Open main menu


A user tagged the etymology with {{fact}}, writing "absolutely not what OED says". - -sche (discuss) 06:26, 4 July 2015 (UTC)

So what does the OED say? --WikiTiki89 20:14, 6 July 2015 (UTC)
Merriam-Webster says it's from Middle English "crocarde", from Middle French "crocard", which is "perhaps from croc hook (of Scandinavian origin; akin to Old Norse krōkr hook) + -ard". This Middle English Dictionary says "AF; ?cp. croquier break in pieces." - -sche (discuss) 19:54, 9 July 2015 (UTC)
The OED says it's "from Anglo-French crokard, of uncertain origin". I've updated the entry to note the various suggestions. - -sche (discuss) 20:44, 15 July 2015 (UTC)


Does the sense of "private room" come from Italian studiolo? Someone tagged the entry but never listed it here. - -sche (discuss) 06:34, 4 July 2015 (UTC)

I've removed the claim. - -sche (discuss) 22:08, 22 July 2015 (UTC)


I don't get the derivation from PIE *diwoh₁. After all, this would give †divō (or something else in ) in Latin, wouldn't it? A more obvious and regular derivation would be from PIE *dyow or *dyew.

In fact, in his Lateinisches Etymologisches Wörterbuch, Walde says that diū in the sense of "during the day" could either come from the locative *dyéwi > *dyowi > *dyow (but in view of Iove, this appears questionable; after *w, *-i is usually preserved) or the endingless locative with lengthened grade *dyēw; however, diū in the sense of "a long time" is probably originally a different word *dū (as in dūdum), which was transformed into diū under the influence of diū "during the day".

(Just in case you are wondering, the usual explanation for the iou- vs. diou- difference is that iou- is from PIE *dyow- and diou- is from the PIE Lindeman variant *diyow-, which is thought to have originally been a sandhi variant used when the previous word in the sentence ended in a consonant.) --Florian Blaschke (talk) 00:44, 7 July 2015 (UTC)

PIE root grades

Would it be a good idea to add an option to {{ine-root}} (and thus the module it invokes) for zero grade, o-grade, and lengthened grade on the headword line of the root? For example, the headword line of *leykʷ-, instead of just saying:


could say:

*leykʷ- (zero grade *likʷ-, o-grade *loykʷ-, lengthened grade *lēykʷ-)

In addition, the other grades could have their own entries as nonlemmas, with definition lines that say things like

  1. zero grade of *leykʷ-

Do other people like this idea, and if so, would someone be willing to implement it? That would be way beyond my module-editing capabilities. —Aɴɢʀ (talk) 19:14, 9 July 2015 (UTC)

What do we currently do with the other grades? --WikiTiki89 19:17, 9 July 2015 (UTC)
As far as I know, nothing. —Aɴɢʀ (talk) 19:37, 9 July 2015 (UTC)
Is it really necessary? All grades other than zero are trivially easy to figure out, and even the zero grade is not that hard once you know the syllabification rules. Making entries that direct the user to the main page aren't really all that helpful in the long run; the real problem is entries that cite nonstandard grades in the first place. Roots should always, and exclusively, be cited and linked in the full grade. —CodeCat 19:46, 9 July 2015 (UTC)
How is this different from any other inflectional information we have for other languages? --WikiTiki89 20:26, 9 July 2015 (UTC)
Root grades alone are not inflectional information though. Rather, different inflectional or derivational formations induce certain grades. So the grades are a consequence of the inflection rather than inflection being derived from grades. —CodeCat 20:31, 9 July 2015 (UTC)
So how is this different from having, for example, под- and подо- or in- and im- and ir-? --WikiTiki89 20:37, 9 July 2015 (UTC)
Let's look at it another way. Would we want different grades for Semitic roots? —CodeCat 20:44, 9 July 2015 (UTC)
What would different grades even be in Semitic roots? --WikiTiki89 21:02, 9 July 2015 (UTC)
i, a, u and zero, plus lengthened of each? —CodeCat 21:06, 9 July 2015 (UTC)
Those are just vowels, can you apply them to a root as an example of what you are talking about (let's go with k-t-b)? --WikiTiki89 21:11, 9 July 2015 (UTC)
The other ablaut grades may be easily derivable from the full grade, but the reverse isn't always true. If I encounter a zero grade *ḱun-, for example, I don't know if the full grade is *ḱwen- or *ḱewn-. —Aɴɢʀ (talk) 08:15, 11 July 2015 (UTC)
Ablaut grades seem like an odd an unsatisfactory compromise, and it's not clear what their purpose would me. Root grades are morphological units after all, not lexical ones; and root entries seem to me like they mainly exist to group together related forms. If this is to introduce non-lemma forms, why not go all the way down to specific inflected forms, as we do with all other languages? --Tropylium (talk) 16:31, 11 July 2015 (UTC)

*jьzkonь or *jьskonь

Etymology of Russian искони?

Latin phrase underlying Portuguese ontem

On Portuguese ontem "yesterday" (likewise Spanish anoche) it claims the underlying Latin is ad noctem "at night" but I think it rather should be hāc nocte "on this night". Compare Spanish hogaño which is clearly hōc annō "in this year" and hoje/hoy from Latin hodie from hōc diē "on this day". Benwing (talk) 08:16, 10 July 2015 (UTC)

@Benwing: At first glance, this seems somewhat unlikely to me. Looking at w:History of Portuguese#Historical sound changes (which specifically mentions this change) it appears that Latin words ending in -e tend to become -∅, whereas words ending in -Vm tend to go to -V (though the reäppearance of the -m in Modern Portuguese is odd. Perhaps etymological hypercorrection?). But taking the Old Portuguese onte, oonte, the etymology ad noctem > *anoite > *aõite > *oõte > oonte > onte seems very nice; whereas hāc nocte might give something like *anoit > *aõi > *oõ > .
Also L&S nox shows that hāc noctu was more common than hāc nocte; though, by Vulgar Latin such a distinction may certainly have disappeared. Just a thought. —JohnC5 14:15, 10 July 2015 (UTC)
Re “the reäppearance of the -m in Modern Portuguese is odd”: Portuguese is chock full of spontaneous nasalisation. The pattern /ˈV.Ce/ → /ˈV.Cẽ/ is an uncommon but well attested case (nuvem, pajem, -agem, outrem). — Ungoliant (falai) 14:27, 10 July 2015 (UTC)
This could be a case of progressive nasalization assimilation rather than spontaneous nasalization, c.f. mim < *mi and minha < *mia. Benwing (talk) 08:51, 11 July 2015 (UTC)
That pattern of minha and mim is NV → NṼ (where N is a nasal consonant). — Ungoliant (falai) 16:23, 11 July 2015 (UTC)
1932, Antenor Nascentes, Dicionário etimológico da língua portuguesa. lists both theories. — Ungoliant (falai) 14:22, 10 July 2015 (UTC)
@JohnC5 I don't think there's any difference in outcome of final -e vs -em. The loss of both occurs after certain consonants but not after stops. Benwing (talk) 08:48, 11 July 2015 (UTC)
@Ungoliant MMDCCLXIV Thanks. Benwing (talk) 08:48, 11 July 2015 (UTC)


Vasmer says this is an early borrowing from Ancient Greek καράβιον (karábion), κάραβος (kárabos). The earliness is evidenced by the /b/ rather than /v/. Would it be reasonable to say it was borrowed into Proto-Slavic? I compiled the following list of descendants based on Vasmer and want to put it at *korabjь:

--WikiTiki89 13:35, 16 July 2015 (UTC)

@CodeCat --WikiTiki89 14:25, 20 July 2015 (UTC)
@Wikitiki89, see Trubačóv, Oleg, editor (1984), “*korabjь / *korabъ / *korabь”, in Etimologičeskij slovarʹ slavjanskix jazykov [Etymological dictionary of Slavic languages] (in Russian), volume 11, Moscow: Nauka, page 44.--Cinemantique (talk) 17:32, 21 July 2015 (UTC)
Thanks! That's enough for me to create the entry (which I just did). There is an overwhelming amount of information in that dictionary, such as dialectal forms in various languages, that I do not have time to go through and add. I would also like CodeCat's opinion from the point of view of timing and language contact between Ancient Greek and Proto-Slavic. --WikiTiki89 17:48, 21 July 2015 (UTC)
Old Church Slavonic certainly had lots of loanwords from Greek, so having a few in Proto-Slavic really isn't that big of a deal. —CodeCat 17:57, 21 July 2015 (UTC)
But that's because of the Greek missionaries and because OCS was used to translate Bible from Greek. This word was before all that, but I don't know how long before and whether it would have been part of Proto-Slavic or entered later into each branch. --WikiTiki89 18:09, 21 July 2015 (UTC)
There definitely were ships in the mutual history of the Slavs and the early Byzantine Empire, for instance, in the constant trade between the two or in the innumerable sieges of Constantinople and other Greek cities by Avars and Bulgars. So my guess would be that the word should have already been known to Slavs by the end of VII century, when the first states with Slavic population were made, but the dialectal changes between the corresponding dialects have not yet shown themselves, so that the word could have been properly transmitted even to the northern boundaries of the people. - Myndfrea (talk) 18:51, 21 July 2015 (UTC)
@Myndfrea: Right. So how do you explain the /b/ sound rather than /v/? At the end of the 7th century, the pronunciation of the Greek beta was already somewhere in between /β/ and /v/ (based on our entry for κάραβος). --WikiTiki89 19:05, 21 July 2015 (UTC)
I think you have to look at Slavic rather than Greek for that. Proto-Slavic had no /v/ itself, and some Slavic languages still don't. The closest was /w/ ~ /ʋ/. We know that by the time Cyrillic was created, Greek /v/ matched with Slavic /ʋ/, but it didn't have to be that way in the past. Older Greek /β/ was still matched more closely by Slavic /b/ than by /w/. —CodeCat 19:10, 21 July 2015 (UTC)
That, and the fact that the loan could come through some obscure Balkan language (Gothic or Gepidic, for instance) or dialect of Greek, or even Latin, which was prevalent in the territories of early Slavic settlement. Ultimately, logic implies that the loan couldn't happen earlier than the times of Heraclius - early VII century, that is. - Myndfrea (talk) 19:23, 21 July 2015 (UTC)

Bécs, Beč, Beç etymology?

What would be etymology of Bécs, Beč, Beç? any thoughts? 13:47, 20 July 2015 (UTC)

According to Wikipedia: "The name of the city in Hungarian (Bécs), Serbian, Croatian and Bosnian (Beč) and Ottoman Turkish (Beç) appears to have a different, Slavonic origin, and originally referred to an Avar fort in the area." That should give a place to start looking. --WikiTiki89 14:22, 20 July 2015 (UTC)


The Latin verb campsō is given as the etymology for the Catalan, Italian, Occitan, Portugese, and Spanish verb cansar/cansare. Cansar in these languages means to tire, whereas the Latin campsō means I turn or sail around a place. How did turning/sailing around become to tire?

*gʷreh₂- or *gʷerh₂-?

The etymologies of gravis, βαρύς (barús) and *kuruz all say that the cluster was -re-. {{R:De Vaan 2008}} also has this form. But {{R:Philippa EWN 2009}} has the cluster as -er-. {{R:gem:Kroonen 2013}} has only the zero grade, so is noncommittal. Which is it? —CodeCat 18:38, 22 July 2015 (UTC)

Is it knowable? It looks like all descendants (those you mention as well as गुरु (guru)) derive from the zero grade. Maybe this is one of those irregular roots like *bʰuH- that didn't have a full grade. —Aɴɢʀ (talk) 19:06, 22 July 2015 (UTC)
But if there is no evidence either way, why do some sources nonetheless commit to one particular variety? What do they base it on? —CodeCat 19:08, 22 July 2015 (UTC)
Their preconceived expectations? Anyway, I just discovered *kwernuz, which looks like it has the full grade in the order -er-. —Aɴɢʀ (talk) 19:16, 22 July 2015 (UTC)
If it's related at all, that is. —CodeCat 19:27, 22 July 2015 (UTC)
Fair enough, but it could be the reason why some authors committed themselves to *gʷerh₂-. And if the full grade is *gʷreh₂-, it will be indistinguishable from the zero grade *gʷr̥h₂- in many languages (Indo-Iranian, Italic, Celtic), which makes the decision more difficult. —Aɴɢʀ (talk) 19:31, 22 July 2015 (UTC)
Latin gravis has a short "a", does that give any distinguishing information? To tell the two possible root shapes apart, we'd need either an attestation with a long ā or ō (indicating -re-) or a Balto-Slavic -er- or -ar- with an acute register (indicating -er-). I don't know anything at all about how syllabic sonorants develop in Indo-Iranian, especially not with laryngeals. —CodeCat 20:31, 22 July 2015 (UTC)
I was mistaken about Indo-Iranian. I was thinking -r̥H- became -rā- before a consonant there, but it doesn't, it becomes (at least in Sanskrit) -īr- or -ūr-. But this root seems mostly to have -r̥H- before a vowel, which means H-loss doesn't trigger compensatory lengthening. —Aɴɢʀ (talk) 05:01, 23 July 2015 (UTC)
BTW, Sihler reconstructs it *gʷr̥ru- without a laryngeal and just says "the obvious inference is that *gʷr̥- before a vowel gives L gra-". He also thinks that prae < *pr̥h₂ey shows a parallel change of prevocalic *pr̥- to pra- and suggests that trāns might also show *tr̥- to tra-. As for the reason why he reconstructs *gʷr̥ru- without a laryngeal, he just says "Evidence bearing on *gʷr̥Hu- is meager by comparison [to tenuis from laryngealless *tn̥u- rather than *tn̥Hu-], but the evidence against a laryngeal is better than the evidence in favor of one." Unfortunately he doesn't say what that evidence is. With or without a laryngeal, the problem is that there are so few instances of a syllabic sonorant before a vowel that it's hard to figure out what the "normal" outcome would be. —Aɴɢʀ (talk) 18:32, 23 July 2015 (UTC)
Proto-Indo-European generally did not have geminates. And if there was a geminate, why is there no trace of it in any descendants? I think his hypothesis is too far fetched. And there are certainly plenty of possible examples of syllabic sonorants before vowels in Germanic: just look for a zero grade -u- + sonorant + vowel, with related forms having -e- or -a- (and not -eu- or -au-). —CodeCat 19:22, 23 July 2015 (UTC)
I don't think he was actually proposing a geminate; at the same time he was writing *gʷr̥ru- he was referring to it as a syllabic sonorant followed by a vowel, so he was clearly thinking of it as *gʷr̥u-. Maybe he was thinking /gʷr̥u-/ phonemically and [gʷr̥ru-] phonetically, though he doesn't seem to come out and say that in so many words, or maybe it was a misprint. I don't know how old the Germanic -uRV- forms are; I can imagine many of them are analogical rather than inherited. —Aɴɢʀ (talk) 20:26, 23 July 2015 (UTC)
Back to the original question, ग्रावन् (grāvan) looks like it can only come from *gʷreh₂-, since *gʷr̥h₂u- gave guru- and *gʷr̥h₂w- would have given gūrv-. So maybe this is a case of Schwebeablaut, with *gʷreh₂- in Indic and *gʷerh₂- in Germanic (assuming the "millstone" word is from the same root as the "heavy" word). —Aɴɢʀ (talk) 20:50, 23 July 2015 (UTC)


Discussion moved from WT:RFV.

Not sure if this is the right place to put a question of etymology, but it seems that satyr has been listed for over three years as being derived ultimately from Hebrew שעיר‎. I haven't found anything confirming such a derivation; I suspect this assumption was taken from the KJV rendering of Isaiah 13:21, which translates שעיר as "satyr." So do some Jewish Tanach translations. But the majority of modern translations have "goats" or some iteration thereof. In any event, it strikes me as incredibly unlikely that the Greek σάτυρος (sáturos), a mainstay of Hellenic paganism, was a Hebrew borrowing, given both the phonetic difficulties of going from שעיר (śa‘ír) to σάτυρος (sátyros) and the lack of suitable horse-men in Jewish mythology (as far as I'm aware) to justify this origin for the Greek myth. Does this etymology seem plausible to any of you, or are there reputable sources giving such an etymology? Aperiarcam (talk) 05:10, 23 July 2015 (UTC)

Looks like bullshit to me. Incidentally, the Etymology scriptorium is the usual venue for questions like this. —Aɴɢʀ (talk) 05:12, 23 July 2015 (UTC)
Makes no sense to me. Someone probably misinterpreted "used to translate" as "derives from". I'll remove it. --WikiTiki89 17:50, 23 July 2015 (UTC)


The term despedir is a Portuguese and Spanish word that comes from Latin de + expetere, from peto, according to some sources like the RAE. It can be confusing as it could come from Latin de + expedire, from pes, which gave out despir instead. So I'm looking for a clarification. :) Thanks. 2001:8A0:4300:B701:F482:A6EA:1705:BCF7 20:46, 26 July 2015 (UTC)

/t/ became /d/ after a vowel in Vulgar Latin, so de+expetere > despedir is expected; /d/ disappeared after a vowel in early Iberian Romance, so de+expedire > despir (with no consonant between the p and the r) is also expected. The only way de+expedire could become despedir in Spanish and Portuguese is if it were a learnèd borrowing from Latin rather than an inherited word. —Aɴɢʀ (talk) 21:32, 26 July 2015 (UTC)
Appreciated. It's just I was reverted (as in other edits) when fixing the etymology of despedir so I wanted to make sure. I then have to return to counter-argue the reverts...


{{R:De Vaan 2008}} and {{R:gem:Kroonen 2013}} both give a very different reconstruction for this, based on a root *h₃(e)rdʰ (De Vaan only gives the zero grade). I don't know what other sources say, or where this particular reconstruction came from. —CodeCat 13:28, 27 July 2015 (UTC)

rapaz / rapariga

Most sources point to Latin rapacem (one who robs, plunders) as the root of Portuguese rapaz (boy) / rapariga (girl), but I have also found one blog post linking the word to Phoenician ḥrph (youth) or rbh (procreate). Anyone familiar with this here who can comment? – Jberkel (talk) 10:58, 29 July 2015 (UTC)

You can’t trust these types of blog. Nearly everything he posted is nothing more than wishful thinking. — Ungoliant (falai) 14:16, 29 July 2015 (UTC)

*jьskati conjugation?

Can someone more experienced look if this conjugation is good. 17:37, 31 July 2015 (UTC) Template:sla-conj-jьskati

What's wrong with using a regular template ({{sla-conj-j/a|jь|sk}}):
 ? --WikiTiki89 17:55, 31 July 2015 (UTC)
*jьskǫ / *jьščǫ ! for OCS has iskǫ, 3sg. ištetъ, 3pl. iskǫtъ (also 1sg ištǫ, 3sg

ištǫtъ, with analogical spread of the palatalized root form išt-) That's problematic, it looks like *jьskǫ is original, and *jьščǫ is maybe a proto-slavized form from OCS innovation. 20:49, 1 August 2015 (UTC)