) 19:32, 22 April 2019 (UTC)

Any thoughts on whether a Proto-Austronesian word for the coconut has been reconstructed, which would be the ancestor of *ñiuʀ and any other words for coconut in the Formosan languages? The Austronesian Comparative Dictionary yields nothing. --Corsicanwarrah (talk) 16:08, 20 April 2019 (UTC)

@Beldonstevens updated the etymology of impeachment to suggest a derivation from Latin impedīre (“to entangle; to fetter”). This is eminently plausible, but I have reverted it for now as the OED suggests Late Latin impedicāre (“to catch; to entangle”), present active infinitive of Latin impedicō (“to entangle; to fetter”). Advice welcome. — SGconlaw (talk) 07:20, 24 April 2019 (UTC)

Le Trésor also gives Low Latin impedicare as the ancestor of French empêcher, earlier empescher, from which Old French empeschement or empechement, which was borrowed into Middle English as empechement, which became today’s impeachment. So this French authority agrees with the OED. Of course, the OED may have relied on the same scholarly sources as Le Trésor. The French term empêchement means impediment, borrowed from Latin, based on the verb impediō, so the alternative etymology is semantically reasonable. However, looking at the similar development of Latin praedicō → Old French preescher → English preach, I think impedicō has the better papers.  --Lambiam 08:10, 24 April 2019 (UTC)

The rationale for the change of etymology supplied in the edit summary is bogus, see “impedico”, in Charlton T. Lewis and Charles Short (1879) A Latin Dictionary, Oxford: Clarendon Press.  --Lambiam 08:19, 24 April 2019 (UTC)

Ah, thanks. — SGconlaw (talk) 08:42, 24 April 2019 (UTC)

On săvârşi and sfârși, the etymology currently reads "From Slavic sŭvŭršiti" which I think I got from this source. Russian совершить – which is cognate – gives its own derivation as Old Church Slavonic съврьшити (sŭvrĭšiti) and so I was thinking of making the Romanian entries more specific. The only problem is ŭ (ъ) and ĭ (ь) represent different phonemes (/ɨ/ and /ɪ/) in Old Church Slavonic. Old Church Slavonic dictionaries (+ ) list съврьшити but Romanian dictionaries (including the Romanian Wiktionary) all give sŭvŭršiti.

It's probably a transcription error, in which case I'm inclined to believe the OCS-specific sources, especially since is closer to the pronunciation of the modern Romanian words which might have motivated the mistake. I just wanted to check first no one has any other ideas. ─ ReconditeRodent « talk · contribs » 12:47, 24 April 2019 (UTC)

/ɨ/ is probably not a good value for *ъ, something like /ʊ/ seems likelier. Note that the Proto-Slavic/Common-Slavic form is *sъvьršiti and that OCS exhibits liquid metathesis (or perhaps this stands for a syllabic liquid). I don't know whether Romanian âr is a direct adaptation of an earlier *ьr or of a later syllabic /r/. Crom daba (talk) 15:26, 24 April 2019 (UTC)

Keep in mind the relative chronology. The vowels ь and ъ were mostly gone by about 1000-1100, so if the word was borrowed with these vowels still in place, it must have been before then. It's then necessary to ask what the state of Romanian phonology was at the time. Did it already have the vowels î and ă?

@Crom Daba The exact realisation of ь and especially ъ was very dialect-specific towards the end of their life, so you have to take into account what the pronunciation was in the dialects closest to Romanian. In Serbo-Croatian, both ь and ъ show up as a, so an open-mid schwa-like realisation is likely. Such a realisation is still found in modern Bulgarian for ъ, while ь became e. So at least for the purpose of Romanian borrowings, neither of these vowels was rounded. A second point to keep in mind is that in Serbo-Croatian, when next to r, both of these vowels appear as a syllabic sonorant, without any vowel. So if the word was borrowed from a SC dialect, we may wonder how Romanian speakers interpreted this /r̩/. —Rua (mew) 15:51, 24 April 2019 (UTC)

Okay, it turns that most Romanian loanwords were borrowed before Old Church Slavonic, as evidenced by the absence of Crom daba's liquid metathesis (Slavic CъR → CRъ). But they are affected by the (partial) South Slavic merging of ь and ъ, as Rua said, so Common Slavic CъR and CьR show up as CɨR in modern Romanian. In which case, "sŭvŭršiti" makes sense as a specifically Proto-South-Slavic reconstruction, one hop from Common Slavic *sъvьršiti.

OCS съврьшити/sŭvrĭšiti would still probably have been recorded after the ь-ъ merger, though, and I know OCS is also South Slavic but maybe that could that just be a reflection of the accent/dialects of non-Southern scribes? ("sŭvrŭsiti" also happens to appear in a word list in the source from a moment ago.) ─ ReconditeRodent « talk · contribs » 00:14, 26 April 2019 (UTC)

@ReconditeRodent I'm not sure if you can put much value in OCS liquid-yer metathesis. Evidence from later South Slavic languages suggests that it never happened. For example, looking at the South Slavic descendants of *bъrkъ, *gъrdlo, *pьrstъ, *sьrdьce, *sьrna, OCS is the only language to consistently show liquid-yer metathesis. Bulgarian has it only sometimes, but not generally, and none of the other Slavic languages have it at all, not even Old East Slavic. Given that OCS is thought to have started off from an old form of Bulgarian, it makes sense that Bulgarian would be the language to show the effects of it, but it's more striking that it usually doesn't. This has led some scholars to question whether the metathesis we see in writing is actually a phonetically accurate representation of the speech of the time, or whether it's really just a way that OCS scribes enforced having every syllable end in a vowel, when in reality the pronunciation may have been a syllabic sonorant in OCS already. —Rua (mew) 11:13, 30 April 2019 (UTC)

Not only that, but some Macedonian Church Slavonic materials show that Macedonian kept the distinction between original *CЪRC and *CRЪC sequences, with the former always becoming syllabic (влъкъ (vlŭkŭ) from *vьlkъ) and the later remaining an actual liquid-yer sequence that could vocalize according to Havlik's law (кровь (krovĭ) from **krъvь, leveled form of *kry), and both were written the same before Havlik's vocalization as влъкъ (vlŭkŭ), кръвь (krŭvĭ).

Also all of these supposed South Slavic isoglosses are innovations dating to Common Slavic or later. Crom daba (talk) 16:14, 30 April 2019 (UTC)

I think that's probably the biggest evidence. If later languages consistently preserve the original ordering of the yer and the r (say that ten times fast), then that's evidence that metathesis, which would have eliminated this distinction, never took place. —Rua (mew) 12:53, 2 May 2019 (UTC)

According to Microsoft® Encarta® 2009, granny --Backinstadiums (talk) 00:07, 25 April 2019 (UTC)

⟨grannam⟩ is not a pronunciation. It may be a pronunciation spelling. Merriam–Webster states that grannam is pronounced /ˈɡranəm/. Collins gives the pronunciation /ˈɡrænəm/. For grandam M–W has /ˈɡranˌdam/, /ˈɡrandəm/, and Collins /ˈɡrændəm/, /ˈɡrændæm/. (I guess by /r/ they mean the phoneme we write as /ɹ/.) The -y ending is, in my eyes, indubitably the diminutive suffix, used here for endearment, so we have gran + -y. The fact that granny is used exclusively for a female grandparent makes it look more likely that here gran is not a shortening of grandmother pronounced /ˈɡɹænˌmʌðə/ but indeed of grandame, grandam or grannam.  --Lambiam 07:12, 25 April 2019 (UTC)

I've noticed that there are some cases where an accent appears on a vowel that is a reflex of a Proto-Indo-European laryngeal, such as ἀρόω (aróō), ἄημι (áēmi), κάπτω (káptō), ἀπάτωρ (apátōr), γενέτωρ (genétōr). Laryngeals, as consonants, could not have an accent in PIE, so this clearly isn't the original state of things. There are also cases of unaccented zero grades appearing with an accent in Greek, like ἄμβροτος (ámbrotos). How did it come to be that way in Greek then? I'm not all that well-versed on all the accentual laws in Greek, but I was under the impression that there are very few of them and Greek mostly preserves the original accent in short words. —Rua (mew) 14:07, 29 April 2019 (UTC)

@Rua: Greek doesn't preserve the original accent exactly. In addition to the restriction of the accent to oxytone, paroxytone and (only with a short ultima) proparoxytone you're probably familiar with, Greek is usually taken to have underwent Wheeler's law (HLĹ → HĹL), Bartoli's law (LH́ → ĹH, perhaps only in trisyllables) and Vendryes' law (LH́L → ĹHL, in Attic only), although you can still find some problems after applying these adjustments. Crom daba (talk) 23:43, 29 April 2019 (UTC)

As far as verbs are concerned, it's a tendency for verb forms to have recessive accent, which doesn't care about where a vowel derives from. So in ἀρόω (aróō) for instance, the ο (o) has to be accented; the accentuations *ἄροω (*ároō) or *ἀροώ (*aroṓ) or *ἀροῶ (*aroô) would break the recessive accent rule. This tendency means that accent in verb forms is not very well preserved.

And there are a number of formations that also tend to be recessively accented, like adjectives with the privative prefix ᾰ̓- (a-) except for third-declension adjectives in -ης (-ēs) like ἀπαθής (apathḗs). This explains ἀπάτωρ (apátōr) and ἄμβροτος (ámbrotos).

And by the rules of accent, γενέτωρ (genétōr) can't be accented *γένετωρ (*génetōr) to follow the Proto-Indo-European *ǵénh₁tōr because words with a long final vowel always have accent on one of the last two syllables, except for some words like πόλεως (póleōs). (My personal notion is that early in the history of Greek, such as during the time when the Iliad and Odyssey were composed, the accent fell back further; the pronunciation that the accent marks are the clearest evidence for is the pronunciation at the time when the accent marks were invented.) — Eru·tuon 04:52, 30 April 2019 (UTC)

Thank you both for clarifying, this is useful information for the future. @Erutuon can you tell me more about the earliest stage of Greek for which we have accentuation information, specifically the points where it differs from later classical Greek? That would be useful for our Proto-Hellenic entries, a lot of which have missing accent at the moment because it was unclear to me how the PIE accent evolved into the classical Greek one. The Proto-Greek article on Wikipedia could also do with some information about accent, so if you have any useful information or sources it would be great for future readers (including me). —Rua (mew) 10:58, 30 April 2019 (UTC)

Tacitus says that this word is what the Germans call their own spears, which implies that the word is Latinized from Proto-Germanic. Are there any attested words of similar form/meaning within the Germanic family that could be related to it/have descended from it, that could be added to an etymology for the term? GabeMoore (talk) 17:36, 29 April 2019 (UTC)

Koebler conjectures a PGmc *framjō (“lance, javelin”), but says the etymology of the PGmc word is unclear. No conclusive tie to *framjaną (“to perform, promote”), nor does he connect it with *frankô (“javelin”) (< *framkô ?) either Leasnam (talk) 04:15, 30 April 2019 (UTC)

Is there not a Latin word lubō ("I please") descending from Proto-Italic *luβēō, in the same way as its third person singular present active indicative *luβēt descended into libet (“it pleases”) ? As such, libet is analyzable as the third person singular present active indicative of lubō. Or, is the said verb unattested that Wiktionary has no entry for the same ? Thanks, — Lbdñk^{(⏳)·(🙊🙉🙈)} 18:23, 29 April 2019 (UTC)

No. Also if it is supposed to be from the same conjugation the hypothetical form is wrong to. It would be **lubeō, like taceō. Fay Freak (talk) 22:41, 29 April 2019 (UTC)

@Fay Freak: In the etymology section of English love, there is mentioning of the wrong word lubō... —Lbdñk (talk) 16:17, 30 April 2019 (UTC)

I have deleted all the cognates of love. This is an example of why one does not smear cognates all around but keeps them hedged at fewer places: it is also done so the forms are checked better to be correct. Fay Freak (talk) 16:32, 30 April 2019 (UTC)

What are the relationships of Pali tāla, Sanskrit ताल (tāla) and Magadhi Prakrit 𑀢𑀸𑀮 (tāla), 𑀢𑀸𑀟 (tāḍa) and Ashokan Prakrit *𑀢𑀸𑀟 (*tāḍa)? More importantly, what should we write here?

Can we say the Sanskrit word derives from the Pali word, the Pali word derives from Magadhi Prakrit, and the Magadhi Prakrit derives from the Ashokan Prakrit? While that agrees with the phonetics, this is not the sort of word I'd expect Sanskrit to borrow from Pali rather than some other old Prakrit. RichardW57 (talk) 07:54, 30 April 2019 (UTC)