Open main menu


Might the existence of another "letter"-bone, T-bone, have reinforced this rebracketing? Or does the bone resemble an H from any angle, which could have caused or reinforced the rebracketing? Or is it just chance? - -sche (discuss) 20:58, 2 August 2018 (UTC)


RFV of the etymology.


Our entry claims that internecine is borrowed from Latin internecīvus, despite the different ending, and without even remarking on it.

Merriam-Webster and, instead, derive the English word from a more obvious Latin source internecinus "fought to the death; very deadly, murderous, destructive". --Florian Blaschke (talk) 12:01, 4 August 2018 (UTC)

Changed by EncycloPetey over ten years ago, I dunno why. It's true that internecinus doesn't seem to be attested in classical dictionaries. Per utramque cavernam 12:05, 4 August 2018 (UTC)
Yes, I saw that edit. From the citations given here, internecīnus is attested in Late Latin, though. I don't see why post-classical Latin should be excluded or ignored. Interestingly, according to Georges, the alternative form internicīvus is attested in Cicero.
The definition given here, namely "nur mit dem Unterliegen der einen Partei od. beider Parteien endigend, v. pr. vom Krieg", i. e. "(of a conflict) only ending when one or both parties are defeated", is more precise than "destructive" or "murderous", more like "fought to the death". --Florian Blaschke (talk) 02:45, 5 August 2018 (UTC)
@Florian Blaschke: "I don't see why post-classical Latin should be excluded or ignored." Sorry if it looked like I was implying that; I wasn't. Per utramque cavernam 16:18, 8 August 2018 (UTC)


The Dutch etymology section has had an excessively long essay for over a decade. Could somebody with expertise in older Germanic languages cut down the second paragraph? ←₰-→ Lingo Bingo Dingo (talk) 14:33, 4 August 2018 (UTC)

It seems that verbiage would be better placed at the Proto-Germanic entry rather than Dutch Leasnam (talk) 16:38, 4 August 2018 (UTC)
I've simply removed it, as it really doesn't pertain to the Dutch entry specifically at all Leasnam (talk) 16:41, 4 August 2018 (UTC)


The English and Ancient Greek entries disagree slightly on whether this is from Hebrew or Aramaic. - -sche (discuss) 17:09, 5 August 2018 (UTC)

I find it also confusing that the Aramaic גולגלתא‎‎ the Hebrew entry גולגולת‎ links to isn’t the Aramaic גּלגּלת‎ the English Golgotha links to, and that the English link turns out to be another Hebrew word.
I think that the Hebrew is from Aramaic is from Akkadian is from Egyptian, because it means in Akkadian and in Egyptian “a kind of vessel” (not the kind of word that tends to be cognate over so long a time, and compare German Kopf from Latin cuppa to see how languages adopt their name for “head” from vessel names). Don’t know what that Arabic جلجلة has to do in the comparison which is some onomatopoeia and else the placename. @Vorziblix. Fay Freak (talk) 20:27, 7 August 2018 (UTC)
The Egyptian word seems to have gone in the other direction; based on attested usage, it basically meant ‘head’, with the ‘vessel’ sense derived from that. In the Old Kingdom, ḏꜣḏꜣ is only found with the meaning of ‘head’ and can be written with head determinatives as
; since the Middle Kingdom, it starts replacing tp as the basic word for ‘head’. The word meaning a kind of vessel, meanwhile, is commonly written with a final semivowel or feminine ending (ḏꜣḏꜣw, ḏꜣḏꜣt, later also ḏꜣḏꜣj), evidently a derivation, and is not attested at all in the Old Kingdom (and only rarely in the Middle Kingdom).
There are problems with assuming an Akkadian borrowing from Egyptian. Egyptian generally doesn’t correspond to Semitic g in loanwords, but only in cognates. The palatalization of Afro-Asiatic g into Egyptian in certain environments is usually considered to have happened shortly before the start of the historical era. In historical-era loanwords, Egyptian is only found rendered by Semitic or z, and rarely ḏ̣ or . If this was a loanword into Akkadian, it would have to be a prehistoric one.
I can’t speak to the Hebrew/Aramaic issues, unfortunately, as I don’t know enough in that field. — Vorziblix (talk · contribs) 09:01, 8 August 2018 (UTC)
According to Brown-Driver-Briggs, Hebrew גֻּלְגֹּלֶת gulgolet "skull" is Aramaic גּוּלְגַּלְתָּא gulgalta, similar to Hebrew גַּלְגַּל galgal "wheel". The whole word group containing "gimel lamed" denotes round objects, (stone) circles, heaps, skulls, piles, rolls, cylinders, scrolls, wheels, even bowls and basins. That site looks fishy. Gesenius says that in Arabic, the second lamed is cast away, so it's ﺟَﻠَﺠَﺔ (Freytag: cranium, et ipsum caput). If there is any Ancient Egyptian cognate, I guess it is 𓎼𓃀𓎼𓃀 gebgeb Coptic ϭⲓⲃϭⲓⲃ, to slay. -- 01:49, 11 August 2018 (UTC)
An older Germanic word for head is gebal, preserved in German Giebel, akin to Russian golova (Kluge). cf. also Latin gabalus "crux, patibulum" (DuCange). -- 03:23, 11 August 2018 (UTC)
Giebel is a gable- it has nothing to do with heads. Russian голова (golova) is from Proto-Indo-European *kalw-, and is unrelated to anything in Germanic (if it were, one would expect the Germanic form to start with an "h"). While it's tempting to bring in Latin calvus from that Indo-European root, the fact that this is a term that comes from Aramaic-speakers makes it easier to explain using the Aramaic etymology mentioned above. Chuck Entz (talk) 19:08, 11 August 2018 (UTC)

-ish (etymology 2)Edit

I'd appreciate an etymology for -ish in Romance words such as publish, relinquish, etc. I don't think it's the same as the adjectival ending that we have, or is it? Thank you in advance.

According to Webster's (1913), the ending -ish appears in certain verbs of French origin, and it corresponds to Latin -escere, an inchoative ending. There is also the adjectival suffix -ish in the sense of selfish, boyish, brutish, etc. -- 01:55, 11 August 2018 (UTC)
But the -ish ending on verbs like finish, diminish, etc. is not a suffix in English. Never has been. It carries absolutely no meaning in English. Leasnam (talk) 18:32, 11 August 2018 (UTC)

wail on, whale on, whale, waleEdit

Which is the original form(s)? Presumably the others should use "alternative form of" or "synonym of" the like, rather than duplicating definitions in so many places. - -sche (discuss) 07:19, 12 August 2018 (UTC)

wail in The Century Dictionary, The Century Co., New York, 1911 has it as a transitive verb, but directing users to wale#Verb. Among OneLook dictionaries only Wiktionary and UD have entries for wail on. whale in The Century Dictionary, The Century Co., New York, 1911 conjectures on the influences of various words referring to blows or beatings (whack, whap, whip) that begin with wh to account for whale, which they view as a variant of wale#Verb. DCDuring (talk) 15:36, 12 August 2018 (UTC)
Interesting. I've centralized the w___ on entries at whale on, and directed readers from there to whale, where the etymology mentions the possible origin from wale. (Whale does seem to be the most common spelling, now that I look at Ngrams for "whale on him", etc.) - -sche (discuss) 22:22, 12 August 2018 (UTC)
My idiolect often falls prey to the etymological fallacy, silly when a popular term is involved. DCDuring (talk) 02:46, 13 August 2018 (UTC)
No other dictionary has an entry for w(h)a(i)l(e) on except UD (whale on). But I've heard it often enough to accept it at whatever the dominant spelling. We should have some cites. DCDuring (talk) 02:50, 13 August 2018 (UTC)
I agree with making "whale on" the main entry. The phrasal verb with "on" seems to be from around 1960. DTLHS (talk) 03:23, 13 August 2018 (UTC)
The "strike to produce welts" sense of wale goes back to at least the 1400s (the Middle English Dictionary has citations of "a wykkyd wownde that hath me walt" and "tyl he be quaylled"), whereas a couple dictionaries I checked think the "thrash" sense of whale only goes back to the 1780s, supporting the idea that the ultimate origin may be wale, although whale on is most common now. Wheal also has a striking-related verb sense, and wheal on may be attestable as yet another alt form / synonym of whale on. - -sche (discuss) 04:47, 13 August 2018 (UTC)
American Heritage Synonyms has all of the following nouns as synonyms: wheal, wale, weal, welt, whelk. Most of these have similar related verb senses. DCDuring (talk) 05:27, 13 August 2018 (UTC)

Khatun خاتونEdit

I think that this necessitates a Common Turkic hub, I'm on it. Crom daba (talk) 12:26, 16 August 2018 (UTC)
@Crom daba: Uh, say what? Proto-Turkic is from 500 BCE and thus could not have been borrowed from Sogdian, but rather it was borrowed into Old Turkic from Sogdian, itself probably a borrowing ultimately from Saka, and disseminated from there. --Victar (talk) 21:20, 16 August 2018 (UTC)
@Victar Proto-Turkic could be as recent as 1st century AD (maybe even slightly more recent), but this is a "Common Turkic" reconstruction, which is analogous to "Old Turkic" in a wider sense.
We use Old Turkic ("Old Turkish" is an obsolete term) here to mean only the attested varieties of pre-13th century Turkic (excluding Uyghur and Karakhanid), which are empathically not direct ancestors to mo.dern Turkic languages.
Tell me more about the Saka connection. Crom daba (talk) 21:48, 16 August 2018 (UTC)
@Crom daba: And Sogdian at its very earliest is from about 300 CE, so again, not matching up with PT. The height of Sogdian words into Turkic languages was during the Late Sogdian Period, from about 600 CE – 1050 CE, and most, if not all, loanwords were through Old Turkic. Any way you cut it, Sogdian words into PT doesn't work from a historical standpoint. The native Sogdian form is 𐼶𐼴𐽂𐼷𐼻(xwtyn /xutēn/), and the non-native form is 𐼶𐼰𐽂𐼴𐼻(xʾtwn /xātūn/), which may point to a reborrowing from Saka *hvatujn, from PIr. *hwatáHuniH.[5] Also note Xiongnu 閼氐 (ʔɑt̚-tei), likely from Saka as well, unless you subscribe to the theory Xiongnu was an East Iranian language. --Victar (talk) 23:52, 16 August 2018 (UTC)
@Victar, and again, we're talking about Common Turkic (the ancestor of all Turkic languages except Bulgar/Chuvash), not Proto-Turkic, which could allow for an older date of breakup, especially if we allow for some early dialectal variance with continuing common innovation.
In any case western Turkic languages do not derive from the dialect of 8th century Orkhon inscriptions (so called Old Turkic or Türkü), and borrowing among them is not usually assumed.
I would like a source for the claim that most Sogdian words were borrowed during the late period, from what we know of history, Kangju were already in contact with the Xiongnu.
It is also not obvious to me that there must be a Saka word involved here. Umlaut is found in both Sogdian and Saka, Saka should change the medial alveolar into /y/. The loss of /w/ in hʾtwn is odd, but we might as well explain it as reborrowing from Turkic.
The Xiongnu word is interesting, but doesn't really speak for Saka origin either. Xiongnu containing East Iranic elements is possible considering recent genetic evidence, although my own suspicion is that some of these groups were Turkic speaking since the Iron age. Crom daba (talk) 12:24, 17 August 2018 (UTC)
@Crom daba: Haha, I see you're one of those that claims the Western Xiongnu were Turks. That's a dangerous premises to lean on -- they could have just as easily been Iranians. Did you see my source above about xātūn being a borrowing? The reason Saka is being suggested is because, unlike Sogdian, Saka exhibits *aw > . I think a more feasible hypothesis is that the word was borrowed into Old Turkic, and then itself borrowed into other Common Turkic languages which would also be consistent with its sporadicness in Common Turkic. Also claiming that the Persian and Armenian words are direct Turkish borrowing is, I'm afraid, just plain crazy. --Victar (talk) 16:06, 17 August 2018 (UTC)
@Victar Perhaps, but other ancient Turkic genomes in the given paper are more similar to Western than to Eastern Xiongnu, in any case I'm not married to the theory.
I've read Dybo's paper, yes, but it is not incontrovertible. The point about *aw > is good, but what if it took place in Turkic, what if hʾtwn represents /xātōn/ which was raised in Turkic (this is a real change, but an attestation in Turkic Brahmi could belie this theory)?
We can of course think up any story of interdialectal borrowing among closely related and sometimes poorly attested languages (why not Sogdian -> Saka -> Turkic, or why not Sogdian -> *Western Turkic -> Old Turkic) to satisfy our ever changing ideas about history, but this is premature optimization, we only know for certain that the word is distributed and behaves in Turkic as if it was inherited and that it was attested in Sogdian with a plausible internal etymology. Adding additional hidden actors to gain slightly more accuracy can only make the model less accurate as its complexity makes it trip over unknown unknowns.
Turkic reflexes of this words are not sporadic in any sense, they are found in all branches except Chuvash and Khalaj, and perhaps even in Khazar if the Armenian word belongs to this period.
I don't see anything crazy about Turkic borrowings in Persian or Armenian, would you care to elaborate? Crom daba (talk) 17:52, 17 August 2018 (UTC)
@Crom daba: I've created a {{rfd}} tag for your *xātun entry. Here is a recap of why:
  1. Iranian *xatun/*xātūn would not have been borrowing into PCT as **xatun/**xātūn, but instead as *qatun/*qātūn.
  2. Sogdian 𐼶𐼰𐽂𐼴𐼻(xʾtwn /xātūn/) is a borrowing, as indicated by non-Sogdian vowel changes.
  3. The native Sogdian 𐼶𐼴𐽂𐼷𐼻(xwtyn /xutēn/) would not have been borrowed into Proto-Common Turkic (PCT) as *qatun/*qātūn
  4. Persian خاتون(xātūn) could not have been borrowed from Ottoman Turkish قادين(qadin), but rather points to a borrowing from the aforementioned Sogdian xātūn. @Vahagn Petrosyan would have to speak to the feasibility of խաթուն (xatʿun).
@JohnC5 --Victar (talk) 20:10, 17 August 2018 (UTC)
  1. Proto-Turkic had no *x, but Common Turkic did, Clauson Studies in Turkic and Mongolic Linguistics:

    Initial x- occurs in a few words listed by Kashgari, the first authority in which this sound can be read with certainty. In most cases it is specifically said to be a secondary form of k-; some of the other words concerned are loan words. The one word in which it certainly occurred as an original initial is xa:n “king” (III, 157). This is the kind of word which might have survived from an earlier period; and taking this with the fact that the Chinese and other foreign authorities consistently spelt the tribal name Hun with an initial x- (sometimes represented by h-), it seems reasonably clear that an initial x- existed in pre-eighth century Turkish, probably as a rare sound which had almost always evolved into k- by the eighth century.

  2. Irregular changes happen, and Sogdian possesses many words derived from *hwat- related to authority while Saka has none. You also provided no arguments against reborrowing from Turkic.
  3. Maybe, but an earlier *xwatōnʲ might.
  4. Ottoman Turkish is only the most recent layer of Turkic loanwords in Persian, you might wanna consult Doerfer's Türkische und mongolische Elemente im Neupersischen
Crom daba (talk) 23:48, 17 August 2018 (UTC)
1. I'll look into Clauson claim, but I'm still not convinced.
2-3. "Irregular changes happen"? Sorry, but no, let's stick to linguistic realities instead of linguistic fantasies. The truth is this is not a native word to Sogdian, which is clearly demonstrated by that fact that there is a whole other actual native word. Actually, the very word Khotan is derived from this same root, as hvatana originally mean "men, kings".
4. Pray tell then, where would the Persian form be coming from in your theory?
--Victar (talk) 04:43, 18 August 2018 (UTC)
Attestation is the prime linguistic reality, and Saka has no kingship related words built from this root (the closest is hauta 'to have power, be able'). Bailey makes no mention of hvatana 'men' meaning 'kings' and also shows that etymologies for you give this word and Khotan are not the only valid ones. If you have a source on Saka you believe makes this one obsolete, please share it.
There are medieval manuscripts from Khorasan written in a literary register somewhere between Karakhanid and Chagatai, and Khwarezmid empire was also influental in the region and even conquered Persia by 1217 which is consistent with numerous non-Oghuz borrowings in Persian which are meticulously investigated in the above mentioned source. Sadly the expenses of your ignorance on this matter are not covered by the snark.
Crom daba (talk) 10:42, 18 August 2018 (UTC)
@Crom daba: when I say Saka, I'm referring to the common ancestor of Khotanese and Tumshuqese, which started to diverge around the same time Proto-Turkic did, not specifically to Khotanese. Here are a few more sources that echo the original one I posted.[6][7][8]
I see we're to personal insults now. I'm asking you for an etymology of the Persian word. Obviously we can't say what the word is straight from PCT. What language are you proposing this was specifically borrowed from?
--Victar (talk) 14:25, 18 August 2018 (UTC)
Old Armenian խաթուն (xatʿun) refers specifically to the title of the Khazar queen or even to her given name. Armenian sources usually derive it from one or several of the Turkic cognates discussed above, without going into detail or providing anything helpful for your debate. --Vahag (talk) 11:43, 18 August 2018 (UTC)
How old is the word? Crom daba (talk) 11:50, 18 August 2018 (UTC)
The oldest attestation is in Ashkharatsuyts, the dating of which has long been debated, but the 7th century is most likely. --Vahag (talk) 12:19, 18 August 2018 (UTC)
Which would then predate Old Turkic (8th - 13th century), making it at least Common Turkic. Crom daba (talk) 12:27, 18 August 2018 (UTC)
Old Turkic stretches back to the 7th c., and Common Turkic languages were most certainly all divergent by then. 7th c. also does not preclude my original argument, that it is a Late Sogdian (600 CE – 1050 CE) borrowing. --Victar (talk) 14:25, 18 August 2018 (UTC)
How about Sogdian → Khazar (as perhaps in לאַפּסערדאַק(lapserdak)) → Armenian? Also, note that Ashkharatsuyts contains interpolations from the 8th century. The ascription of Barsil ethnicity to the Khazar queen in the quoted passage is considered such an interpolation. --Vahag (talk) 15:02, 18 August 2018 (UTC)
I don't know enough about Armenian borrowings from Turkic to say either way, but just to mention, the Khazar-Ashkenazi theory has been roundly decredited. --Victar (talk) 03:28, 20 August 2018 (UTC)
Okay, so can I get an attestation in Tumshuqese? How do we know it was Proto-Saka language even, and not some other (equally) unattested East Iranian language? How do we know where and when Proto-Saka was spoken? If you have an answer could you cite two other Iranists who agree with your assessment?
All of the given sources go back directly to Dybo's claim. Dybo is a good Turkologist, but she is not free of controversy [9] [10].
What you're asking for is false precision, medieval Turkic langauges have only subtle differences and we cannot ascertain which dialect could have borrowed the word or even if even if it was a dialect we have any exact knowledge of. I don't even claim that the word is necessarily borrowed from Turkic, this is merely a distinct possibility.
Attested Old Turkic starts with the 8th century and this is a natural cut-off point, supported by for example Clauson's work and Rachewiltz and Rybatzki's Introduction to Altaic philology, there are different usages of the term "Old Turkic", but this one is valid and it's what us active editors working on Turkic prefer (no inheriting from Old Turkic).
There is of course some language diversity in any given language, even before the end of the period of commmon innovation. Common Turkic (not "Proto-Common-Turkic") should here be understood as a period, not as a singular point in which the unity first broke up (this point might have been older than Proto-Turkic itself as diversity has a way of waxing in actual spoken languages).
We find it useful to reconstruct PIE terms even when we don't have an Anatolian cognate, or to use the usual instead of Holzer's convention for Proto-Slavic, or PIE reconstruction which might be only Proto-Italo-Celto-Germanic, or reconstruct "Sanskrit" terms or merge Serbian and Montenegrin or American and British under one language.
In the same way, we (Turkic editors) find it useful to reconstruct terms which are at least old as the 8th century, that have regular reflexes in many languages, that have entries in etymological dictionaries of Turkic, but which cannot safely be projected to Proto-Turkic as "Common Turkic" and this convention is not in any way an affront to the use of the term elsewhere. Crom daba (talk) 16:52, 18 August 2018 (UTC)
@Crom daba:
  1. Why does Dybo claim the word is from Proto-Saka? As I already mentioned the vowels best fit with that of Saka. What other Eastern Iranian language do you propose fits better? Aside from Dybo, I recommend you read Parpola and Tremblay for Proto-Saka date estimates.[11][12]
  2. We could spend all day pettily discrediting scholars, but what does that accomplish?
  3. No, I'm asking you a simple question, which contemporary Turkic languages would have rendered in Perian the long vowels and initial *x- we find in خاتون(xātūn)? The word being borrowed from Sogdian 𐼶𐼰𐽂𐼴𐼻(xʾtwn /xātūn/) is far more straightfoward than guessing it derives directly from some unknown Turkic language.
  4. I've quite often seen the earliest Orkhon inscriptions dated to the 7th century, albeit, late 7th c. Regardless, as Vahagn mentions, dating the Ashkharatsuyts is also debated and the word may just as well be a 8th century interpolation.
  5. Who's recommending this word was inherited from Old Turkic into other Turkic languages? I'm saying that perhaps it was borrowed from Old Turkic.
  6. Your PIE analogy doesn't hold water, because we're talking about a borrowed word, not a natural one.
--Victar (talk) 03:28, 20 August 2018 (UTC)
1. Her theory probably has to do, besides linguistic data (which might include the dubious results of glottochronology), with her interpretation of anthropological data from archaeology, genetics and history; but these are liable to change as we gain new information.
Instead, language study should be grounded in language facts first, the most important of all being actual attestation, as was done by previous scholars that worked on this subject.
In any event, it is not clear how *hvatujn gets us any closer to *xātun or hʾtwn, and presumably a very similar form could have existed in the history of Sogdian anyway unless we have a reason to believe that *va > ū must precede the contraction of *-VvV- sequences and the transfer of palatality or that it must postdate umlaut.
You seem to believe that only things we know about can exist. The fact is, we have no direct attestation of the languages spoken by Iron Age Scythians from the Kazakh steppes to Western Mongolia, an educated guess is that they spoke something similar to what was spoken in Tarim a thousand years later, this is our best guess but it's not a good guess. Were ancient Saka documents never discovered, our best guess would be they spoke Alanic or Soghdic, this guess would be wrong and our guess about Siberian Saka is probably also wrong.
2. It shows that her dixit is not enough to ignore previous scholars who wrote about the subject, as her opinion is likely to be contradicted by other linguists in this branch.
3. The question is simple, but it doesn't have a singular answer. All early Common Turkic dialects would have reflexes that look like xātun, you might as well split hairs about German or Serbo-Croatian, let alone something like Chinese or Arabic. Most diversity in early Turkic languages (as in any closely related group) is found in the lexicon or morphology, what little regular phonetic innovation took place will only be recognizable in a small precentage of words and this isn't one of them. We can tell it wasn't the dialect of Kashgar since Kashgari gives the form as قاتُونْ(qātūn). If you are wondering about the long ū, we don't reconstruct long vowels for non-first syllables as they may be allophonic, but they were certainly phonetically present, especially in this word (see forms given in Clauson, although he unexpectedly doesn't have it in the headword).
4. Maybe you mean earliest Runic inscriptions, as a minority opinion dates some of the Yeniseian and Talas inscriptions (mostly fragmentary) to the 7th century, but Orkhon monuments indisputably start with the 8th. If we accept Dybo's claim that the word is found in Xiongnu or Doerfer's that it's found in Xianbei, then it is certainly old enough to belong to Common Turkic.
5. Why not borrowed into Old Turkic then? Other Turkic tribes were closer to Sogdians (or even Persians maybe?) than eastern Göktürks were.
6. Sometimes you can know if a word is borrowed and sometimes you can't, that isn't the point, the point is that all languages are spread across geography and have dialectal diversity. So did Common Turkic (pre-8th century Turkic by the definition I'm using), so did PIE, but their descendants form a unitary picture of a singular reconstructed language and trying to pick them apart into dialects we have zero information about is not only inconvenient, it is impossible. Crom daba (talk) 14:30, 20 August 2018 (UTC)
I would say no evidence for borrowing from Old Turkic into modern descendants. Besides Oghuz sense and voicing suggests it couldnt not have been reintroduced later on either.
What is the attestation of the word at the Persian side? It may very well be borrowed directly from Turkic, q-->x is not regular but not unusual, Kashgari's notes on the dialects prove this; but I dont know if the word really entered Common Turkic with x- or q-.
The word is attested in the earliest Turkic records with a lot of parallels in medieval and modern languages, I dont understand what is wrong with Common Turkic reconstruction. We must also question PGmc reconstructions from Latin too, especially when there is no Gothic cognate. Lets mark them for deletion and argue from which Germanic language they spread to whole family. --Anylai (talk) 15:43, 20 August 2018 (UTC)
@Crom daba: Her theory, which, again, mind you, is cited by multiple scholars, is based on well understood developments in both Saka and Sogdian. A stage of Proto-Sogdian would have indeed had *xwV-, but it would not have had *-ū-, and that's the key for identifying that this form does not have a native Sogdian origin. You can't just try and hammer a puzzle piece in place. If you want to formulate a theory of some other Eastern Iranian origin, I'm all ears. Every scholar is going to have others that disagree with them. Trying to discredit them outright for it is absurd.
I haven't looked into it beyond that they claim OT is first attested in the 7th century, but perhaps they are indeed referring to said Runic inscriptions. If you want to propose the word was borrowed though Xiongnu, I'm open to exploring that.
The goal here is to get the best picture we can using what we know. As I see it, you can either a) derive the word from the Sogdian borrowing 𐼶𐼰𐽂𐼴𐼻(xʾtwn /xātūn/), but it would have to be post Common Turkic (I also disagree with your assertion that everything per-8th century is CT), or b) derive the word from some other East Iranian language contemporary to Common Turkic.
--Victar (talk) 06:14, 21 August 2018 (UTC)
  1. Sogdian does have *-aw(V)- > ō (which later umlauts to ē in this word if I understand correctly) doesn't it? (That should later develop to ū during stress IV in Novák's model, although that seems irrelevant to the situation here)
  2. Development of non-first syllable *o/*ō in Turkic is a topic of importance here (I apologize for the incoming tl;dr). Modern languages don't differentiate between *u and *o in this position, except maybe Yakut and Chuvash, but details are unclear and controversial (note Yakut хотон (xoton) that might or might not imply something about this reconstruction. Stachowski gives an instance of *a..ō > o..o, but also a pattern of *a..u > a..y ~ o..u ~ o..o where he places *xātun). Old Turkic (meant here as any 8th-13th Turkic source) doesn't even differentiate o and u due to the scripts used, except for Brahmi documents (that show that *o and *u were differentiated in non-initial syllables) and perhaps Manichean documents (that might show an unrounded *o with an aleph). The third recognized evidence for Turkic non initial *o is Mongolian which can show unrounding to *a, but the earlier Mongolic documents already contain very young (Proto-Mongolic is late 12th century or so) Turkic borrowings, so it could have borrowed *katun from a Turkic language that already went through raising of *o.
  3. Dybo believes that Xiongnu were the Proto-Turkic speakers (or at least contained them), and I would tend to agree, so borrowing through Xiongnu would just mean borrowing into PT (or into Common Turkic if late enough). This seems possible, but I do not think it excludes the possibility of the word deriving from (Proto- ?)Sogdian, either through intermediaries or even directly considering that we know from Chinese histories that Kangju and Xiongnu were in contact.
    If I was to speculate on the language spoken by the Scythian tribes that were responsible for the earliest Turkic-East Iranian contacts, I would consider some sort of (Para-)Soghdic not less likely than Proto-Saka, Kunlun-Mongolia might be as hard to traverse as Transoxiana-Mongolia considering patterns of later Turkic expansion.
  4. In any case, my goal here isn't to choose or create the best theory, the goal is to create an entry that will be able to integrate information that we have on this etymon in a single location so that all concerned words can refer non-reduntantly to it and all etymological theories, including those that are less likely and those that are yet to be created, can be presented side by side for the benefit of a curious reader or an interested researcher.
Crom daba (talk) 14:54, 21 August 2018 (UTC)
@Crom daba: The i/u-umlaut is a (Indo-)Iranian mechanic and not specific to Sogdian or Saka. There was no intemediatry *-ō- period in Sogdian for this word. The development of the Sogdian form would have followed this: *hwatáHuniH > *xwatā́wnī > *xwatyʷā́yn[13] > *xwətyʷḗn[14] > 𐼶𐼴𐽂𐼰𐼷𐼻(xwtʾyn), 𐼶𐼴𐽂𐼷𐼻(xwtyn /xutēn/). --Victar (talk) 19:05, 21 August 2018 (UTC)
Interesting, I have a lot of questions about this, but we're already off-topic. Do you still seek to delete the reconstruction? Should we ask more people for opinion? Crom daba (talk) 19:19, 21 August 2018 (UTC)
@Crom daba: Are we off topic? This is the very topic for me -- that this word could not be borrowed from a native Sogdian form. As long as we have PT entry deriving from Sogdian, I will contest this entry. --Victar (talk) 01:47, 22 August 2018 (UTC)
In that case I would like to hear details of how we know that this is exactly the way Sogdian developed, proof that umlaut proceeded through exuding a semivowel rather than palatalized/labialized consonants (perhaps a false distinction phonetically, but you explanation seems to depend on such and analysis), proof that metathesis of the second syllable w occured, explanation of how this w later disappeared as well as indications for the suggested chronology (including āu > ō that wasn't to occur here). Crom daba (talk) 02:04, 22 August 2018 (UTC)
@Crom daba: I added the sources for the reconstructions. I think you're confusing the development of *āu with that of *au > ō. --Victar (talk) 02:12, 22 August 2018 (UTC)
Okay, so basically *wi, *wai, *wa_i and *au_i merged at some point into a phoneme that was possibly realized as /ɥe/. As far as I can see, nothing here excludes the possibility of monophthongization either before or after umlaut with later breaking or even the possibility of wy and whatever was the Brahmi equivalent being merely ways to represent front rounded vowels. Furthermore, I doubt that there is enough data to claim that dissimilation ɥe > ē in the presence of labial elements is a regular change rather than a sporadic phenomenon, so that *xvatɥen could also dissimilate into *xatɥen which could yield xatūn (although the long vowel of the front syllable would be unexplained). Finally it is not excluded that Turkic could adopt *xvatɥen as *xātun/xāton as it does not have /w/ or diphthongs generally (at least by Common Turkic times). Crom daba (talk) 03:05, 22 August 2018 (UTC)
@Crom daba:
  1. Just about. *wi actually became /ɥɨ/, however, and *wa needed to be either long (*wā) or stressed (*wá).
  2. Please detail the chronology of how you believe *āw became a monophthong.
  3. Why would you claim /ɥḗ/ > /ḗ/ in a labial environment as a "sporadic phenomenon"? Beside this example, we also have *hwaypaθyah > *xyʷā́ypθ > *xyʷḗpθ > 𐫟𐫏𐫛𐫔(xypδ /xḗpδ/). Do you have examples to the contrary?
  4. Do you know of any borrowings into Turkic with /ɥe/, perhaps though Chinese, that can reinforce your claim? I would be interested in the same for CwaC.
--Victar (talk) 08:12, 22 August 2018 (UTC)
Monophthongization could have happened any time or in any dialect, it wouldn't have been a phonemic change anyway and doesn't contradict other changes.
Maybe not sporadic, but perhaps dissimilation could have went the other way in this case.
I don't have a good index for Turkic Sinicisms, Clauson uses an obsolete transcription scheme so I can't easily find modern Middle Chinese reconstructions for the given terms, Dybo's paper has a lot of cases going in the other direction, Turkic rounded vowels written as falling w-diphthongs.
I am not interested in continuing this discussion, you cannot expect me to construct a theory of Sogdian diachrony that will be better than ʟ̌ubomír Novák's just to win a wiktionary dispute, but I highly doubt that his theory fits the data so absolutely perfectly and is so watertight that it can prove a negative that a predecessor or a close cousin of xwtʾynh could under no condition yield Turkic *xātun with such certainty that it necessitates invoking a reconstructed language that perhaps wasn't spoken in the region and whose descendants don't even have this word.
You object to calling pre-8th century Turkic "Common Turkic", I'm sorry but this is the periodization most useful for our work and I'm not gonna break it up into four reconstructed languages with identical entries to satisfy some classification scheme that will be obsolete in 10 years time.
We seem to have a fundamental disagreement on the level of philosophy, you want all etymologies to be structured around true theories following Neogrammarian rules, while I believe that "All models are wrong, some are useful" and that our enduring value is in collecting data (attested words and references to relevant works) and making it accessible. I am willing to live and let live, but you are resolute to show those people who haven't read the latest .pdf that supersedes all previous work how it's done.
I would ask other editors @AryamanA, Metaknowledge, JohnC5, Wyang, Vahagn Petrosyan, Tropylium, Anylai, Allahverdi Verdizade, Borovi4ok to give some comments on the arguments given here to prevent this exchange continuing in perpetuity. Crom daba (talk) 17:55, 22 August 2018 (UTC)
@Crom daba, please see my edit to the etymology *xātun and let me know if it is amenable to you. --Victar (talk) 18:29, 22 August 2018 (UTC)
@Victar, I've made a counterproposal. Crom daba (talk) 20:12, 22 August 2018 (UTC)
@Crom daba, it was still worded to strongly for me. How about this? --Victar (talk) 21:47, 22 August 2018 (UTC)
@Victar Hm, I didn't like the wordiness of my version, but this one seems to actually downplay the problematicness of Sogdian origin. I would like to keep it as deriving from Sogdian since this is what the sources usually claim (perhaps only for simplicity but still), but I'm happy to qualify it as being doubtful with our current understanding of Sogdian, is this a deal-breaker for you? Even keeping in mind that "Common Turkic" (perhaps this should be put as a label on the definition) means here Turkic partially contemporary with Sogdian?
Also this point is not important to me, but should we reconstruct Proto-Iranian *hwatáHuniH to Proto-Iranian, was *-iH not productive in the descendants? Also what's the status of *hwatáHuan? I would suggest (noninsistently) collating the Sogdian and Turkic word there. Crom daba (talk) 22:28, 22 August 2018 (UTC)
@Crom daba, I've made a new edit. Are you OK with a compromise of {{bor|trk-pro|ira-sgc}}? The original form would have to be at least Old Iranian because it had a *-iH suffix, triggering the i-umault. I'm fine deriving it from perhaps a more basal form, like *hwatáHwah. --Victar (talk) 22:58, 22 August 2018 (UTC)
@Victar, good enough, sold. Crom daba (talk) 23:20, 22 August 2018 (UTC)

Greek ΠαλαιστίνηEdit

Do we know for a fact that this was a borrowing of Hebrew פְּלֶשֶׁת‎? It apparently first appears in the works of Herodotus, who says it is between Phoenicia and Egypt. I would think it would be far more likely to have been borrowed from Phoenician. פֿינצטערניש (talk) 12:28, 17 August 2018 (UTC)

Or Aramaic, since that would have been the lingua franca in the Levant at the time of Herodotus, if I'm getting my dates right this time. פֿינצטערניש (talk) 15:31, 19 August 2018 (UTC)


RFV of the etymology. May be a folk etymology.  — Mr. Guye (talk) (contribs)  02:51, 19 August 2018 (UTC)

What exactly do you find hard to believe about it? We can look at the dates of attestation- there are no uses from before this century. That means it certainly wasn't inherited from a Middle English word. The only other possibility is affixation within English. DTLHS (talk) 02:55, 19 August 2018 (UTC)
Yes, and even if it were different it would still be correct to view it thus synchronically. Such obvious analyses do not even get stated in books and are therefore not to be found. With the same argumentation you could put a RFVE for every Hebrew or Arabic word because in most cases there isn’t anyone who has written its root down. Therefore I will remove that template now. Fay Freak (talk) 16:44, 19 August 2018 (UTC)

bus, bus bar, busbarEdit

Is the electrical sense from the automotive one or from omnibus#Adjective (short for omnibus bar)? Also, the electrical sense deserve its own homonym entry (or, better, two, for the power and data communication senses). Шурбур (talk) 08:32, 20 August 2018 (UTC)


Etymology 3, sense "Zero, no score":

"From the phrase Neither for love nor for money, meaning 'nothing'."

This is confusingly worded. Is it saying that "love" in that phrase means "nothing"? If so, I somewhat disagree. While doing something "for love" can mean doing it for nothing (in the sense of no monetary reward), I think that "neither for love nor for money" is a poor choice to illustrate the idea. In any case, I find this whole origin explanation, which is moreover presented as a definite fact rather than just a theory, rather unconvincing. I would at least demote it to "just someone's theory". What are other people's views? Mihia (talk) 19:52, 21 August 2018 (UTC)

Online Etymology Dictionary offers: The sense "no score" (in tennis, etc.) is 1742, from the notion of playing for love (1670s), that is, for no stakes. DCDuring (talk) 14:38, 22 August 2018 (UTC)
Yes, thanks, I did come across that one. Others, such as [15], are less certain, and say it is just one theory. The more general idea that it is from doing something "for love", meaning doing it for nothing, or no money, makes makes more sense to me than mentioning the specific phrase "Neither for love nor for money", in which "love" can easily be construed as a positive or valuable thing. Anyway, I have changed the explanation accordingly. Mihia (talk) 17:49, 23 August 2018 (UTC)

Indian English "only"Edit

Since this is a calque of ही, would this require a separate heading as "Etymology 2"? It doesn't currently seem to be defined under only at all and I was interested in adding it. Sorry if this is the wrong place to ask this. פֿינצטערניש (talk) 11:41, 22 August 2018 (UTC)

See [[if only]]. Is that close to or part of the meaning of ही. If so, the derivation would seem to be from that phrase by elision, perhaps under the influence of ही. DCDuring (talk) 12:16, 22 August 2018 (UTC)
I don't speak Hindi (though I have worked in India) so I can't exactly say, but "if only" is definitely not used the same way as only in Indian English. Here are some usage examples: You'll see it is used purely for emphasis. פֿינצטערניש (talk) 13:15, 22 August 2018 (UTC)
To say emphasis doesn't help me very much. The examples help a bit more. I can see that only is being used in a way that I evidently don't understand. It is interesting that I want to interpret the sentences in ways consistent with my use of only. The StackExchange discussion shows that many there persist in not understanding the usage, seeing it only as an error. The StackExchange examples would be more helpful if they were 'translated', that is, if some other appropriate English word that preserved the 'only' meaning were substituted for only. In StackExchange, indeed was substituted in one case, but it didn't seem to work in other sentences. DCDuring (talk) 14:32, 22 August 2018 (UTC)
I am focused on the definition because the curious aspect of the etymology is the source of the meaning in Indian English, not the source of the form of the word. DCDuring (talk) 14:42, 22 August 2018 (UTC)
Yes, if the calquing claim can be substantiated. DTLHS (talk) 22:15, 22 August 2018 (UTC)
Shouldn't this be {{sl}}? I would have it with the same header. Crom daba (talk) 02:40, 23 August 2018 (UTC)
That makes sense; I apologize, I'm not always entirely clear on the exact terms for things. I don't think it should be hard to substantiate, so I will do so in the near future. פֿינצטערניש (talk) 11:41, 23 August 2018 (UTC)
@פֿינצטערניש: It's not necessarily just calqued from Hindi. Most Indian languages have an emphatic particle, it could be that anyone who speaks Indian English is just calquing from their native Indian language which doesn't have to be Hindi. But yeah ही () is the equivalent Hindi word. —AryamanA (मुझसे बात करेंयोगदान) 21:30, 23 August 2018 (UTC)


Does anyone know why the gender of this word changed from masculine in Latin to feminine in French? Indeed, in the other descendants of Latin, it's masculine still. This would be great info to put into the etymology section of that word. - 02:29, 23 August 2018 (UTC)

Previous discussion: User talk:JohnC5/2017#Why is "erreur" feminine?. —Μετάknowledgediscuss/deeds 21:36, 23 August 2018 (UTC)
All those Latin nouns suffixed in -or have become feminine in French (both the inherited ones, as douleur, and the borrowed ones, as terreur), with the notable exception of amour, a borrowing from Old Occitan. Per utramque cavernam 21:50, 23 August 2018 (UTC)

כוח / قوةEdit

I'm probably wrong, but despite the difference between כ/ك and ק/ق, as well as the difference between ח/ح/خ and ה/ת/ة, I find it hard to resist the thought that these words (which both mean power/strength) are related. The sounds /k/ and /q/ are similar enough that something might have happened along the way to cause a shift in pronunciation, and this could also be true in the case of the final consonant, seeing as the original realization of ח was similar to the Arabic ح, which is close to h.

But I can't demonstrate all of this, partly because I'm not a linguist, so I was wondering if someone more competent might be able to weigh in... פֿינצטערניש (talk) 11:40, 23 August 2018 (UTC)

According to Semitic languages#Consonants, Hebrew /k/ can correspond to Arabic /q/, but for the rest of the consonants, no dice. mellohi! (僕の乖離) 13:59, 23 August 2018 (UTC)
The Hebrew corresponds apart from a not so widespread Aramaic to Arabic كَاحَ(kāḥa, to defeat), كَاح(kāḥ), كِيح(kīḥ, rugged surface of a mountain), (Freytag, Georg (1837), “كاح”, in Lexicon arabico-latinum praesertim ex Djeuharii Firuzabadiique et aliorum Arabum operibus adhibitis Golii quoque et aliorum libris confectum (in Latin), volume 4, Halle: C. A. Schwetschke, pages 68–69) and وَكَحَ(wakaḥa, to batter down) (Freytag, Georg (1837), “وكح”, in Lexicon arabico-latinum praesertim ex Djeuharii Firuzabadiique et aliorum Arabum operibus adhibitis Golii quoque et aliorum libris confectum (in Latin), volume 4, Halle: C. A. Schwetschke, page 499), and Ge'ez ኰኵሕ (kʷäkʷəḥ, rock), so Koehler-Baumgartner, p. 430 of the third edition, and Dillmann, August (1865) Lexicon linguae aethiopicae cum indice latino (in Latin), Leipzig: T. O. Weigel, column 859. I don’t see any cognates for the Arabic except Mehri qəwwēt(strength) from a well-developed root in Mehri. Fay Freak (talk) 15:23, 23 August 2018 (UTC)


Can anyone add to this list of descendants? Especially regarding Portuguese abacate and where it fits in. There are also many European languages with just "avokado" whose etymologies should be sourced. Ultimateria (talk) 00:41, 24 August 2018 (UTC)

Absolutely, thank you for pointing that out. Ultimateria (talk) 23:18, 24 August 2018 (UTC)


Does the French really come from the Spanish? The RAE lists the Spanish as coming from the French, and the Trésor doesn't mention Spanish. Ultimateria (talk) 01:46, 24 August 2018 (UTC)


We have conflicting etymologies at بيطار (from Arabic root) and albéitar (from Greek). Ultimateria (talk) 23:20, 24 August 2018 (UTC)

@Profes.I. Fay Freak (talk) 01:17, 25 August 2018 (UTC)
Nişanyan states at baytar that the Arabic term stems from Ancient Greek ἱππίατρος, referring to Vollers 620 and Fraenkel sf. 265. Although farriers may need to trim a hoof to fit a horseshoe, their main job is the shoeing and not the cutting. This makes me think that the derivation from a triconsonantal root meaning “to cut” may be a folk etymology.  --Lambiam 11:32, 27 August 2018 (UTC)
Additionally, it looks like the primary meaning, at least in present-day Arabic, is "veterinarian specialized in horses".  --Lambiam 11:50, 27 August 2018 (UTC)
And according to several scholarly dictionaries, the meaning of the root ب ط ر(b-ṭ-r) is something like to mutilate, to amputate, to cut off prematurely or utterly, to maim; actions that do not behoove an expert supposed to merely trim the poor horsey’s hooves.  --Lambiam 12:11, 27 August 2018 (UTC)
I have now fixed the etymology, assuming it is directly from Greek by the material Fraenkel proferred, from that Greek-Arabic translation environment. Fay Freak (talk) 15:12, 27 August 2018 (UTC)

boeren (coarse)?Edit

The etymology section for burlap states that the first part stems from Dutch boeren (coarse) (although this links to boeren#Latin!). However, no dictionary appears to give “coarse” as a possible sense of the word “boeren”. A boer is a peasant, and so a boere(n)lap would mean “a peasant-style piece of cloth”, which would presumably not be the finest kind of cloth. But I see no source for this theoretical etymon. I’m more inclined to seek a connection with burel.  --Lambiam 02:30, 26 August 2018 (UTC)

See it now. I see no reason to refer to Middle English, since the compound was created in Modern English from burel + lap, both modern English words. Further etymology of those components can be found at their respective entries. I've also maintained the possible derivation from Dutch boeren, since does. Leasnam (talk) 03:15, 27 August 2018 (UTC)
As far as I can tell, the mention of boeren to mean "coarse" stems from the perceived notion that boeren means "of boers". Such confusion is pretty common when non-native speakers analyze another language with which they have only rudimentary knowledge of. Leasnam (talk) 03:21, 27 August 2018 (UTC)
I have changed boeren to boer, in analogy to the etymology of Dutch words like boerenkool, which are analyzed as a compound whose first component is boer. I have accordingly removed the translation coarse, as it is inappropriate as the translation of a noun.  --Lambiam 11:06, 27 August 2018 (UTC)
The way Bauer and related words in Low German dialects is used in a depreciatory way is distinct from how “peasant” is used in English. When the German – say in the time of Luther, or Luther himself – called someone by that name, it indeed meant “coarse bloke”, “bulky fellow”, a bit old-fashioned with the technology use by farmers these days of course, whereas an Englishman who says “peasant” rather highlights that there is a distance between him and the other, he being U, the other a commoner. Fay Freak (talk) 15:22, 27 August 2018 (UTC)
According to both the Dutch and the German Wiktionary, the word may be used as a term of insult for someone with bad (or coarse) manners. See also the Dutch adjective boers, which is purely pejorative. Used as a prefix, as in Bauernwurst (lit. “farmer’s sausage“, it has acquired a positive connotation of having been produced using small-scale, traditional, artisanal, “honest” techniques, contrasted with large-scale industrial factory methods. Such pre-industrial techniques tend to result in coarser textures which industrial production then may seek to mimic. How this got transferred to an English compound with bur- is a mystery, unless there is some early Dutch carrier *boerelap. In any case, the alleged coarseness is merely a connotation, not the meaning, of the putative etymological component.  --Lambiam 17:24, 27 August 2018 (UTC)
Yes, the words cannot be glossed with such a meaning, this is misleading or at least inexact. Fay Freak (talk) 18:39, 27 August 2018 (UTC)
I've expanded the etymology based on what I could find, including with information on who suggests the (unattested) Dutch word and on what basis (other compounds do extend the sense to something like "coarse"). - -sche (discuss) 19:12, 27 August 2018 (UTC)


RFV of the etymology.

In one of Michael Quinlon's (World Wide Words - now retired)weekly emails, he sais:

The origin of "okay": United States, USA. From the USA, early 1800s.

Some college fraternity men had developed for their own use, several words of slang. A North Carolina(?)newspaper ca 1824, published an article about these "college words", and the word "okay" was mentioned. When the presidential campaign of 1828 began, one of the campaign organizations began using "okay". Usage of "okay" spread out New England, then eventually around the world.

Ukrainian громадаEdit

Any reason for splitting it into two etymologies? It's pretty clearly the same word to me. @Guldrelokk? Per utramque cavernam 13:23, 29 August 2018 (UTC)

The sense ‘community, assembly, gromada’ comes from Polish gromada. Because the two words are identical, it’s hard to decide whether to treat it as a homonymic loanword or as a semantic borrowing. Perhaps it’s worth checking if general dictionaries consider it one and the same word. Guldrelokk (talk) 22:33, 1 September 2018 (UTC)

Latvian tūteEdit

Can someone help me derive Latvian tūte (power, energy) from PIE *(h₁)tewh₂-? what would the intermediary PIE and PBS forms be? Thanks. --Victar (talk) 17:53, 29 August 2018 (UTC)