Wiktionary:Etymology scriptorium/2016/September

The Online Etymology Dictionary finds the verb to be traceable as far back as the 13th century, but finds the noun to be traceable only as far back as the 14th century.

We state that the noun derives from Old English nōt, and was reinforced by Old French note, but whence came the verb? Is it also from Old English, or... what? Tharthan (talk) 19:28, 1 September 2016 (UTC)

I couldn't find an attested use of a verb in Old English (i.e. *notian (“to mark; make a notation”)), so Middle English is the first record we have of it (not saying that it could not or did not exist prior...we just cannot prove it). The Middle English is good enough though ;) Leasnam (talk) 15:54, 3 September 2016 (UTC)

If nōt has a long vowel, it can't be the ancestor of the modern term. You'd expect *noot. —CodeCat 16:09, 3 September 2016 (UTC)

Leasnam: Well, I reckon that if the verb has older attestations in Middle English than the noun, and the noun comes at least partially from Old English, then it wouldn't be unreasonable to assume that an Old English *notian ("to mark; make a notation) existed. It must have just been unattested.

CodeCat: Both nōt and not were in use, so we can assume that the form that gave rise to the modern term was the latter. Moreover, Old French note impacted the modern term as well, so it is possible that a *noot form of the word was rooted out early on due to the influence of the Old French term. Tharthan (talk) 17:50, 3 September 2016 (UTC)

In Middle English both note and noote were in use, and were synonymous. There must have been some confounding of the forms, because we often see the long vowel form noote being used for what were senses clearly derived from the French: melody/musical note Leasnam (talk) 23:47, 3 September 2016 (UTC)

Do both verbs actually have the same etymology? The conjugations and meanings are significantly different, but not knowing that much about Latin, I didn't want to assume anything. Andrew Sheedy (talk) 01:56, 2 September 2016 (UTC)

I've been bold and made the change, but if I am in error, please revert my edit. Andrew Sheedy (talk) 02:02, 2 September 2016 (UTC)

@Andrew Sheedy: here's what I read in the Lewis & Short about our first verb: "as a secondary form of the preced. (cf.: jungere, jugare), to drive to or toward, to go to in order to accost, make a request, admonish, etc.; like adire, aggredi; hence like these constr. as v. a. with acc., to accost, address, to speak to, call upon (very freq. and class.)." --Fsojic (talk) 20:57, 1 October 2016 (UTC)

I'm not entirely sure I understand the abbreviations, but it looks to me like I was wrong to divide the etymologies. Is that accurate? Andrew Sheedy (talk) 18:27, 2 October 2016 (UTC)

@CodeCat, could you field this one? I created the entry for *pelh₂- in order to deal with this a while ago, but I was having trouble coming up with a good way of representing De Vaan's division of these two verbs in the entry. —JohnC5 23:24, 2 October 2016 (UTC)

Purely morphologically, the difference is that the 1st conjugation verb suffixed an additional -ā- to the stem, which turned it into a 1st conjugation verb instead of just inheriting the base verb's conjugation. Exactly why this happened, or why that's significant for the meaning, I don't know. —CodeCat 00:14, 3 October 2016 (UTC)

Calabash says it derives "from Spanish calabaza (“pumpkin, gourd”), possibly from Arabic from Ancient Greek καρπός (karpós), or from a pre-Roman substrate of Iberia word *calapaccia". Calabaza admits to only the substrate hypothesis. Carabassa also admits only the substrate hypothesis, but says it initially "denot the envelope or shell of certain plants and animals", which sounds like a carapace; that entry indeed suggests a connection between the words.

1. Is the theory that calabaza is from arabic solid/referencable? Then the Spanish entry should mention it or direct users to the English entry for it (I sympathize with the goal of not putting content in too many places since we can see even right here that it will not be kept in sync).
2. Is there, in fact, a (possible) connection between calabash and carapace? then they should mention each other.

- -sche (discuss) 03:26, 8 September 2016 (UTC)

For example *němьcę and *mǫžę. Where does it come from? Hard o-stems have -y in this place. --WikiTiki89 15:17, 9 September 2016 (UTC)

The Balto-Slavic ending is, I believe, -ūns. Does that explain it? —CodeCat 15:24, 9 September 2016 (UTC)

PIE o-stems ended in -ons in the accusative plural, from which we'd expect PSl. -ǫ after hard consonants and -ę after soft consonants. So I'd say the question is rather where does the -y in hard o-stems come from? Maybe from the accusative plural of neuter u-stems, which in PIE was -uh₂? PSl. does seem to have mixed up the o-stems and the u-stems. —Aɴɢʀ (talk) 15:27, 9 September 2016 (UTC)

o > u before n in final syllables was a regular Balto-Slavic sound change, predating the merging of o with a. And there was a second change that lengthened the u. —CodeCat 15:31, 9 September 2016 (UTC)

Maybe the fronting of o after soft consonants inhibited that sound change, so that -ons > -uns > -ūs > -y after hard consonants, but -ons > -ens > -ę after soft consonants. —Aɴɢʀ (talk) 15:33, 9 September 2016 (UTC)

That doesn't work chronologically. The change of ons > ūns is pre-Balto-Slavic, so it was already complete well before the Slavic fronting came into play. —CodeCat 15:36, 9 September 2016 (UTC)

Note that Lithuanian also reflects -uns in this ending, though the -n- is strangely missing. —CodeCat 15:38, 9 September 2016 (UTC)

(double e/c) I found my answer in w:History of Proto-Slavic#Vowel fronting. The vowel was originally *ų̄, which was fronted to *į̄ after soft consonants and then merged with *ę. Unfronted *ų̄ I guess merged with *ū, losing its nasalization (and then became *y). --WikiTiki89 15:34, 9 September 2016 (UTC)

Matasović has unC/ūnC > uC/ūC as a general rule, see Russian лыко (lyko) and Lithuanian lunkas or Proto-Slavic *sъto. Crom daba (talk) 18:28, 13 September 2016 (UTC)

Many of our derivation use the word "ultimately". Whether it should ever be used in etymologies or not, it is often misused, eg stoga, "ultimately from Conestoga, Pennsylvania." Before that was the English name of the Indian tribe, Conestoga, presumably from the name in their language for themselves or for the nearby river. It is only the current state of our knowledge that would make that a stopping point.

I propose that we completely rid all etymologies of the terms ultimate and ultimately. Alternatively we need to create our own definition for how we use it in etymologies and link to the location of this definition in Appendix:Glossary. DCDuring TALK 21:30, 9 September 2016 (UTC)

I would use it when I skip steps, and show only the oldest known form. —CodeCat 21:31, 9 September 2016 (UTC)

Oppose. It is often useful to indicate that a word is from a language without giving the intermediate borrowings (usually if the exact path is unknown or ambiguous). --WikiTiki89 21:34, 9 September 2016 (UTC)

What does your objection have to do with the proposal?

I don't care about incompleteness of etymologies, because every single one is necessarily incomplete. I only care about the use of the word ultimately. DCDuring TALK 21:35, 9 September 2016 (UTC)

What word would you propose to explicitly indicate that the etymology is incomplete? --WikiTiki89 21:56, 9 September 2016 (UTC)

Whatever word might be selected by my betters should at least not do violence to the ordinary meaning(s) of the word used. In what sense is Conestoga, Pennsylvania "ultimate" in a derivation? In what sense is an inferred PIE root "ultimate"? "Ultimately" indicates that one can go no further, that one has reached some point of origin. What word or root is ever such a point?

Incompleteness has no need to be marked, being the normal state of things. Any indication that implies completeness in etymologies is simply wrong. It is also true that we rarely have a complete set of derivations, with no gaps, even for the stages in last 2,000 years, though our derivations may be the best ones we can provide or the best ones anyone can provide. DCDuring TALK 22:57, 9 September 2016 (UTC)

There's a difference between an unfinished etymology section, and an etymology section where we know for sure that we don't know something. It's important to be able to differentiate those things. --WikiTiki89 16:04, 10 September 2016 (UTC)

Yes, but I don't think the word ultimately doesn't mean what you want it to mean, at least as normally defined. If you are going to use the word in a special way where normal users can see it, you should link to a suitable definition which you might take the trouble to produce and put in Appendix:Glossary. I assume that our etymologies are written to communicate information other than to ourselves.

I'd like to see a definition of ultimately as used in etymologies. DCDuring TALK 20:32, 10 September 2016 (UTC)

Oppose. If an etymology says, say, "Ultimately from Latin 'lorem ipsum'", that simply means that the origin of the term in question is based in that. It doesn't mean that one couldn't trace the etymon itself isn't in and of itself rooted in something further, it just means that the relevance of that is somewhat arbitrary. Mind, if the etymon has a page itself on Wiktionary, further etymology can usually be found by following the link to the etymon's page. In such cases, we often say "Further etymology at etymonxyz".

In addition, to Wikitiki's point, examine the etymology of "bamboozle". Is there really more we can say about such, for instance, than has already been said? If so, I would encourage whoever can provide more information to do so. Tharthan (talk) 02:17, 11 September 2016 (UTC)

So what would the definition be? Or how would you say that a normal definition of ultimately applies? I think this abuse of the language is inexcusable in a dictionary written for normal people. Of course, it is increasingly clear that we don't care one iota about normal people. We consume resources they provide to indulge ourselves. DCDuring TALK 11:29, 11 September 2016 (UTC)

I propose that we redefine "Ultimately".Crom daba (talk) 18:17, 13 September 2016 (UTC)

I've added a non-gloss definition restricted to linguistics, but it may be more broadly applied. Other dictionaries, even those that use etymological usage examples, don't quite capture the sense, IMO. DCDuring TALK 20:08, 13 September 2016 (UTC)

I know where you're coming from with these. When I use it, I hesitate to do so because I know that it doesn't truly mean ultimately. As you said, linguistically speaking, there is always an earlier root. But like others have pointed out, how else can we indicate that a term certainly comes from another term we know, but likely through some (uncertain or unknown) intermediate? That's when I tend to use ultimately. Also, as mentioned, in this context it can mean either the furthest known or reconstructed root, or simply the furthest back you can go while maintaining a relevance in meaning for the term at hand. Especially in an expression, when you don't need to go all the way back to the Indo-European root for each word. Word dewd544 (talk) 18:12, 9 February 2017 (UTC)

I think we can safely assume that many, many of our etymologies are incomplete in the sense that some intermediate form may exist or that there may be some other influence on the derivation. Simply omitting the word ultimately would be a less presumptuous way of presenting our etymologies, which should not be pretending to a "completeness" that they cannot have. I don't think that claiming that we are using an alternative definition for ultimately is very satisfactory solution, whether or not we link to our special definition. To whatever extent etymology can claim to be scientific, the use of absolute terms is inappropriate and misleading. DCDuring TALK 19:43, 9 February 2017 (UTC)

@Simboyd is claiming that a lot of the Georgian numerals come from IE or Afro-Asiatic... Aren't numerals generally one of the most native things in a given language family?

I'll back down if Simboyd can cite credible sources for his theories. Hillcrest98 (talk) 23:10, 9 September 2016 (UTC)

Those aren't my theories, and they can't be since I'm way too illiterate in linguistics to form my own hypotheses. I just copy-paste shit from either Gamkrelidze/Ivanov, Klimov or Starostin. (see this for example). --Simboyd

Thanks. Hillcrest98 (talk) 23:48, 9 September 2016 (UTC)

Some languages do use loanwords for their numerals. Japanese has multiple names for numerals, but the basic "ichi, ni, san, shi" series that people learn first are all from Chinese. I once watched a Tagalog weather forecast on TV and noticed that the temperatures were all given in Spanish. There are claims that some languages don't really have words for numbers, just "one, a few, a lot" and the like; it's quite plausible that such a language would use loanwords for numerals after coming in contact with a language that does have words for numbers. —Aɴɢʀ (talk) 06:56, 10 September 2016 (UTC)

@CodeCat, Anglom Hey, I made the entry for *nebʰ-, ran across some inconsistencies with which I would like some insight.

1. Kroonen proposes that this root comes from a schwebeablaut of the root *dʰembʰ- (“to smoke, to steam”), whence Germanic *dimpaną ~ *dimbaną (“to fog”), etc. This would lead to *dʰnébʰ-os ~ *dʰnébʰ-es- > *nébʰos, *dʰm̥bʰ-rós > *n̥bʰrós, and so on. My question is whether there are any other cognates that have *dʰembʰ- as an ancestor. I don't find this schwebeablaut argument very convincing, particularly if this root only appears in Germanic. The LIV has 3 entries for *dʰembʰ-, but none meaning “to smoke”. Can anyone find any other sources for this?
2. Several authors reconstruct *n̥bʰrós as *m̥bʰrós, which is borne out in all of the descendants. What are the findings about this type of assimilation in PIE proper? Obviously this could just be similar assimilation occurring in all of the descendants, so I thought I'd ask.
3. CodeCat and Anglom had had a discussion about the *u in Germanic *nebulaz here. Kroonen argues that Old Norse njól could have arisen from *nebalaz or *nebulaz, though he prefers *nebalaz slightly. Where does this leave us given also Ancient Greek νεφέλη (nephélē) and Latin nebula, and can we construct a unified adjective that supports them all?

Thanks for any help! —JohnC5 20:29, 10 September 2016 (UTC)

At least nebula does not point to earlier *-u-; it's from the usual Latin medial-syllable reduction, which before results in u from any earlier vowel (per Sihler also e.g. *komsolō > consulo, *ensalto > insulto, *-tlom > *-klom > *-kolom > -culum). --Tropylium (talk) 20:47, 10 September 2016 (UTC)

@Tropylium: Trust me, I knew about Latin's ambiguity. ;) Thanks for the help though! —JohnC5 20:56, 10 September 2016 (UTC)

It looks to me like Kroonen is trying to set things up for a connection between *dʰewh₂- and *nebʰ- along the same lines as the connection between *h₂éwis and *h₂ōwyóm. That still, of course, leaves the question of where the *mbʰ might come from, but *dʰewh₂- does have lots of variation in what gets added to it, and some kind of combination of reduplication, n-infixation, assimilation and syncopation also seems vaguely half-plausible (but what do I know?). Chuck Entz (talk) 00:35, 11 September 2016 (UTC)

1) It's an ingenious solution, but we should be wary (though not dismissive) of schwebeablaut explanations. 2) Even if it hadn't assimilated in PIE times, it would have assimilated in the daughter languages, see *h₂m̥bʰí which likely came from earlier *h₂n̥t-bʰí. 3) Greek and Latin (according to De Vaan) agree on *nebʰelos, which would make Germanic *nibilaz the closest formal relation. *nebalaz/|*nebulaz would then show ablaut variation or represent intra-Germanic derivation. I don't understand all the derivational patterns of Germanic l-suffixation yet, but I'll start digging into it soon. Anglom (talk) 21:12, 11 September 2016 (UTC)

Thanks for this advice! Also, whence does the form *n̥bʰrís listed at *n̥bʰrós come? Are there descendants that reflect it? —JohnC5 21:43, 11 September 2016 (UTC)

Latin imber and Oscan anafríss are i-stems, presumably derived from an adjectival form of the word, according to De Vaan page 299. Anglom (talk) 23:21, 11 September 2016 (UTC)

@CodeCat, Victar, UtherPendrogn, Nayrb Rellimer, Florian Blaschke, Anglom, Angr, Chuck Entz, and anyone else who cares:

Where does the name Gildas come from? — This unsigned comment was added by UtherPendrogn (talk • contribs) at 20:46, 12 September 2016.

No idea. And this goes in WT:ES. —CodeCat 20:52, 12 September 2016 (UTC)

Please see: Isomorphyc (talk) 21:48, 12 September 2016 (UTC)

Does anyone have any idea where this suffix comes from? This awfully resembles some femininized derivative of -arium but I'm mostly sure I've jumped on the false cognate bandwagon after looking at -aire and -ier. Hillcrest98 (talk) 02:29, 13 September 2016 (UTC)

-ier + -ie as far as I can make out from this (I don't actually speak French).Crom daba (talk) 18:04, 13 September 2016 (UTC)

Yes, pretty much. -ie is from Latin -ia. Apparently there's no full consensus on the -er-, but it's probably from -ier and/or the verb ending -er. Later on it was simply interfixed in many cases. Kolmiel (talk) 19:51, 13 September 2016 (UTC)

However, if I understood correctly — and I don't know very much about Vulgar Latin and Old French — for -ia to become French -ie, the stress must have been shifted to the penult. If that had happened in -aria as well, it might have led to -erie, too. Kolmiel (talk) 20:03, 13 September 2016 (UTC)

@Kolmiel, @Hillcrest98 could you guys update the French/Old French Etymology sections now that this mystery has been solved? Crom daba (talk) 19:17, 20 September 2016 (UTC)

I left a message on Reconstruction talk:Proto-Indo-European/wóyh₁nom two months ago, but it doesn't seem to have attracted much attention. "Wine" is actually a classic early Mediterranean word, seen in some non-IE languages (such as Semitic) where it's unlikely to have been a borrowing from IE. I'm concerned that this word never actually existed in true Proto-Indo-European, but was borrowed relatively early into various different Indo-European branches which happened to be in or near the Mediterranean area. Certainly if some form of the general "Kurgan" hypothesis of Indo-European origins is true, then neither the word nor the physical object is at all likely to have existed on the Pontic-Caspian steppe ca. 3000 B.C. In this case, the laryngeal could well have never actually existed, but would merely be a theoretical (and historically spurious) artefact of reconstructing a word which didn't occur in the proto-language. I'm not sure how to edit the entry to reflect these probabilities... AnonMoos (talk) 23:13, 13 September 2016 (UTC)

Reconstructing this word as PIE seems to be standard practice. And there's no particular problem with it, as far as I can see. Wine is native to southern Russia and the Caucasus region. And Semitic languages may well have borrowed it from Indo-European. Why not? So the reconstruction is probably justified. — The problem is simply that it isn't agreed upon. It's a good possibility, but some authorities prefer the approach that you mentioned. I think that we need a way to distinguish between "fairly probable" reconstructions and "scientifically safe" reconstructions. Our not doing so is a real problem. Kolmiel (talk) 03:13, 14 September 2016 (UTC)

Wine was not significantly present at an early date in areas associated with the archaeological "Yamnaya horizon" or colored with the darkest color in map File:IE expansion.png. (For one thing, the Yamnaya and connected peoples did not tend to be sedentary for years, as vine cultivation would require.) Wine seems to have originated south of the Caucasus, and then spread further south into the Middle East at a time when there were no Indo-European peoples there, according to most views (except the still somewhat controversial Anatolian origins hypothesis). AnonMoos (talk) 12:18, 14 September 2016 (UTC)

I'm not an archaeologist. However, the darkest territories in the map actually include the northern Caucasus. So I don't see how that could be turned into an argument against PIE origin of the word "wine". It actually implies that the Indo-Europeans knew of wine early on. Nobody says that they were the first peoples who made wine. — Now, this doesn't prove anything, but it doesn't disprove anything either. Kolmiel (talk) 14:41, 15 September 2016 (UTC)

The darkest magenta/purple area on the map is to the north of the main Caucasus mountain range (or should be), while early wine-growing areas were apparently to the south of the main Caucasus mountain range (as far as I can understand). It's theoretically possible that some of the southernmost steppe peoples could have been exposed to wine, but there's no evidence that they made wine themselves -- and the majority of the magenta area is at least 200 miles from the Caucasus mountains, at a time when long-distance trade in bulk goods (as opposed to small precious or useful objects) barely even existed in the whole world. So it's hard to see how wine could have been a significant part of the PIE way of life. In any case, most of the IE migration paths (the arrows to the north of the Black sea and to the north of the Caspian sea) would have led the various IE groups to be many hundreds of miles distant from any wine-growing region for many centuries... AnonMoos (talk) 00:24, 17 September 2016 (UTC)

By the way: personally I also favour the non-PIE approach, but I've seen the PIE one proposed in several reliable dictionaries, so I don't see any reason to delete it. It should just be labelled in some way to show the reader that it's not agreed upon. Kolmiel (talk) 14:52, 15 September 2016 (UTC)

I would greatly appreciate any useful advice as to the technicalities and wording of such a labeling... AnonMoos (talk) 00:29, 17 September 2016 (UTC)

From both Latin vadere and aditare? DTLHS (talk) 01:09, 15 September 2016 (UTC)

If it's like other Romance words for "go", then some forms will be from vado and some from ambito and/or adito, and maybe some from eo as well. See the Etymology section of andare. The problem in this case is just that the full conjugation isn't given. —Aɴɢʀ (talk) 08:26, 15 September 2016 (UTC)

What happens in sequences of vowel-resonant-laryngeal-vowel? For example -oyHo-. Once the laryngeal disappears, the y will become a consonant, so you end up with -oyo-. Normally, syllables ending in a laryngeal become acuted, but the syllable is now no longer long, so it's no longer able to show the acute-circumflex distinction. Is this sequence distinguishable at all from -oyo-? —CodeCat 14:16, 16 September 2016 (UTC)

• East Slavic:
  • Belarusian: -∅ after е/э without reflexive ending, otherwise -ць
  • Russian: -т
  • Ukrainian: -∅ after е/є without reflexive ending, otherwise -ть
• South Slavic:
  • Bulgarian: -∅ in singular, -т in plural
  • Macedonian: -∅ in singular, -т in plural
  • Serbo-Croatian: -∅
  • Slovene: -∅
• West Slavic:
  • Czech: -∅
  • Polish: -∅
  • Slovak: -∅

I'm wondering how this ending so nearly uniformly vanished in almost every modern Slavic language. --WikiTiki89 17:47, 16 September 2016 (UTC)

Supposedly, there was a t-less ending as well, though the details are unclear. I think at least once source has suggested that -ti was the athematic ending, and that thematic verbs had -Hi or something similar. —CodeCat 17:50, 16 September 2016 (UTC)

What is meant by "thematic" and "athematic"? --WikiTiki89 17:56, 16 September 2016 (UTC)

They are terms related to IE grammar. Thematic stems end in a vowel e/o, athematic stems end in a consonant. Some of the endings change depending on this, most notably athematic -mi versus thematic -(o)h₂ (where the o is the thematic vowel). There's not a full agreement on all the details of the endings; some linguists suggest that the thematic endings were originally a lot more different from the athematic ones, but that they mostly converged on each other. —CodeCat 18:04, 16 September 2016 (UTC)

See also Category:Proto-Indo-European thematic nouns and Category:Proto-Indo-European athematic nouns. —CodeCat 18:05, 16 September 2016 (UTC)

In this case, surely Category:Proto-Indo-European thematic verbs and Category:Proto-Indo-European athematic verbs would be more relevant if they existed. In either case, though, "thematic" basically means there's a vowel e/o between the root and the ending, while "athematic" means the ending is attached directly to the root without any vocalic "glue" between them. When I was a student of Jay Jasanoff (a leading expert in PIE verbs), he taught that the 3rd person singular ending of the thematic verbs was -e, e.g. the third person singular of *bʰer- was *bʰére, not *bʰéreti (which was a later form created by tacking the athematic ending -ti onto the original form). To the form *bʰére, it was possible to add the "hic et nunc" suffix -i found throughout the present tense, yielding *bʰére-i, the source of Ancient Greek φέρει (phérei), the third person singular present of φέρω (phérō). —Aɴɢʀ (talk) 19:13, 16 September 2016 (UTC)

I always considered it to be coined in reference to Zappen, which are black birds (fulica atra). Does anyone have any hints for/against either theory? Korn (talk) 09:56, 17 September 2016 (UTC)

From what I can see, zappenduster often refers specifically to nightfall, which adds credence to the idea it refers to a last shift. "zappenduster" appears earlier (1917) than "zapfenduster" (1937) in Google Books, but both terms are relatively recent. Beyond that, the only thing I can point to either way is that in English at least, it's not the blackness of a fulica atra (coot) that's interesting, but the whiteness - someone with a shiny head is bald as a coot. Smurrayinchester (talk) 12:03, 23 September 2016 (UTC)

My line of thinking was that duster is a northern term, Zappe (for the bird) is a northern term, and Zapfenstreich is a formal term, which thus is less likely to be used in a local form. But of course that's just my personal folk etymology. I don't trust the Duden's theory to be anything else until substantiated, though. 'Bald' in this context seems to mean 'having white feathers on the head', so it invites the word play. Korn (talk) 12:35, 23 September 2016 (UTC)

Can we find references for derivation from leren? A user has just pointed out that the usual descendent of that word would seem to be lere, and other dictionaries do seem to consider the "teach" sense to have the same etymology as the "acquire knowledge" sense. They say it is a relic of (formerly standard) transitive use, like the two-syllable adjective "learned". - -sche (discuss) 15:43, 17 September 2016 (UTC)

@Angr The Old and modern Irish descendants seem a bit off here. For an initial w- you'd expect f- in Old Irish, like in the Scottish Gaelic descendant given. —CodeCat 21:54, 18 September 2016 (UTC)

On a somewhat related note, what's that desctree thing and why did you only do it there? UtherPendrogn (talk) 21:58, 18 September 2016 (UTC)

Anyway, Old Irish probably generalised the lenited variant of the word, explaining the absence of the initial f-. UtherPendrogn (talk) 22:01, 18 September 2016 (UTC)

Is that something that happens often? —CodeCat 22:09, 18 September 2016 (UTC)

I'm not an expert on Irish, but I'd say it's probably a very rare thing for the lenited version to become more common. However, íaru is indeed attested. UtherPendrogn (talk) 22:11, 18 September 2016 (UTC)

I'm not doubting that it's attested, but rather whether it descends from this particular Proto-Celtic form. The final -u is also odd. —CodeCat 22:15, 18 September 2016 (UTC)

True, Matasovic does mention it's a stretch. But "fiaru" is similar to one declension wiweru, I guess. As Matasovic said, it's a stretch. Shall we see what the others think? UtherPendrogn (talk) 22:18, 18 September 2016 (UTC)

Pace Matasović, the Brythonic words are much more likely to be borrowed from Latin (as GPC says) than inherited from Proto-Celtic. I have no idea where the Goidelic words come from, but it's definitely unlikely that sga generalized a lenited form, when exactly the opposite is what usually happens (and is apparently precisely what happened in gd, where the diminutive iaróg became feòrag). If this is a Proto-Celtic form, I'd say the sga word is more likely to be a dissimilation of *wiw- to *iw-, but in fact more than the missing f is problematic. Old Irish íaru is an n-stem and looks at first blush like it ought to come from *ērVCū/ērVCon-, where V is some back vowel (since the r isn't palatalized) and the C is some consonant that disappears between vowels (like a glide or s). The initial ē could have come from iwe- or īwe- I suppose (just as Ériu comes from *ɸīweryū), but that still leaves open the question of what the -VCon- is and where it came from. The later Goidelic vowels are all over the place too; if the sga word is íaru there's no obvious reason why the ga word isn't *iara (with a long diphthong) instead of iora (with a short vowel) or why the gd isn't *fiarag instead of feòrag. And why does Matasović reconstruct the word as a masculine when it's feminine in both branches of Celtic as well as in Latin? Even Homer nods... —Aɴɢʀ (talk) 18:23, 19 September 2016 (UTC)

RFV of the etymology:

• The expression dates back to Old English, where it literally meant to go to Hēl, who was the Goddess of Hell (also called Hēl). It was not an insult.

All the quotes for hel in Bosworth-Toller are for the Christian concept of Hell, and none use this phrase. For this to be valid, we would need evidence for usage of the phrase in Old English, not just for change in the concept of hell. What's more, even by the pre-Christian concept of hell, this phrase would still be telling someone to die- not exactly nice. Chuck Entz (talk) 13:43, 19 September 2016 (UTC)

Are we even sure this isn't a Latin calque or something like that? Crom daba (talk) 14:06, 19 September 2016 (UTC)

It's in Shakespeare (Merchant of Venice) Leasnam (talk) 02:58, 20 September 2016 (UTC)

It has a parallel in early Dutch ter helle vaaren to go to hell Leasnam (talk) 03:01, 20 September 2016 (UTC)

Okay, that takes us back to Early Modern English, but that's less than half of the way there. Given that pre-Christian Old English writing was limited to no more than a few words on various artifacts, it would seem we would need some description of pre-Christian customs in Bede or other Christian writers, which are very rare. I notice that there's basically nothing in Wikipedia on Hēl as an Old English deity. One can deduce her presence by extrapolating from Norse mythology, but I didn't see anything in Bosworth-Toller that wasn't about the place rather than the deity (I suppose I might have missed something in one of the supplements, though). Chuck Entz (talk) 04:15, 20 September 2016 (UTC)

The OED simply says to refer to "go to the devil", where there is a quote from 1394 in Latin: "Malverne Contn. Higden (Rolls) IX. 33 Excanduit rex et..dixit ei , ‘Quod si tu mihi imponas..vadas ad diabolum’.]". So maybe it is a Latin calque. DTLHS (talk) 04:21, 20 September 2016 (UTC)

Well, isn't Latin at this time really just calqueing any/all other languages ? I mean, Latin in the Middle Ages was no one's native language. Everyone who spoke Latin learnt it as a second or third language, and spoke another language first (Middle High German, Old French, Middle English, etc.), so it's more likely to have come into this Latin from someone's mother tongue, possibly even English Leasnam (talk) 21:36, 27 September 2016 (UTC)

Chaucer (1395) has: To a poure man men sholde his vices telle, But nat to a lord thogh he sholde go to helle. ; earlier still is a cite from 1300: Þe time is comen, I go to helle. Leasnam (talk) 04:49, 20 September 2016 (UTC)

Even Chaucer is much too late for a supposed pagan root. The Christianisation of Anglo-Saxon England happened in the first half of the seventh century, so that's well over five hundred years for the phrase to borrowed into English from Christendom. Plus, I'm not sure how many speakers of Anglo-Saxon English followed the Norse religion. The Old English Gospel of Nicodemus may be some help here, but it's not clear whether the "Hell" it talks about is the goddess Hel or just a feminine-gendered generic personification of the underworld. Smurrayinchester (talk) 13:57, 20 September 2016 (UTC)

I disagree on principle that Chaucer is too late for a pagan origin. Things tend to stick around in the collective oral memory, and easily could have even over those 700 or so years. Having said that, in this particular case the pagan origin makes no sense to me and the Christian hell is a much more likely and logical explanation. --WikiTiki89 14:16, 20 September 2016 (UTC)

People love to falsify pagan history, so if we have to err, it's better to err on the side of suspicion. Crom daba (talk) 15:05, 20 September 2016 (UTC)

(Sorry, I should have been clearer. I meant "too late for proving a supposed pagan root". Certainly, a pagan expression could certainly have stuck around in the language - Yule has survived for about a millennium after the actual Yule ceremony died away) Smurrayinchester (talk) 18:37, 20 September 2016 (UTC)

*kareð says the form is ultimately without the i, but caire has the i. Which one is correct, and how would you tell the difference? —CodeCat 14:26, 19 September 2016 (UTC)

They both have to go back to an immediate pre-form with the i, but that could go back to an earlier form without it. Matasović reconstructs it without the i because he wants to take it back to a PIE *kr̥yeh₂. —Aɴɢʀ (talk) 17:54, 19 September 2016 (UTC)

If the immediate pre-form had no i, what would the outcomes have been? I'm mostly accustomed to Latin, where the distinction is often moot, and Germanic, where Sievers' law has made them allophones. I'm guessing there's no such thing as Sievers' law in Celtic. —CodeCat 17:59, 19 September 2016 (UTC)

I think there's something Sievers-like in Celtic, but it may not have depended on the weight of the preceding syllable. If *karyā had gone straight into Old Irish it would probably be *cair /karʲ/ and in Welsh it would probably be *cair /kair/. —Aɴɢʀ (talk) 18:26, 19 September 2016 (UTC)

Thank you. Since all the actual attestations point to *kariyā as you say, I've created the entry there. —CodeCat 18:39, 19 September 2016 (UTC)

@CodeCat It occurs to me that in Old Irish, all yo- and yā-stem nouns end in -e, suggesting that they all ended in -iyos/-iyā before the loss of final syllables. In Welsh, on the other hand, they do not all end in -ydd/-edd, suggesting that only some of them ended in -iyos/-iyā, and some of them ended in -yos/-yā. —Aɴɢʀ (talk) 13:16, 23 September 2016 (UTC)

If Old Irish only had -iy- stems, how can we know that -y- alone would not give -e? —CodeCat 16:10, 23 September 2016 (UTC)

Strictly speaking, I suppose it could have, but that would have involved a nonsyllabic y becoming syllabic e, which is a little less likely than syllabic i becoming syllabic e. But basically it's probably fair to say that pre-Irish is like Latin here: there is no meaningful distinction between CiyV and CyV. Some other data to keep in mind, which I haven't entirely figure out yet, are: 3rd person singular present conjunct ·gaib < *gab(i)yet(i) < *gʰh₁bʰ-yéti contrasting with ·léici < *linkʷeyet(i); but 2nd person singular present conjunct ·gaibi < *gab(i)yes(i) < *gab(i)yih(i) contrasting with ·bir < *berih(i) < *beres(i); and genitive singular céili (Primitive Irish ᚉᚓᚂᚔ (celi)) < *kēl(i)yī contrasting with maicc (Primitive Irish ᚋᚐᚊᚔ (maqi)) < *makʷkʷī. I wonder if there are any Primitive Irish forms in -(i)yos attested. —Aɴɢʀ (talk) 21:17, 23 September 2016 (UTC)

At *wóyh₁nom, Persian وین (vīn, “black grape”) is listed as probably a descendent. Does this make sense phonologically? Otherwise it seems more likely that it came from Arabic. --WikiTiki89 15:22, 23 September 2016 (UTC)

*wóyh₁nom > PII/PIr. *wayina- > Persian vīn would appear to be roughly in order, but most of the descendants don't seem to show a whole lot of evidence for the laryngeal there. It would also be suspicious if cognates elsewhere in Iranian cannot be found. --Tropylium (talk) 13:30, 24 September 2016 (UTC)

The modern Goidelic descendants' entries say that this is actually a borrowing from Latin. How valid is that? —CodeCat 13:16, 24 September 2016 (UTC)

Well, you wouldn't expect a Latin loanword to have geminate /dd/ since crēdō has a singleton /d/. But both Welsh and Irish have /d/, which can only come from /dd/. Also the Celtic words have a short /e/, while the Latin has a long ē. So a Latin loanword should have given *crwyddu in Welsh and *créidid /ˈkʲrʲeːðʲiðʲ/ in Old Irish. —Aɴɢʀ (talk) 14:11, 27 September 2016 (UTC)

Ok, thanks for fixing. —CodeCat 14:32, 27 September 2016 (UTC)

No problem. I think I was the one who said the Goidelic words were Latin loanwords in the first place, so it's my own mistake I fixed. —Aɴɢʀ (talk) 14:51, 27 September 2016 (UTC)

This is given as the origin of Proto-Brythonic *kėnɨd. However, the Old Irish form appears to be canaid rather than *cainid which you'd expect from such a Celtic form. What's going on? —CodeCat 13:19, 24 September 2016 (UTC)

I don't know why the n of canaid isn't palatalized—it isn't in one of the usual depalatalization environments—but the 3rd person singular conjunct is ·cain (likewise the 2nd person singular imperative is cain) and the imperfect passive is cainte, both with a palatalized n where you'd expect it in a plain e/o-verb. —Aɴɢʀ (talk) 21:35, 27 September 2016 (UTC)

De Vaan reconstructs a ye-stem *brusyeti for this instead. Can we determine whether this is (more) correct than the current thematic inflection? —CodeCat 14:19, 24 September 2016 (UTC)

@Angr, Anglom, JohnC5 ? —CodeCat 14:02, 27 September 2016 (UTC)

I doubt the two would be different in Old Irish. —Aɴɢʀ (talk) 14:05, 27 September 2016 (UTC)

If it's of any interest, LIV and Matasović also reconstruct a *ye-stem. —JohnC5 01:02, 28 September 2016 (UTC)

I wonder what evidence they base it on. The only thing I can think of is the fact that it's a zero grade root, but zero-grade simple thematics aren't unheard of (compare the Sanskrit "tudati" type). —CodeCat 16:39, 28 September 2016 (UTC)

RFV of the etymology:

• From Old High German ganc

Duden says "mittelhochdeutsch, althochdeutsch ganc", but I can only find ganc as Middle High German, not Old High German (OHG seems to be gang). Of course, it could just be my lack of experience with OHG, but I notice that the Luxembourgish entry on the same page refers to OHG gang.

Given the variability of OHG spelling, it wouldn't surprise me if ganc were attested in OHG, but would gang or ganc be the lemma? Chuck Entz (talk) 01:29, 25 September 2016 (UTC)

Old High German did not originally feature final obstruent devoicing, which is a purely Dutch thing in the first third of the middle ages and only then creeps east and south. When it reaches High German, it doesn't make its way into the alps (and hasn't until this day) and my understanding is that "Old High German" per default is exemplified by dialects from these regions. While these lects are of course prone to general devoicing of the phonemes, i.e. initial + final, I think distinctive spelling is the most representative way to write OHG, also with regards to the modern situation, which cannot be understood if we pretend final devoicing was general. All this does already seem to be the practice, by the way: tag. Korn (talk) 08:37, 25 September 2016 (UTC)

I couldn't find ganc attested, only gang Leasnam (talk) 02:37, 26 September 2016 (UTC)

If it's a noun, why the "-"? Sobreira (talk) 14:44, 25 September 2016 (UTC)

Presumably because the descendants don't let us decide on a particular nominative case. Maybe it should be called a root instead, though. —Aɴɢʀ (talk) 18:59, 25 September 2016 (UTC)

I've been bold and changed it Sobreira (talk) 08:51, 26 September 2016 (UTC)

By the way: shouldn't we create Reconstruction:Nostratic/*ʔer-a or Reconstruction:Nostratic/*ʔarV̄ (and other Nostratics, in the cases of oṭṭi and pȧda) to cover/include/move the etymology given en R:PIE/h1er-? (even if you don't like/believe/give credit to the thesis/theory) Sobreira (talk) 21:51, 26 September 2016 (UTC)

1. Why is Category:Proto-Indo-European terms with IPA pronunciation included in Category:Proto-Indo-European entry maintenance with others needing attention? Do they have problems?
2. Is this a real preffix: Category:Proto-Indo-European words prefixed with *smḱ-? (I mean, following the phon rules, being applied to a root and the lemma created?

Sobreira (talk) 08:06, 27 September 2016 (UTC)

As to point one, it isn't just PIE. Any language's "Category:XXX terms with IPA pronunciation" is included in that language's "Category:XXX entry maintenance", though it isn't clear to me why that should be so. The entry maintenance categories are usually for problems that need to be solved, but obviously having an IPA pronunciation isn't a problem that needs to be solved. On the other hand, I wonder why there are any PIE terms with IPA pronunciation since the pronunciation of PIE is purely speculative. —Aɴɢʀ (talk) 21:41, 27 September 2016 (UTC)

Yes, I supposed also that the maintenance class should be for "terms withOUT IPA pronunciation" (unless for technical/modules supervision). I guess that info is there because it can be tracked to a source, which I find respectable. Sobreira ►〓 (parlez) 07:01, 28 September 2016 (UTC)

@Angr, JohnC5 The root here suggests a derivation from *linékʷti ~ *linkʷénti, from which you'd get Proto-Celtic *linkʷeti. But the Old Irish verb is a class A II present, with a present in *-īti in Proto-Celtic, and this class reflects causatives in *-éyeti and some denominatives, not plain thematic/athematic root verbs. So this is not compatible with the nasal-infix present reconstructed for PIE. So what's going on and how could this be resolved? —CodeCat 22:38, 27 September 2016 (UTC)

As per usual, Matasovic gives *li-n-kʷo- with no useful explanation. LIV also claims a nasal infix but has the decency to say:

“Mit *-i̯e- erweitert; zum Lautlichen vgl. SCHRIJVER, Ériu 44 (1993) 39-42, wogegen MCCONE, FS WATKINS 474-5 (lautgesetzliches *li° analogisch nach Konj. ersetzt durch *lē°).”

So we are supposedly looking at something like *lēkʷīti. As to the nasal itself, I have no notion. —JohnC5 00:43, 28 September 2016 (UTC)

I think a nasal is required, because the c in Old Irish isn't lenited. A k(ʷ) between vowels would lenit to ch. So the Old Irish form seems to require *linkʷīti, which is weird. —CodeCat 11:15, 28 September 2016 (UTC)

Yes, the c in this word represents /ɡ/, which comes from /*ŋk(ʷ)/. It seems to come from *linkʷīti/*linkʷeyeti, but I don't know why it comes from that rather than from *linkʷeti, which would have given the same absolute form léicid, but would have given the conjunct *·léic rather than attested ·léici. —Aɴɢʀ (talk) 14:35, 28 September 2016 (UTC)

I've seen the occasional "doubly-characterised" present in Ancient Greek, where there's both a nasal infix and some other present suffix as well. However, the causative -éye- is a secondary derivative suffix and was not used in any IE branch to make primary verbs (which this is), and was certainly never used together with a nasal infix. So it's quite bizarre, but I don't see any other option. —CodeCat 16:27, 28 September 2016 (UTC)

I'm trying to figure out what "lautgesetzliches *li° analogisch nach Konj. ersetzt durch *lē°" means. Normally *-ink- becomes /iɡ/ in OIr. (e.g. ·icc), while /eːɡ/ comes from *-enk- or *-ank-. Is McCone saying that the subjunctive started with *lankʷ- or *lenkʷ-? Why would it? Unfortunately I only started receiving Ériu in 1994, so I don't know what Schrijver said, nor do I have access to the Watkins Festschrift, so I don't know what McCone said. —Aɴɢʀ (talk) 19:43, 28 September 2016 (UTC)

*linkʷ- was a present verb and would be expected to be limited whatever forms that verb gave rise to in Celtic/Irish. An -ē- could only arise from a full grade of the root, which never occurs with a nasal infix. So the source of such an ē should be sought in the non-present stem. However, it looks at first glance like Old Irish regularised the nasal present stem léic- across all forms. Otherwise you'd expect to see forms with ch too, reflecting Celtic kʷ without the preceding nasal. —CodeCat 15:40, 29 September 2016 (UTC)

But the nasal present stem should be lic-, not léic-. If the subjunctive stem was *lēkʷ-ā- (from the full grade with no nasal infix) and the present indicative stem was *linkʷe/o-, then maybe the vowel of the subjunctive and the consonantism of the present indicative joined up to give *lēnkʷ- throughout the paradigm. Then Osthoff's law shortened that to *lenkʷ-, which is the source of Old Irish léic-. Still no explanation for why it's an ī (eye/o) present instead of a plain thematic present, though. If nothing unphonological had happened, the present would be *liccid, ·licc and the subjunctive would be *líachaid, ·líacha. —Aɴɢʀ (talk) 11:56, 1 October 2016 (UTC)

Ironically enough, the modern Irish verb is lig, exactly the form we would have expected from *linkʷeti (or *linkʷīti). —Aɴɢʀ (talk) 11:58, 1 October 2016 (UTC)

I've noticed recently that PIE roots ending in laryngeals insert a p in the causative. Here are a few examples:

• श्रद्धापयति (śraddhāpayati) < *ḱred-dʰoh₁-éye-ti (same root as श्रद्दधाति (śraddadhāti))
• ज्ञपयति (jñapayati) < *ǵnoh₃-éye-ti
• स्नापयति (snāpáyati) < *(s)noh₂-éye-ti (c.f. Avestan 𐬁𐬯𐬥𐬀𐬫𐬁𐬝 (āsnayāt̰), Khotanese (ysänāj-), Manichaean Middle Persian 𐫀𐫀𐫘𐫗𐫀𐫏 (ʾʾsnʾy), Sogdian (snʾy))
• भापयते (bhāpayate) < *bʰoh₂-éye-ti
• दापयति (dāpayati) < *doh₃-éye-ti
• धापयति (dhāpayati) < *dʰoh₁-éye-ti
• हापयति (hāpayati) < *ǵʰoh₁-éye-ti

I'm sure I could find more. Does anyone have information about this phenomenon? As the Iranian cognates show, it does not appear to extend outside of Indo-Aryan (or even Sanskrit proper). Was this via analogy to the normal shape of causatives, CāCáyati, and the p was chosen at random? Is there some bizarre sound law in PIA like āHá > āpá? Does this happen in any other situation besides roots ending in laryngeals? Is there some literature pertaining to this? Thanks for any help! —JohnC5 05:31, 28 September 2016 (UTC)

No idea. It would be great if besides the Grimm laws, the Szemer. law and others we would reach here an undiscovered JohnC5 law. Sobreira ►〓 (parlez) 07:07, 28 September 2016 (UTC)

I don't know either, but my historical linguist's intuition says it's probably analogy. There was probably some verb or group of verbs that had a causative ending in -payati for a good phonological reason (e.g. a root ending in p), and in these verbs the p was reinterpreted as part of the suffix, so at some point speakers started treating -payati as the postvocalic allomorph of the usual causative suffix -ayati. Whitney's Grammar discusses these forms here but he doesn't venture any explanation of their origin. If you have access to old issues of Language, Franklin Edgerton has an article in Vol. 22 (1946), pp. 94–101 on "Indic Causatives in -āpayati (-āpeti, -āvei)", which is on Scribd but I can't get to it without a subscription. —Aɴɢʀ (talk) 14:49, 28 September 2016 (UTC)

Thanks for this info. I'll look into Edgerton's stuff later, but for the moment, Whitney's description points strongly to analogy (but with what?). Also the presence of unepenthesized causatives:

“gāyaya (also gāpaya) from √gā sing, chāyaya, pāyaya from √pā drink (or pī), pyāyaya from √pyā or pyāy; sāyaya from √sā (or si); also, later, hvāyaya from √hvā (or hū)”

implies that some causatives did evolve in the expected way (though their reconstruction sometimes seems to result in an epenthetic -y-). Now my question is, from a PIE reconstruction perspective, do we count all -p-epenthesized causatives as secondary and not relevant to potential PIE causatives? —JohnC5 15:44, 28 September 2016 (UTC)

I certainly would. The causative was quite productive in Sanskrit, so I wouldn't take the presence of any causative—even a canonical one in -ayati—as evidence of a PIE causative unless it had parallels in other IE languages. —Aɴɢʀ (talk) 15:54, 28 September 2016 (UTC)

@Angr: I have gotten access to Edgerton's paper (I can send it to you if you'd like). It pertains to the spread and semantic meaning of the suffix, but only in one footnote mentions the source saying: “Its origin is still obscure, despite numerous attempts at explanation, mostly summarized by Ghosh 69 ff (see f. 1).” The paper to which he is alluding is Batakrishna Ghosh's “Les formations nominales et verbales en p du sanskrit.” I do not have access to this work, but this review seems to imply that he drew few useful conclusions. —JohnC5 18:14, 1 October 2016 (UTC)

Well, maybe someone has come up with something in the 70 years since Edgerton's paper, but if so, I couldn't find it after a 3-minute Google search! —Aɴɢʀ (talk) 18:19, 1 October 2016 (UTC)

Yeah, that's why I asked for help: I haven't the faintest idea where to look for this. —JohnC5 18:22, 1 October 2016 (UTC)

Can someone verify the source of the etymology for moonlight? The OED doesn’t mention anything about a surname. In the time being, should I add {{Template:m}} with Middle English language tags for the terms? —britannic124 (talk) 17:39, 28 September 2016 (UTC)

Jeezum Crow, even if it's attested as a surname, that doesn't make the surname the etymology of the common noun. Just remove it. —Aɴɢʀ (talk) 18:05, 28 September 2016 (UTC)

I think it's supposed to be early attestations rather than origin. Makes me wonder: If we have the etymon, do first attestations (which these probably aren't) belong into the etymology or do we only put that into the citations? Korn (talk) 20:07, 28 September 2016 (UTC)

I'd put them in the etymology if they're from an earlier stage of the language (e.g. Middle English) and into the citations if they're from the same stage of the language (e.g. Early Modern English). —Aɴɢʀ (talk) 20:17, 28 September 2016 (UTC)

I found a Middle English attestation of monelight meaning moonlight ("light of the moon"). I've removed all reference to the name, as it's unneeded. Leasnam (talk) 15:20, 29 September 2016 (UTC)

What is the etymology of the verb to gruel (“to exhaust”)? --Fsojic (talk) 11:31, 1 October 2016 (UTC)

The Online Etymology Dictionary says "from late 18c. slang get (or have) one's gruel 'receive one's punishment'". —Aɴɢʀ (talk) 12:01, 1 October 2016 (UTC)

Ok, thanks. --Fsojic (talk) 13:26, 1 October 2016 (UTC)