User talk:JohnC5/2016

Greek Nasal Accent

Hi did you know if there is an nasal accent in Ancient Greek ? I need it pro my varg ? 91.180.225.145 14:39, 15 January 2016 (UTC)

I don't know what you mean by "nasal accent", but the accent represented by acute/circumflex was a pitch accent. Nasalisation of vowels may have occurred, but it was never phonemic. —ObsequiousNewt (εἴρηκα|πεποίηκα) 14:59, 15 January 2016 (UTC)

Are you referring to any word in particular, Nemzag, or generally? I would advise using this copy of the LSJ which provides accents very accurately. —JohnC5 18:00, 15 January 2016 (UTC)

Kanth you ! I don't know who is Nemzag, your page don't talk about accent, and the list is limited to some A terms. 91.180.227.172 22:54, 19 January 2016 (UTC)

It has everything, you just have to use the search bar. A reference table for the search bar encoding may be found at {{R:LSJ}} under the “Transcription characters”. That page does list the accents in the entry of each word. —JohnC5 23:56, 19 January 2016 (UTC)

I'm searching pro a website listing all terms using Qoppa, San, Digamma (both version) and others ancient removed letter, I can't find any in web, do you know something ? 91.180.227.172 09:57, 20 January 2016 (UTC)

{{R:LSJ}} and {{R:DGE}} will have digammata when known. For the others, I do not know. —JohnC5 15:25, 20 January 2016 (UTC)

This is a fair point, actually; English has categories like this. On the other hand, this appears to be manual. Is it possible to put something like this into {{head}}? —ObsequiousNewt (εἴρηκα|πεποίηκα) 15:51, 20 January 2016 (UTC)

@ObsequiousNewt: If we create module:grc-headword, easily. Without that, it would become more difficult and would require you to juryrig something with invocations of the match function in Module:string in the headword templates. —JohnC5 16:16, 20 January 2016 (UTC)

Why would we need to create a separate module? We can easily modify the show_headword_line function in Module:headword. Perhaps add a category in Module:languages/data for a list of 'notable' characters (e.g. "[0-9À-ÿ]", although it would certainly have more characters), then something like for i in mw.ustring.gmatch(<headword>,<list>) do <add category> end. (Alternatively, you could do the reverse, and make a list of characters that are typical, then your function would be the same except the list would be "[^ 'A-Za-z]". This may be easier.) CodeCat, comments?—ObsequiousNewt (εἴρηκα|πεποίηκα) 17:22, 20 January 2016 (UTC)

@ObsequiousNewt: That's fair. Though specifying them in each headword template for a language is a pain. The optimal method, I think, would be to specify notable = "[0-9À-ÿ]", in Module:languages per language and then reference that list in Module:headword. I think it would be nice to have a "notable" category that get categories by character and then a catchall which will get anything outside of the "standard characters" + "notable characters". This would also be useful for debugging. —JohnC5 18:34, 20 January 2016 (UTC)

I hope you realize that you and and ObsequiousNewt suggested the exact same thing. It's a good idea though. --WikiTiki89 22:34, 20 January 2016 (UTC)

@Wikitiki89: Lol, I do now! I skimmed over this earlier and clearly misread everything. Newt shows him/herself to be as prudent and wise as ever; whereas I am always making careless misakes. —JohnC5 23:01, 20 January 2016 (UTC)

@CodeCat, what do you think of this suggestion? —JohnC5 00:55, 25 January 2016 (UTC)

FWIW, I think it's a great idea. — I.S.M.E.T.A. 02:46, 1 February 2016 (UTC)

@I'm so meta even this acronym: so, should we bring this up in the GP? You know how much I hate starting discussions in the main rooms... —JohnC5 03:57, 1 February 2016 (UTC)

Done; see Wiktionary:Grease pit/2016/February#Autopopulation of Category:Terms by their individual characters by language via Template:head. — I.S.M.E.T.A. 23:48, 1 February 2016 (UTC)

I think we will want to have some kind of global list of characters that are not interesting in any language. This would include spaces, punctuation and the like. For individual languages, a list of uninteresting characters would be far more effective than a list of unusual characters, too. —CodeCat 01:00, 25 January 2016 (UTC)

John, what do you think are the chances that Nemzag (who is probably this user) is Wikinger? —ObsequiousNewt (εἴρηκα|πεποίηκα) 22:25, 20 January 2016 (UTC)

So, that is a very interesting question. The use of pro my varg (for my work), Kanth, and generally ungrammatical English are extremely characteristic of Nemzag. While Wikinger does seem interested in AG letter forms, it seems to me that Wikinger is far more sophisticated and intentional in style of abuse compared with Nemzag, whom I would consider just plain crazy. This person may well be Wikinger, if the user's claim to recognize the name Nemzag is to be believed. Wikinger did seem to use similarly broken English to this and claimed to be on a mission (“This is not obsession. This is my Ultracatholic inquisitorial fanatism to fight for preserving of heritage of my Catholic Faith. God obliges me here: [1] in general to fight for His Catholicism, especially against stubborn sinners who perform removals of Catholic Heritage in part or in whole.”). I think we have fairly well established that Nemzag was an Albanian Muslim, which does not jive with Wikinger's narrative. I believe they are distinct people and that the above user is more likely to be Wikinger than Nemzag. —JohnC5 22:56, 20 January 2016 (UTC)

Maybe, but by all appearances Wikinger pretended to be several different people. It's true that Nemzag has had a fairly consistent alibi throughout his time at Wiktionary, though. —ObsequiousNewt (εἴρηκα|πεποίηκα) 01:17, 21 January 2016 (UTC)

I my little JohnC5 & ObsequiousNewt, I'm the son of Ouranos also know as the crasy caesar sartena rovel kami kamli Nemzag, the one who will make your life a real nightmare :) !!! No I can confirm, I'm not that Wikinger and I don't know who he is, hope that he will make you became crazy as I'm you will be surely beddër and less boring... Good dayë. 60.39.58.187 00:10, 21 January 2016 (UTC)

Some comments by Werdna Yrneh Yarg

@Werdna Yrneh Yarg Hello Andrew, I've been following your user page for a while now. I believe I am right in saying that you wish to demonstrate the Celtic words that appear in English. It seems that, in the above statement, you are claiming there are words in English inherited from Celtic as opposed to borrowed from Celtic. This is not the case. English is a Germanic language, and any non-Germanic word (and some Germanic ones, for that matter) must have been borrowed into the language at some point. —JohnC5 17:31, 25 January 2016 (UTC)

@JohnC5 Thank you for your message. It conveys what I always supposed at one time. Recent scientific finds have disproved that, however: I was only advised as to this properly last night, by my ex English post-graduate English lecturer; that except in the East of England, the Angles and Saxons found it relatively easy to settle with the natives due to their tongue being already a from of Germanic akin to Brythonic. So just because a lexeme can be traced to a Germanic root does not in every case mean that that root was not already in Britain! I go by Barber's Story of Language', where it clearly states that the number of Celtic words borrowed into English are like 'a drop in the ocean'. Practically every European language has its small group of older words that have been assimilated into its vocabulary. Both Cornish and Welsh are both Brythonic, as you know; but a number of words have been similated into Welsh from the Silures - the Iberian race that remained after the conquests. If I ask: "Are you an etymologist" (that I would not have the audacity to do), I would be speaking in Germanic (except obviously for the lexeme, 'etymologist'); but if I were to ask: "Do you do a number of edits?", I am no longer using a Germanic idiom, but a Celtic one. The dilution of old Germanic grammar, where Old English once had not only 'neuter' but duals in some of the nouns and adjectives, surely demonstrates in itself the influence of the non-educated element that possibly predominated in the British race and carried forward elements of their own vocubulary. All of this, you would be surely more aware of than I; but my user page is not simply to defend any Celtic element in the English vocabulary, but to be a simple restricted treatise of etymology as a whole. Kind Regards, Andrew H. Gray 13:40, 26 January 2016 (UTC)Andrew

P.S. I have transferred your message from the exhaustive Talk Pages of Leasnam to mine to save room on his page. Andrew H. Gray 13:40, 26 January 2016 (UTC)Andrew

So, I have a few comments, and I by no means wish to offend:

• Brythonic would never have been similar or mutually understandable to Germanic speaking groups like the Angles or Saxons.
• “Do you wish to do a number of edits?” is perfectly normal Germanic sentence structure. I have no idea where you got the idea that it is Celtic.
• Old English grammar did not “disintegrate” by any means. Languages change over time, and the ignorance of the population does not cause any form of degradation. This also presumes that Proto-Germanic grammar is somehow more formal and complex. Languages tend to substitute one form of complexity for another over time.

I am still very confused as to the point you are trying to make and its purpose. Fortunately, you are very friendly and seem to be acting in good faith. —JohnC5 04:34, 27 January 2016 (UTC)

@ Dear John[[User talk:JohnC5|C5] Thank you so much for your reply: to start with, I need to apologise for mis-quoting 'C. L. Barber, on "the Story of Language" - not "History", as I put - and had to correct this in my previous paragraph! Also, what I put about a form of Germanic already existing, I now see was quite misleading. What I should have stated was that recently science has unveiled that when the Germanic tribes moved around Europe they also settled in England, thereby facilitating the settling of conquering Saxons and Angles, (in that order), to understand the language. I will delete the term "disintigration" in accord with your reply. Also, I freely admit that I had real difficulty in believing any similarity between Brythonic (that is of the Celtic branch as you know) and Proto-Germanic.

The point of comparing the Germanic and Celtic idioms, to which you made reference, was to show that due to Celtic being thrifty with its verbs, the repetition of 'do' with a verb, is not found in any other Germanic language that I am aware of, and therefore is of earlier grammatical structure. The Germanic grammatical idiom is "edit you?" - surely not; "Do you edit". In the following sentences you will notice that I have used this bizarre idiom at least twice, when I mean: "it means not" and I believe not", respectively! By 'assmimilation' into Old English, I meant 'absorbed', so if that is what you meant by 'borrowing', your term is quite suitable and is reflected by other authors. It just does not mean that those few lexemes were not in existence when the Germanic languages pervaded parts of Britain - they could have been borrowed from what existed beforehand. I do not believe that there is any proof that the previous vocabulary was totally wiped out: 'tungolcraft' for astronomy, is an example of this. "-craft" is obviously Germanic; but where does one find a similar lexeme to "tungol" (star)? One can only summise its formation as possibly being related to 'don' (to carry, bear) and gol-ow (light), < Proto-Celtic *wlugos. This, among a number of others have become completely lost to English now; but that in itself, with others, like DOWN, DUN, BROCK and COOMB, et cetera, have had no valid proof as to their non-existence before the Saxons, Angles, Jutes and Vikings conquered England. Another familiar lexeme is that of PEAT < Anglo-Latin PETA, already in use at the time of the Germanic invasions. I trust that this clarifies the facts; although, for myself, I have had to learn as much as I can, since a little knowledge of etymologies is like a little knowledge of electricity! Kind Regards. Andrew H. Gray 09:53, 28 January 2016 (UTC)Andrew

"edit you?" is Germanic; and "do you edit?" is also Germanic. Both have the verb first place in interrogative. The roots of the do-periphrastic can be found in the Middle English causitive use of don (e.g. Þe king dēde þe mayden arise "the king made the maiden arise") and in the use of don as an empty, meaningless verb (e.g. He dude writes sende, where writes sende kinda behaves like a verbal noun (i.e. "he did writs-sending")). English is not the only Germanic language to use do as an empty verb: Dutch does also, (though not as elaborately as English, and not in exactly the same way). OE tungol is clearly traced to PGmc *tunglą. Leasnam (talk) 21:17, 1 February 2016 (UTC)

@ Dear Leasnam What you state is instructive and helpful, but still does not provide any effective examples in other Germanic languages to prove the point of being Germanic. It has been established that the Brythonic vocabulary has largely taken over that of the Punic, but the grammar of the latter carried through in some Celtic dialects. So, as you know better than I, the Germanic tribes, before conquering the Dutch, encountered the Celtic race there. Also PGmc *tunglą is simply conjecture as a root of lexemes in other Germanic languages, if no other examples exist; but if in Old English alone, that is quite feasible, but not as from invaders! The Germanic word for STAR is steorra, and correspond with the Common Germanic forms STER-N, et cetera. I am not trying to be awkward here or elsewhere, but simply to arrive at accuracy, in view of what is logically acceptable by the general reader. Unfounded assumptions have been made by some 'etymologists' that are neither true, nor are logically acceptable. Kind Regards. Andrew H. Gray 16:26, 2 February 2016 (talk)

Comments:

• It is Leasnam who responded to you.
• What do you mean that “It has been established that the Brythonic vocabulary has largely taken over that of the Punic”? Punic is not only geographically and temporally distant from Proto-Brythonic but also generally unrelated.
• What do you mean by “the Germanic tribes, before conquering the Dutch”?

I will admit to being more confused than less. —JohnC5 16:38, 2 February 2016 (UTC)

@ Dear John[[User talk:JohnC5|C5] Thank you again for your comments; it is time that I just stop wasting your time! However, regarding Punic, it is totally unrelated to both Brythonic and Goidelic, but cognate - although distant - with the Bronze Age languages between around 1500 BCE and 500 BCE in the British Isles, possibly that of the Parthusians or Milesians that arrived before the Formosians or Formorians, being the latter wave of Celtic invasion, into Ireland. An example of a derivative of that language is Old Cornish KĒR, cognate with Welsh CAER, akin to Punic QERETH (town, city). "Angr" has clarified your explanation about the auxiliary verb todo, and confirms your presentation of this, thank you. My due apologies for ignorance there. The Teutonic peoples, when moving west into Holland met Celts there, so they did not conquer the 'Dutch', as I badly put, but became 'Dutch' - as we say - afterwards. If you did need to reply, could you please do so on my talk page, rather than use up your space with my replies. Thanks in anticipation. Kind Regards. Andrew H. Gray 11:56, 3 February 2016 (UTC)Andrew

yep I think it was a rollback. why asking?

? --Horsesongrassland (talk) 19:42, 28 January 2016 (UTC)

@Horsesongrassland Short answer: we don't really support Altaic theories. Longer answer: see here. —JohnC5 19:49, 28 January 2016 (UTC)

Actually you do (I know what I am talking about), and you are not an academician. so at the end I will have to revert you. --Horsesongrassland (talk) 19:56, 28 January 2016 (UTC)

@Horsesongrassland To be honest, you have no idea whether I am an “academician”, academic, or anything else. Please bring your point up in the Etymology scriptorium before continuing. —JohnC5 21:39, 28 January 2016 (UTC)

Sorry for my last statement. But.. come on Johnny, don't be so diffident :P

https://en.wiktionary.org/wiki/Wiktionary:Votes/2013-11/Proto-Altaic Decision

"Only option 2 is supported, and it has consensus. "Proto-Altaic is allowed in appendices and entries can only link to it." DAVilla 09:18, 31 December 2013 (UTC)" The matter I was working on was not even about Proto-Altaic anyway. So... --Horsesongrassland (talk) 05:58, 29 January 2016 (UTC)

https://en.wiktionary.org/w/index.php?title=Wiktionary:List_of_languages,_csv_format&oldid=25676571

kc_kennylau (talk) 05:30, 11 February 2016 (UTC)

Never mind, I fixed it myself. --kc_kennylau (talk) 05:36, 11 February 2016 (UTC)

Hmmm, what happened? —JohnC5 05:45, 11 February 2016 (UTC)

Try previewing https://en.wiktionary.org/w/index.php?title=Wiktionary:List_of_languages,_csv_format&oldid=25676571 with https://en.wiktionary.org/w/index.php?title=Module:languages/data3/x&oldid=37197005 --kc_kennylau (talk) 07:41, 11 February 2016 (UTC)

Grammatical error in your user page.

I just undid my edit fixing an error on your user page, because it is not okay to edit other people's user pages.

I have corrected a mistake in the following sentenceː "I'm particularly interested in expanding Wiktionary's etymological resources"(the mistake that you've made is in the bold text). Mountebank1 (talk) 00:46, 14 February 2016 (UTC)

@Mountebank1: Yeah, that was just a typo that I had missed. I tend to make them often―much to my annoyance. You may rest assured, however, that I know that particular construction, and I thank you for noticing the error and for taking the time to peruse my userpage. —JohnC5 22:33, 14 February 2016 (UTC)

Romani

Do you think Romani is wide enough to warrant its own family? —CodeCat 03:18, 15 February 2016 (UTC)

@CodeCat Definitely. Should we also do South, Western, Central, Eastern, Northern, and Northwestern Indo-Aryan? —JohnC5 03:21, 15 February 2016 (UTC)

I'm hesitant to use anything that has less than good consensus as a family. Some divisions are less accepted than others. —CodeCat 03:25, 15 February 2016 (UTC)

Southern, Central, and Eastern are all derived from specific Prakrits and thus seem relatively uncontroversial to me, but I don't care too much at this moment. Would you do the honors of creating inc-rom? —JohnC5 03:28, 15 February 2016 (UTC)

Done. —CodeCat 03:42, 15 February 2016 (UTC)

aromantic

You listed it as both synonyms and derived. Which one is it? And can I ask you not to "revert" others' edits, it is unfriendly. Thanks. Ubuntuuser13 (talk) 03:21, 16 February 2016 (UTC)

On second, I had fixed the part where you said it is derived and not synonymy. Ubuntuuser13 (talk) 03:22, 16 February 2016 (UTC)

It is both synonymous and derived. It need not be one or the other. I apologize if reversion seemed rude to you, but I needed to get your attention. I will say that you should not take it as a slight when you are reverted―it is a tool. I have been reverted many times for many reasons, and frequently it helps me become a better editor. Also, I'm actual one of the least revert-happy admins on here (pace my colleagues). Trust me that I know what a revert entails. —JohnC5 03:30, 16 February 2016 (UTC)

No problem. It's OK. You're forgiven. Ubuntuuser13 (talk) 04:05, 16 February 2016 (UTC)

User Gadrian01

This new contributor has been making some good-faith, but poorly-formatted edits to Latin entries. I've cleaned up one or two, but I don't know Latin well enough to judge their content (I have my doubts). I'd appreciate it if you took a look. Thanks! Chuck Entz (talk) 15:06, 22 February 2016 (UTC)

@Chuck Entz: It's quite odd. A bunch of the additions are close enough to be plausible but far enough not to be easily verifiable. The user does seem to be acting in good faith, but I wonder what sources/agenda (s)he is pursuing. —JohnC5 15:52, 22 February 2016 (UTC)

@Chuck Entz, JohnC5: It looks like these senses being added might be being inferred from a parallel translation that Gadrian01 is reading. — I.S.M.E.T.A. 01:40, 23 February 2016 (UTC)

@I'm so meta even this acronym: That seems quite possible. As for cumque, I certainly have seen it used as in cum + -que, when a second clause starts with cum and needs to be adjoined the the preceding clause (sort of like maybe vs. may be). Don't we have a template for an SOP sense of a term? —JohnC5 02:00, 23 February 2016 (UTC)

Something like {{&lit}}, do you mean? I can't find anything pertinent in Category:Definition templates, but there may be one. @Chuck Entz? — I.S.M.E.T.A. 15:06, 23 February 2016 (UTC)

@I'm so meta even this acronym: Yes, something like Used other than figuratively or idiomatically: see cum,‎ -que., though that would fall under a conjunction headword, I think. It's probably not worth keeping, but there are some ambiguous cases where -que cliticizes with words like nē, cum, or quō outside of the normal idiomatic use that should maybe noted under -que. —JohnC5 15:17, 23 February 2016 (UTC)

Agreed. Re cumque, how's this? — I.S.M.E.T.A. 23:22, 23 February 2016 (UTC)

Module:la-noun/data

Regarding the length of -am:

• Indeed, Lewish & Shorts omits the length of the final syllable, and Gaffiot ( micmap.org/dicfro/introduction/gaffiot ) has "Abrăhām, patriare hébreu.". Sorry, and thanks for pointing it out.
  P.S.: Gaffiot also has "Adām, ind., et Adam, dæ, m." So, Gaffiot has both forms. "Jeanneau" (prima-elementa.fr/Dico.htm , which was mentioned at micmap.org/dicfro/ ) also mentions the length of the final syllable, and has "Abraham" and "Ādām, indécl. m. (Ādam, dae, ou Ādāmus, i, m.)". TTL seems to have "Ādām" and "Abraham" (degruyter.com/databasecontent?dbid=tll&dbsource=%2Fdb%2Ftll seems to show the headlines). Georges (zeno.org/Georges-1913) mentions the length of the final syllable too, and has "Abraham" and "Adam". Though maybe Georges omits the length here, as he doesn't know it? That rather looks like the length is doubtful. Wiktionaries and Wikipedias have transcriptions like "ʼAḇrāhām", "Avrāhām" or "avrahám" for the Hebrew name. So maybe the Hebrew is also doubtful.
  Also, could there be both forms, with -ām and -am? For example it could be like this: In older times it was with -ām (like the Hebrew transcriptions ʼAḇrāhām and Avrāhām), and in later times (Middle Ages, Church Latin) it was -am, like in Modern German and French.
• If nominative and vocative are -ām, then how about the accusative? The normal acusative is -am, but here it could be -ām too, couldn't it? Maybe both accusatives, -ām and -am, do exist.
• Depening on the actual length, entries like Abraham#English (which has "Ābraham") might be incorrect.

-Balebatim (talk) 08:54, 24 February 2016 (UTC)

TLL, which we should take as our main authority (since OLD doesn't list names), lists Ābrāhām and rarely Ābrahām in its prosdia section, 129 (and does mention the alternate Ādam). But generally it seems like -ām is the default for this declension. Lengthened final vowels tend to persist in final syllables when they are borrowed (as in {{la-decl-1st-Greek-Ma}}) since the process that shortened them after Old Latin did not have affect later. I would be interested to know whether there are other first declension nouns in -am.

My other issue is that we don't get this specific within the base templates, or else either we create an irregular noun or mention the minor variances in the entry but not in the overall declension. I'm not seeing quite enough evidence to merit both -ām and -am across the board, but -ām seems to be the scholarly standard. Generally, there should be more discussion before changing the declension backend unless the change is completely uncontroversial.

I do appreciate the attention you are paying to these issues, though. —JohnC5 15:26, 24 February 2016 (UTC)

• Nouns in -am are: Adam, Abraham, Abram (another name of Abraham), Mathusalam (also in -em, i.e. Mathusalem), Cham (a son of Noah), also Elam, Aram, and maybe also Adoram (in English Hadoram). Of course there might be more names, but in Genesis, for example, many names are just mentioned a few times or even just once (as Meshech in "The descendants of Japheth: Gomer, Magog, Madai, Javan, Tubal, Meshech, and Tiras.").
• Nouns with -ās, -ae, m. are older than nouns in -am. So names like Ionas/Jonas could have retained the long a, while -am was introduced with a short a.
  After looking at "Ādām, indécl. m. (Ādam, dae .." again, it could also be like this: When the noun is indeclinable then it's -ām, but when it's declined then it's -am like the Latin accusative. So both would be true: The name retained the long final syllable (and the Greek declension of Ἀδάμ, ὁ indecl.), while it was also used with a short final syllable.
  PS: Putting that into a table to sum it up:

Case	"Ādām, indecl. m." like "Ἀδάμ, ὁ indecl."	"Ādam, ae, m."
Nom. Acc. Voc.	-ām	-am
Abl.		-am or -ā
Gen. Dat.		-ae

• The TTL has "nomina hebraica quae nominativo et accusative et vocativo et ablativo in -am terminantur [...], sed ideo primae declinationis sunt propter genetivum et dativum in -ae". So, it should also the support the ablative -am/-ām. The TTL later has "prosodia: Ābrāhām [...] gen. Ābrāhāē [...] Ǎbrǎhām". Combing the two statements, it could be -ām in nominative, accusative, vocative and ablative, that is it's also -ām in the accusative. But, as one can also find the ablative in -a (actual spelling) or -ā (with length), which earlier was the only mentioned ablative here, there could be an accusative in -am (short). That is, some people declined it like a normal Latin word except the nominative and vocative.
  PS: After some further thinking, I guess nominative -ām and accusative -am could sound awkward. So maybe there is -ām, -ae, m. (acc. and abl. -ām) and -am, -ae, m. (acc. -am, abl. -ā). As for the dictionaries: Latin dictionaries are often restricted to ancient Latin and do not usually include medieval and modern Latin. So the declension -am, -ae, m. (acc. -am, abl. -ā) could be modern Latin.
  I've just look into the Vulgata (Wikisoruce, Vatican, Nova Vulgata and latinvulgate.com) again. In Genesis 17:22 it is "ab Abraham" (in English "from Abraham") and in Genesis 21:5 it is "Cum Abraham", while "Abrahae" occurs in Gen. 19, 21, 22, 23, 25. So the more original ablative could be in -am and that might be the more common form (like the Grammar book only had that form). Then the ablative -ā should rather be an error (prescriptively seen), or a younger (maybe even New Latin) form. That is, it could be like this:

Case	Vulgata, classical, common	younger [maybe New Latin]
Nom. Acc. Voc.	-ām	-am
Abl.	-ām	-ā
Gen. Dat.	-ae	-ae

and NOT like this which I mentioned after looking into the TTL but before the PS:

Case	(a)	(b)
Nom. Voc.	-ām	-ām
Acc.		-am
Abl.		-ā
Gen. Dat.	-ae	-ae

Though, there could be a third form: The younger form with -am (short) could also have ablative -am. Like using the Vulgata declension in written form, but younger (maybe New Latin, maybe also French or German) length for -am. But I'd rather reject that too.

• As for discussions: I first tried to discuss the issue at WT:ID#Moses, and Abraham (Latin). Later, after my change at the module got reverted, I tried to discuss it again by mentioning the discussion page. Then one could have reverted it with a comment like "Let's first discuss it; I'm gonna reply at the discussion". Then I would have waited, and tried to discuss it there too. What posibilities were there?
  • I could have added another comment at WT:ID#Moses, and Abraham (Latin). But then one maybe wouldn't see it, or wouldn't reply there.
  • I could have posted here earlier. But I didn't want to repeat myself (except if my text wasn't understable). So I only would have posted a link to the discussion. But IMHO it was easier and faster to post a link to the discussion in the edit summary. If there were doubts to my change, one could have reverted it again AND replied to the discussion.
    • Also another reason for posting the link in the edit summary and not on a user talk page, could be this: I got the impression that some Wikipedia/Wiktionary users and even adminstrators are impolite, and dislike (or even avoid) disussions. For example, some admins revert good-faith edits without giving s reason, some admins block IP users who made good-faith edits without talking to them, and some reply in a very impolite (or sometimes IMHO rather insulting) way. So subconscious I might have prefered reverting too.

Anyway, two persons discussed the declension via edit summary.

• Maybe you can also take a look at
  • WT:ID#pascha (Latin) (especially whether it's accusative -am or -a)
  • WT:ID#Iesus (Latin) (whether it's pronounced with i (three syllables) or j (in English often spelled y; two syllables).

-- Greetings Balebatim (talk) 08:43, 25 February 2016 (UTC)

@JohnC5, Balebatim: Bear in mind the possibility of hybrid declension: It is common for an author to use a word in an indeclinable-seeming form in the casus rectus whilst, in the same text, using regularly-declinable forms in the casus obliqui; for example, I have read a short text which used Bēthlehem in the nominative, but Bēthlehēmum in the accusative. — I.S.M.E.T.A. 13:43, 25 February 2016 (UTC)

@I'm so meta even this acronym: That's a very good point. Thanks for making it. @Balebatim: I'm afraid, again you are going into waaaaaaaaaaaay too much detail for the general template. You may (and should) add a summary of these details to the appropriate pages for Adam and Abraham, etc. and probably copy this discussion on their discussion pages, but very little if any of this should make it into the template. Latin authors disagree about declension for every word imaginable, and when you start talking about Vulgar and Ecclesiatical Latin and terms borrowed from other languages, all bets are off. In truth, all the Latin inflection templates (prefix la-) were written for Classical Latin only. There has been some attempts recently to make Vulgar Latin inflections, but there is some serious disagreement about even which cases to include. More generally, there are some Wiktionarians who may seem rude; though on the whole, they deal with a high volumes of vandalism and need to be glib. When I reverted you the first time, though, we should have had a proper discussion, and not editwarred, as editwarring helps no one. Admittedly, I am somewhat to blame here for not posting on your talk page. The other thing, and I know this is annoying, is that Wiktionarians tend to be leery of letting new or unknown editors edit templates or more particularly modules. I'm sorry for my scepticism, but people come and add POV nonsense all the time. —JohnC5 16:14, 25 February 2016 (UTC)

 @I.S.M.E.T.A.: Yes, such usages do occur.

@JohnC5: Vulgar Latin could be different from Middle Latin and New latin. New Latin has all the cases that classical Latin has. Vulgar Latin maybe lost some cases on it's way of becoming the Romance languages.
The ablative "Ābrahā" might be non-classical too (cf. the quoted TTL statement "nominativo et accusative et vocativo et ablativo in -am"). So I guess the ablative "Ābrahā" should be removed, if one ignores non-classical forms. And the accusative might than rather be "Ābrahām" too instead of "Ābraham" (which is the current form in Abraham#Inflection).

-Balebatim (talk) 20:25, 27 February 2016 (UTC)

sicken

How do you write ‘the rain will make you ill’ in Latin? --Romanophile ♞ (contributions) 18:57, 24 February 2016 (UTC)

Interesting question.

“(The) rain will make you sick.”:

• Pluvia/Imber tē aegrum faciet. masc.sg.
• Pluvia/Imber tē aegram faciet. fem.sg.
• Pluvia/Imber vōs aegrōs faciet. masc.pl.
• Pluvia/Imber vōs aegrās faciet. fem.pl.

“You will become sick because of (the) rain.”:

• Pluviae/Imbris causā aegrēscēs. masc./fem.sg.
• Pluviae/Imbris causā aegrēscētis. masc./fem.pl.

Use pluvia for light rain and imber for heavy rain. This is assuming you mean real sickness, as opposed to seasickness/nausea. —JohnC5 20:51, 24 February 2016 (UTC)

Doesn’t Latin have an entire verb that means ‘to make (somebody) sick?’ Literally speaking. --Romanophile ♞ (contributions) 21:09, 24 February 2016 (UTC)

nauseō is “make sick” as in nauseated/disgusted. As for “make sick” as in an ailment, I would guess *aegrificiō, but no such verb exists. infirmō is to “weaken, make infirm”, but not sick. —JohnC5 21:22, 24 February 2016 (UTC)

Oh well. Gratias tibi ago pro tentare (or however it should be said). --Romanophile ♞ (contributions) 21:27, 24 February 2016 (UTC)

Apparently it's something like: Gratias tibi ago, quod tentavisti, “I give thanks to you because you tried.” That sounds really depressing, though. I think you could also squeak by with something like Gratias tibi ago pro tentando. —JohnC5 21:47, 24 February 2016 (UTC)

Venisne ex imperio romano? --Romanophile ♞ (contributions) 21:52, 24 February 2016 (UTC)

Utinam, utinam, mi amice. —JohnC5 21:57, 24 February 2016 (UTC)

@Romanophile, JohnC5: Pluviā aegrēscēs. --kc_kennylau (talk) 14:14, 25 February 2016 (UTC)

@Romanophile, JohnC5: Pluviā aegrēbis. (Sorry, I just have an OCD of making things as short as possible) --kc_kennylau (talk) 14:24, 25 February 2016 (UTC)

@Kc kennylau Yeah, I thought about an ablative of means, but I wanted to make it as explicit as possible. Thanks for the input, though. —JohnC5 15:43, 25 February 2016 (UTC)

Would contamino suffice as a translation? --Romanophile ♞ (contributions) 18:48, 28 February 2016 (UTC)

Figuratively, I suppose, but not literally. —JohnC5 18:57, 28 February 2016 (UTC)

Quomodo dicuntur download uploadque? --Romanophile ♞ (contributions) 19:00, 28 February 2016 (UTC)

@Romanophile: Secundum The New College Latin and English Dictionary, “to download” est “ex rētī prehendere” vel “ex rētī exprōmere”. Ratiōne illā, suādeō “in rēte pōnere” vel aliquid simile prō “to upload”. —JohnC5 19:26, 28 February 2016 (UTC)

smoothie quoque? --Romanophile ♞ (contributions) 21:05, 28 February 2016 (UTC)

@Romanophile: Fortasse “pōculum fructuum commixtōrum”? —JohnC5 21:28, 28 February 2016 (UTC)

@JohnC5: Comprêndés sermoné vulgaré, etiam proto-romanicu? --kc_kennylau (talk) 10:33, 29 February 2016 (UTC)

@Kc kennylau: Éo comprêndo, sed non sapjo sí posso illu scríbere. Habés tu referentja por mẹ? —JohnC5 16:35, 29 February 2016 (UTC)

@JohnC5: Nõ avjo multas referentjas; solã-mentẽ scrivo como volo. --kc_kennylau (talk) 16:41, 29 February 2016 (UTC)

@JohnC5: Mjũ comprênsionẽ de illã evolutjonẽ de Latinã initjav quando quese quomodo illũ verbũ "habere" en Latinã potest devenire "avoir" en linguã Francogallicã. --kc_kennylau (talk) 16:58, 29 February 2016 (UTC)

@JohnC5: Per esemplũ, potes videre illũ paravlã abreuver, ubi scrivav illã evolutjonẽ de-llã paravlã sine referentjas. --kc_kennylau (talk) 17:14, 29 February 2016 (UTC)

@Kc kennylau: Grasias! —JohnC5 17:22, 29 February 2016 (UTC)

@JohnC5: Cũ plakʲere. --kc_kennylau (talk) 17:26, 29 February 2016 (UTC)

Quomodo dicuntur «buvons à sa santé»? --Romanophile ♞ (contributions) 00:59, 9 March 2016 (UTC)

complurēs

Vēritāsne coniugātiō? Venitne adiungente "com" "plurēs"? --kc_kennylau (talk) 13:50, 29 February 2016 (UTC)

Secundum L&S, dēclīnātiō neutra plūrālis “complura” in Latīnitāte Classicā facienda est sed quoque “compluria” in Latīnitāte Anteclassicā. —JohnC5 16:17, 29 February 2016 (UTC)

breviseta

Declinatio prima fit tertia? Affirmāturne? --kc_kennylau (talk) 14:35, 1 March 2016 (UTC)

@Kc kennylau: Nesciō triste. @I'm so meta even this acronym, suntne sententiae tibi? —JohnC5 16:00, 1 March 2016 (UTC)

@kc_kennylau, JohnC5: Cogito duo lemmata esse, viꝫ brevisētus, -a, -um et brevisētis, -e. — I.S.M.E.T.A. 19:35, 2 March 2016 (UTC)

@I'm so meta even this acronym: Fontem habes? --kc_kennylau (talk) 13:19, 3 March 2016 (UTC)

@kc_kennylau: Nihil, praeterquam ingenium quaerens Librorum Gugulae. — I.S.M.E.T.A. 13:51, 3 March 2016 (UTC)

@I'm so meta even this acronym: Sic omnia mutamus? --kc_kennylau (talk) 07:27, 6 March 2016 (UTC)

Mutanda sunt. — I.S.M.E.T.A. 22:46, 6 March 2016 (UTC)

T:de-verb

Are we still manually converting those? --kc_kennylau (talk) 07:21, 6 March 2016 (UTC)

@Kc kennylau Yeah, though slowly. Is there anyone else we can conscript to help us? I only go so quickly. —JohnC5 07:31, 6 March 2016 (UTC)

Genus and species names

What genders are they actually? Do they follow genders at all? Please kindly go through all the names in w:Bouteloua and kindly provide your analysis. --kc_kennylau (talk) 07:33, 6 March 2016 (UTC)

@Kc kennylau, Pengo, I'm so meta even this acronym. Those look femin-ish to me. The species normally agrees with the genus in my experience, but there are many exceptions because scientists are not as good at pseudo-Greco-hogwash-Latin as they should be. Some of those (ending in -ii) are faux genitives and do not agree with the genus. I dunno. Our policy for Translingual ~ New Latin needs some work. —JohnC5 07:41, 6 March 2016 (UTC)

They look like neuter plural instead of feminine to me. At least animalia is proven to be neuter plural. There are also some -es and -ii suggestive of plural, although -ens slips in sometimes. --kc_kennylau (talk) 07:45, 6 March 2016 (UTC)

Fair point. I'm a bit too tired at the moment anyway. It is an interesting if intractable question, nonetheless. —JohnC5 07:49, 6 March 2016 (UTC)

The rule is that specific epithets are either adjectives modifying and agreeing in gender/number with the generic names, nouns in the genitive, or nouns in the nominative in apposition (the nouns never agree with the specific epithet). Some scientists prefer not to change the gender of the specific epithet when a species is moved to a new genus, and others make mistakes like assuming neuters ending with -a are feminines, so there's lots of variation in gender for specific epithets. As for Bouteloua, the list of specific epithets is what one would expect from a feminine generic name. The specific epithets ending in -ii or -orum are genitives, which agree with their (male) referents, not with the generic name (female referents would mean -ae or -arum endings). Specific epithets ending in -oides could be either masculine or feminine, depending on the gender of the generic name. Chuck Entz (talk) 08:21, 6 March 2016 (UTC)

@kc_kennylau: I'm sure Chuck Entz's explanation has clarified matters. Are there any specific binominals that his analysis doesn't cover? — I.S.M.E.T.A. 22:26, 6 March 2016 (UTC)

@Chuck Entz, I'm so meta even this acronym: Vōbīs grātiās agō. --kc_kennylau (talk) 10:45, 7 March 2016 (UTC)

Praefīxa

Sapiēns, doceās me via praefīxōrum: cur verba significantiārum eārundem praefīxīs disparibus ūtuntur? ad&visō = sub&gerō = prō&videō? Praefīxa quae significant? --kc_kennylau (talk) 17:02, 7 March 2016 (UTC)

@Kc kennylau: CCTPT (Correxī commentārium tuum prō tē) Interdum fōrmātiō verbōrum ratiōne eget. Utinam explicātiō melior mihi foret. —JohnC5 19:21, 7 March 2016 (UTC)

Latin word formation

Is it correct, that if a prefix is added to a word before Proto-Italic, then the first syllable of the stem must have its vowel changed to "i" or "u" (making the vowel higher)? For example:

ad + iaciō = adiciō

ab + teneō = abstineō

bis +‎ capitis = bicipitis (although the nominative doesn't follow this rule)

cum + legō = colligō

ab + caedō = abscīdō

ad + causa = accūsō

--kc_kennylau (talk) 12:18, 11 March 2016 (UTC)

Cribbing mostly from Fortson, I'll summarize the rules of vowel change from Proto-Italic to Latin. First of all, it is important to understand that Proto-Italic and Old Latin had stress on initial syllables (until it became the w:Dreimorengesetz in Latin). This resulted in weakening of vowel post-tonically (viz. in non-initial syllables).

• Short vowels:
  • In open syllables weakened to i (amīcus > inimīcus, legō > colligō, locus > īlicō).
  • In closed syllables:
    • a > e (factus > affectus, aptus > ineptus)
    • o > u (PItc *onostos > onustus)
  • Also, o > u in almost all non-initial syllables.
  • e raise to i sometimes before a nasal (OL en > in)
  • In final syllables, same as above, except:
    • e before -s / -t > i (*leget > legit, *leges > legis)
• Long vowels:
  • In final syllables shorted in late Old Latin before ever consonant except -s. (*dūcām > dūcam, *dūcāt > dūcat, but dūcās)
• Diphthongs:
  • ai > ae†
  • au left unchanged
  • ei > ī (dative *patrei > patrī)
  • oi:
    • Sometimes oe† (foedus, moenia)
    • Sometimes ī (OL meois sokiois > meīs sociīs)
    • Sometimes ū (OL oinom > ūnum)
  • eu merged with ou in Italic. OL ou > ū (PIE *h₁ews- > PItc *ourō > ūrō, OL iouxmenta > iūmenta)

†Merely a spelling convention change. No phonological shift.

This is by no means exhaustive, but I hope it will help. —JohnC5 18:42, 11 March 2016 (UTC)

@Kc kennylau, secondary ping so that you actually get pinged. —JohnC5 18:45, 11 March 2016 (UTC)

Regarding ai > ae (and oi > oe), why would the spelling have changed if there were no phonological shift? --WikiTiki89 18:50, 11 March 2016 (UTC)

@Wikitiki89: Beats me, but it's pretty well accepted that OL ai and oi became L ae and oe respectively with little to no known phonetic shift. —JohnC5 18:53, 11 March 2016 (UTC)

I would postulate that unless there is evidence to the contrary, we can assume that the second part of these diphthongs had become noticeably laxer or lower. --WikiTiki89 19:03, 11 March 2016 (UTC)

@Wikitiki89: That would both make sense and be consistent with the output of {{la-IPA}}. —JohnC5 19:07, 11 March 2016 (UTC)

John, the diphthongs part is a bit inaccurate because it seems to be about stressed syllables, while the question was about unstressed syllables. For unstressed syllables:

• ae > ī (abscīdō)
• au > ū (accūsō)
• oe > ī (rare in verbs, but the o-stem nominative plural ending -ī is an example of this)

I hope this helps. —CodeCat 19:42, 11 March 2016 (UTC)

If *leget became legit, then where does leget come from? *legēt? --kc_kennylau (talk) 23:57, 11 March 2016 (UTC)

@Kc kennylau: Yep! @CodeCat: Thanks fo the point. I've whipped up this handy dandy table based on someone else's and a deep dive into Sihler. Would you check it out? Should we add it to WT:AITC, maybe? @Wikitiki89: In reading Sihler, he supports your postulation that the respelling of ai and oi in the early 2nd century BC represents a lowering in the second element of the diphthong. —JohnC5 04:10, 12 March 2016 (UTC)

faliscus

Scisne falisce loqui? --Romanophile ♞ (contributions) 01:37, 19 March 2016 (UTC)

@Romanophile: Bene satis. Cur quaeris? —JohnC5 06:51, 19 March 2016 (UTC)

Sapisne jam quomodo dicebat «quando»? --Romanophile ♞ (contributions) 08:48, 19 March 2016 (UTC)

@Romanophile: Nesciō īnfēlīciter. —JohnC5 18:17, 19 March 2016 (UTC)

-u- perfects

It seems to me, that Category:Latin verbs with perfect in -av-‎, Category:Latin verbs with perfect in -ev-‎, Category:Latin verbs with perfect in -iv-‎, and Category:Latin verbs with perfect in -u-‎, have the same source. What would be the PIE and Proto-Italic forms of those perfects? --kc_kennylau (talk) 05:26, 19 March 2016 (UTC)

@Kc kennylau: Fortson and Sihler both say that the Latin -v- stems are not Italic but a variously/dubiously explained Latin phenomenon. Sihler treats it fairly extensively in §528-529, if you have access to it. —JohnC5 07:16, 19 March 2016 (UTC)

@JohnC5: Wikipedia briefly mentions that it comes from PIE -wos, and would you like to write down the derivation as argued by Sihler? --kc_kennylau (talk) 08:41, 19 March 2016 (UTC)

@Kc kennylau: Was Sihler useful in answering your question? —JohnC5 03:32, 20 March 2016 (UTC)

PIE *bʰleh₁-eh₂ > PI *flēai > PI *flēwai > flēvī

PIE *ser-h₂-eh₂ > PI *seraai > PI *serawai > OL *serewai > *seruwai > seruī

--kc_kennylau (talk) 06:30, 20 March 2016 (UTC)

@Kc kennylau: Do we believe that *w insertion was in PI or post-PI? —JohnC5 06:48, 20 March 2016 (UTC)

Scrībit ut in PI fuit. --kc_kennylau (talk) 06:52, 20 March 2016 (UTC)

@Kc kennylau: Sihler scrībit “These are a type particular to L―they are not even Italic.” Mihi vidētur *w-suffīxum sērius appāruisse. —JohnC5 07:09, 20 March 2016 (UTC)

Quōd modōd deiketor "I see"? --kc_kennylau (talk) 07:20, 20 March 2016 (UTC)

Quōr nōn “videō” vel “intellegō”?

-cipes

In your table, nothing of the sequence Vs can develop into es, but -cipes clearly came from Proto-Italic. Then where did it come from? --kc_kennylau (talk) 10:13, 19 March 2016 (UTC)

I've changed it to a {{rfv-inflection}}. --kc_kennylau (talk) 11:05, 19 March 2016 (UTC)

@Kc kennylau: are my changes satisfactory? —JohnC5 18:41, 19 March 2016 (UTC)

sanctā•māter•deiwī --kc_kennylau (talk) 00:12, 20 March 2016 (UTC)

usus

Where does ūsus come from? --kc_kennylau (talk) 16:34, 19 March 2016 (UTC)

@Kc kennylau: De Vaan says the PI is *oissos, which would from PIE *tt > *tst > PI *ss. If we propose a PIE verbal adjective, you'd expect something like PIE *h₃ittós > PIE *issos > L *issus, which is clearly not the case. According to Sihler, PI geminate *ss > L s after a consonant, diphthong, or long vowel. So I think we have a PI leveling of PI *oitōr to *oittos > *oissos. —JohnC5 18:40, 19 March 2016 (UTC)

Sed PIE bhit-s-tós facit PI fitsos? --kc_kennylau (talk) 00:15, 20 March 2016 (UTC)

PIE *bhidtós > *bhid-s-tós > PI **fissos > L **fissus. Sed quoque cum leveling, Pi *feissos > fīsus. —JohnC5 00:38, 20 March 2016 (UTC)

Vergivius

Can you provide a citation for vergivius, and especially its etymology? It's not in L&S, and it seems entirely backwards to say it came from Greek. Which is not to say it couldn't have been from Celtic, but I actually can't find (in Beekes) any examples of Greek vocabulary being borrowed directly from Celtic, and -ivius is an established Latin suffix. —ObsequiousNewt (εἴρηκα|πεποίηκα) 15:24, 22 March 2016 (UTC)

@ObsequiousNewt: Interesting. I can't seem to remember the reference I used the first time. Here, however, is a similar-ish reference. —JohnC5 00:33, 23 March 2016 (UTC)

Early Proto-Italic *CujV

What would it develop to? --kc_kennylau (talk) 09:42, 23 March 2016 (UTC) kc_kennylau (talk) 09:42, 23 March 2016 (UTC)

@Kc kennylau: /j/ is always lost intervocalically (VjV > VV). Like vowels contract to a long vowels (so Cuju > Cū), but /i/ and /u/ do not contract before a following unlike vowel. Therefore, Cuj(ā̆|ē̆|ī̆|ō̆) > Cu(ā̆|ē̆|ī̆|ō̆). Why do you ask? —JohnC5 03:15, 24 March 2016 (UTC)

Could it have assimilated to become *CyjV > *CijV > *CīV? --kc_kennylau (talk) 10:19, 24 March 2016 (UTC)

I do not think so. Particularly since intervocalic PIE *-y- became *-j- and disappeared before PI. What is your example? —JohnC5 12:57, 24 March 2016 (UTC)

Could you provide any example for *CujV > *CuV? Could the "u" have been assimilated before PI? --kc_kennylau (talk) 13:08, 24 March 2016 (UTC)

@Kc kennylau: The only example that is close-ish is L clueō < PI *kluēō < PIE *ḱlu-éh₁-(ye)- < *ḱlew-. It doesn't have the *-y/j- but it does have *-uV-. I feel like I could be more helpful if I knew why you were asking. —JohnC5 21:09, 24 March 2016 (UTC)

*bhu-ye-h3 > *φiiō > *fīō. --kc_kennylau (talk) 23:08, 24 March 2016 (UTC)

@Kc kennylau: It seems like De Vaan walks though the possibilities for this fairly well. —JohnC5 05:04, 25 March 2016 (UTC)

sēnsus

Where does sēnsus come from? --kc_kennylau (talk) 13:22, 23 March 2016 (UTC)

@Kc kennylau: PIE **sn̥t-tós? > *sn̥tstós > PI *senssos > L sēnsus (with characteristic V → Vː / _N[fricative]). —JohnC5 03:06, 24 March 2016 (UTC)

But it must have come from *senssus... --kc_kennylau (talk) 10:17, 24 March 2016 (UTC)

@Kc kennylau: Sorry, I wasn't paying attention. PIE **sént-tus > late-PIE *sn̥tstus > PI *senssus > L sēnsus. —JohnC5 13:04, 24 March 2016 (UTC)

What is *-tus? --kc_kennylau (talk) 13:07, 24 March 2016 (UTC)

@Kc kennylau: L -tus (Etymology 2) from PIE *-tus). It is closely related to *-tis. Both show root zero-grades in virtually all reflexes. —JohnC5 15:35, 24 March 2016 (UTC)

ῥοδοδάκτυλος‎

Hey, I know that ῥοδοδάκτυλος theoretically has a neuter form ῥοδοδάκτυλον and all those other inflected forms, but they don't actually exist because ῥοδοδάκτυλος‎ is only used in one phrase at the end of a line in Homer: ῥοδοδάκτυλος Ἠώς. The inflection table says "not all forms are necessarily attested", but in this case to be accurate it ought to say "Only the feminine nominative singular is actually attested, and the other forms are only hypothetical". I would prefer instead not to have an inflection table and neuter form at all, and to simply give the one form that actually exists.

(Addendum: Oops, it does occur outside of Homer, but still only in the f. nom. sg.) — Eru·tuon 04:40, 24 March 2016 (UTC)

@Erutuon: That makes sense, I guess. The article looks like a stub though. Also, I like having the other forms for if someone wanted to compose some Greek for themselves. I'd prefer to leave a note saying that it is only attested in one form but give it a full treatment. @ObsequiousNewt, what do you think? —JohnC5 04:45, 24 March 2016 (UTC)

If we can infer and disambiguate the full inflection from the attested forms, I don't see why the full paradigm shouldn't be given. We do this with countless other entries. —CodeCat 23:11, 24 March 2016 (UTC)

@CodeCat Never responded. The only form attested is ῥοδοδάκτυλος‎ (rhododáktulos‎), but yes, that does unequivocally point to second declension. I know it's the practice to give all the hypothetical declined forms, no matter how many of them actually exist, but I think it's bizarre in this case. For words with highly restricted usage like this, I'd rather give only the form that is attested (and only list the attested gender in the head line), or mark all unattested forms with asterisks and unlink them so that nobody ever creates entries for never-used forms.

@JohnC5 I'm not sure why anyone would ever need the oblique singular feminine forms of "rose-fingered" in a Greek composition, unless they were composing an extended hymn of praise to the dawn, or trying to invent more and more syntactically convoluted ways to introduce new days in an epic poem. It's even harder to imagine how they could use the other genders, and the plural, since the adjective only ever modifies a singular feminine noun. — Eru·tuon 03:30, 5 April 2016 (UTC)

@Erutuon Meh, whatever. My gut says just let it have its table for the lulz, but do whatever you think is best. —JohnC5 03:44, 5 April 2016 (UTC)

@Erutuon: This word may be a hapax, but it is an entirely regular formation (exocentric compound). Additionally, simply because you can't think of a reason for a plural to be used doesn't mean it shouldn't exist. It is possible that a poet would have wished to say ῥοδοδάκτυλοι κόραι "rosy-fingered maidens" (especially given that the singular does exist), and this form just does not survive. If we only supply the singular, that seems to imply that there is some special grammatical or semantic property of this word, which is simply not the case. —ObsequiousNewt (εἴρηκα|πεποίηκα) 16:02, 5 April 2016 (UTC)

@ObsequiousNewt: Yeah, nothing preventing there from hypothetically being a feminine plural in one of the many Ancient Greek works that haven't survived. But I'd rather not display the dual and plural, since if they're displayed in a table entries will be created, and we shouldn't have entries for forms that aren't attested. Meh. Really this line of reasoning means either no inflections should be displayed, or I abandon the line of reasoning and agree with displaying the feminine forms of all numbers. — Eru·tuon 16:32, 6 April 2016 (UTC)

Unless we're dealing with something like ἐΰς, all forms should be displayed. Ancient Greek is not Hittite or Old Irish; it is easy to infer an entire conjugation from only one form. —ObsequiousNewt (εἴρηκα|πεποίηκα) 18:30, 6 April 2016 (UTC)

@ObsequiousNewt: So you're not concerning with, in future, a bot creating pages for inflected forms that readers will never encounter in a text? — Eru·tuon 18:34, 6 April 2016 (UTC)

@Erutuon: It's done for Latin. —ObsequiousNewt (εἴρηκα|πεποίηκα) 18:42, 6 April 2016 (UTC)

@Erutuon: Yeah, our general policy is that if we can infer the full inflection from the data given (even for hapax legomena) we give the full inflection. I think we should stick with this policy. Does it give forms that may have never existed? Yes, but so does the Perseus inflection generator. —JohnC5 19:24, 6 April 2016 (UTC)

@JohnC5, ObsequiousNewt: Well, I'd love to change that policy for rarer words at least (though certainly I can see it would require much more work to only display forms that actually exist). So, in the case of this word, how many forms does the policy require to be displayed? — Eru·tuon 19:35, 6 April 2016 (UTC) — Eru·tuon 19:46, 6 April 2016 (UTC)

Or short of policy change, perhaps there could be a way to (optionally) mark the forms in inflection tables that are known to be attested. That would be easiest for rarer words. — Eru·tuon 20:02, 6 April 2016 (UTC)

There is already a way to mark words that are only attested once: {{label|grc|hapax legomenon}}. —ObsequiousNewt (εἴρηκα|πεποίηκα) 21:16, 6 April 2016 (UTC)

That's not what I mean. That's for lexical items, not inflected forms. — Eru·tuon 21:47, 6 April 2016 (UTC)

How many terms are there in Ancient Greek that are attested for every form in the inflection tables? I doubt there are that many. Chuck Entz (talk) 01:29, 7 April 2016 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ Probably not that many. But there are quite a few words that are only attested in a few inflected forms. And for some of those, it wouldn't be too hard to figure out which forms are actually used. So perhaps some entries could have unattested forms marked if an editor wants to take the trouble of figuring out which forms are attested. I would be interested in doing that for Homeric epithets, such as the one discussed here.

Anyway, I have an idea (see here for what it looks like): suppose attested inflected forms were coded by |attested=s[n], indicating that only the nominative singular is attested, or unattested forms by |attested=~s[gdav]/d[ngdav]/p[ngdav], using ~ to indicate negation, as with dialect codes in {{grc-decl}} and {{grc-adecl}}. Then the module could display unattested forms in gray without links, and the default note "Not all forms ... are attested" would be replaced by a note explaining that the grayed forms are unattested. Then anyone who wants to compose hymns to the Dawn, or poems with groups of rosy-fingered girls, could find the inflected forms they need, and anyone who is curious which forms of the word happen to actually be used in Ancient Greek literature (for whatever reason) could find that out too. — Eru·tuon 07:56, 7 April 2016 (UTC)

If there is one attested form, the word can be marked as a hapax; and the form should be cited in the definition (as all words are.) If there are multiple forms, it's already obvious whether the word was used freely. If ῥοδοδάκτυλος were attested several times, but only with Ἡώς, I would find it appropriate to mark the word with "Only attested in the set phrase ῥοδοδάκτυλος Ἡώς (rhododáktulos Hēṓs, "rosy-fingered dawn")." As it is, however, it is attested with other words, showing that it could be used freely (albeit only poetically). Note also that there is in fact a small note that appears on every inflection table that states, and I quote,

Not all forms, especially dialectal forms, are necessarily attested. Use with caution.

It is not our job to provide information about every citation of a word, unless of course that word is a hapax (or perhaps a dis or tris). This word alone has forty-four citations in Perseus, and potentially more not in Perseus. I find, therefore, that a general disclaimer is much more efficient—and accurate—than attempting to track down and display every possible citation, an endeavour bound to fail. —ObsequiousNewt (εἴρηκα|πεποίηκα) 17:44, 7 April 2016 (UTC)

@ObsequiousNewt: Thanks for the in-depth response. I took a look at the occurrences on Perseus, and have to conclude you're right. About half are from the Iliad and Odyssey, but there are authors or works that are not listed in the LSJ entry, and two occurrences of the accusative, making my table above inaccurate, since it grayed out that form. So it makes sense to list and link all forms in case there are even obscurer works not on Perseus that use other forms of the word. — Eru·tuon 21:40, 7 April 2016 (UTC)

*fakiō or *fakjō?

Where does the 3rd-IO verbs come from? Sihler says *-jō while De Vaan says some from *-iō and some from *-jō. --kc_kennylau (talk) 12:37, 24 March 2016 (UTC)

Appendix:Latin fourth declension and Appendix:Latin fifth declension

You're in error. But what's the reason for the revert? -84.161.32.66 23:40, 24 March 2016 (UTC)

I reverted you for a few reasons. For one, all of your sources are wildly out of date and have been superseded. Secondly, we have many Latin editors on Wiktionary whom I trust more than you. Finally, your comments thus far have been rather rude. —JohnC5 23:52, 24 March 2016 (UTC)

That "You're in error." might have been rude. But the the reverting was rude too and asked me to tell you that you're in error.

As for 4th declension: I provided sources. If you think it's differently, then please provide other sources. "Sympathy" for or against users or "rangs" of users aren't sources.

As for 5th declension: You reinserted the locative. So please proof that it is correct. To proof it, there has to be a Latin word of the 5th declension with a locative. AFAIK no such noun exists. -84.161.32.66 00:02, 25 March 2016 (UTC)

The rudeness to which I was alluding was not the phrase “You're in error”. It was the tenor of your dismissive and clearly disgusted edit summaries.

In truth the OLD recognizes only nominative and accusative (sometimes in -ōn) uses for Argo, echo, Calypso, and Io. It is only for Sappho, Clio, and Dido that the OLD finds genitives in -ūs. Giving them a unified declension does not seem to be a good idea anyway. Given Dido is the only word for which we have a dative (that I can find) and Dido is clearly declined sometimes as a normal third declension noun, sometimes as a Latinized Greek noun, and sometimes as an indeclinable name as is common with foreign borrowings, this should not be the archetype for a paradigm. As to whether these nouns are third or fourth declension, such a question is a little silly, but I would agree the should probably put them in the third declension, if any declension at all. I agree with you that those dative, ablative, and locative forms seem speculative at best.

Here's a source for the 5th declension locative. There does not seem to be much good evidence for the locative plural, but I'd advise putting and asterisk next to it to show it is unattested but would likely be correct.

Also, I haven't the faintest idea what the word “rangs” you used above means. —JohnC5 00:50, 25 March 2016 (UTC)

There are only three type of edit summaries:

1. Me providing sources. The formating maybe was bad as the number of characters is restricted, but nothing else should be wrong with it.
2. "There shouldn't be any any 4th/5th declension word with locative, thus this should be made up. Else please provide an example with locative." There is a typo "any any". But except from this, I can't see what should be wrong with it. If the locative exists, it should be easy to give an example. So, it shouldn't be too much to ask for an example. Well, I could have asked for a source on the talk page, but I doubt that anyone would provide a source in in the next few days. So as far as I can see, the best way was to remove it.
3. " "reconstructed" = made up by an Wiktionarian (?). A declension like of echo or Dido should be more likely. Dictionaries also mention the acc. -on ("Prop. 1, 20, 17 M.") and state "Abl. ō". Here I guess only the first part, ""reconstructed" = made up by an Wiktionarian (?)", could be problematic. But here it was a question (see "(?)"), and there was no blaming or anything.

So as far as I can the, the only comment which might sound rude is "You're in error.", which followed after your revert. Else please be more specific.

"rangs": "rangs" was supposed to be "ranks" ("a level in an organization"), refering to users' ranks. A statement from an admin isn't correct just because he is an admin, and a statement from an IP or newbie isn't wrong just because he is an IP or newbie.

5th declension locative: Quoting from your source: "The Locative form of this declension ends in -é. It is found only in certain adverbs and expressions of time". Does that qualify as a regular locative? 1st, 2nd and 3rd declension have real locatives for city names. So "adverbs and expressions" is something different. Adverbs shouldn't attest a Latin locative. But maybe expressions can. I'm not sure about it, but I'm okay with it. So thanks for proving the source.

OLD is just one dictionary. Lewis and Short for example have: "Argo, ūs, f. (gen. Argūs, Prop. 3, 22, 19; acc. Argo, Varr. ap. Charis. p. 94 P.; Argon, Prop 1, 20, 17 Müll." So the OLD should miss some forms.
Some borrowed names were indeclinable. But are there any non-biblical indeclinable Greek names? "Isaac" for example is a indeclinable name, but it's biblical. Even if there are non-biblical indeclinable Greek names, is "Dido" sometimes indeclinable? The dictionaries that I used didn't mention a gentive "Dido".
Lewis and Short also mention regular 3rd declension forms for Calypso. And in New Latin such forms might have become more common. googling for forms of *echo, -onis (like genitive echonis or plural echones) brought up some results which could indeed be forms of "echo".
It's not just dative, ablative and locative. (a) The accusative and the plural is also doubtful. The accusative usually is (said to be) -ō and sometimes -ōn or -ūn [see the PS], but not -um. (b) Grammar books back up dative and ablative -ō. But I have seen none which has dative -uī or ablative -u. So if dative and ablative are unattested, the forms in -ō are the more common constructed forms. And maybe they were also used in New Latin.

As I provided sources, both dictionaries and grammar books, and as the old declension mentioned in Appendix:Latin fourth declension is unsourced and incorrect compared with dictionaries (see the accusative), the version after my edits should without any doubt be better. Whether or not the dative and ablative are stil doubtful, and whether or not a declension table should be present at all are other questions.

PS: (1) Other examples which should support the declension I added (quoting Georges):

• "Callistō, ūs, f. [...] Dat. Callisto, Catull. 66, 66: Akk. Callisto, Hyg. astr. 2, 1: Abl. Callisto, Hyg. fab. 155."
• "Allēctō (Alecto), Akk. ō, f. [...] Abl. Allecto, Serv. Verg. georg. 2, 98."
• "Hērō, (Erō), ūs, [...], Ov. am. 2, 16, 31 (wo Akk. Heron). Serv. Verg. georg. 1, 207 (wo Akk. Ero)"

(2) "-ōn or -ūn" maybe should be "-on/-ōn or -un/-ūn". Georges sometimes has it as -ōn, sometimes as -on. Georges' "Heterokl. Akk. Argon" could mean that it's (sometimes) the second declension ending -on. But other dictionaries don't say that "Argon" is a heteroclitic form. So my impression is that Georges here is somewhat inaccurate.

-84.161.32.66 02:19, 25 March 2016 (UTC), PS 02:52, 25 March 2016 (UTC)

1. Your TR post begins with “The declension there (especially dative and ablative) obviously were made up by wiktioanrians ("reconstructed").” which is rude beyond belief. I don't care if you don't think you are being rude; if you continue in this manner, someone is going to block you.
2. I meant to mention the accusative singular and the plural forms. They are spurious too.
3. Latin treatment of borrowed words is as wild as might be expected. I've seen the same author decline nouns in the third and second declension in the same sentence. Furthermore, the common borrowing of AG -οι in Latin is -ī, not -ō. I'd say that the only things we can reconstruct at all are nom. -ō, gen. -ūs (sometimes), acc. -ō / -ōn / -ūn, and maybe dat./abl. -ō. I still see little evidence that these should not be considered a handful of borrowed AG forms with the rest filled out by indeclinable forms. This is not a stable declension worthy of a template and certainly not one that should be generalized elsewhere, especially since the evidence is scattered over a few isolated forms. The best that can be done is to mention the forms available a page by page basis. I'm not defending what's on the 4th declension page, but to say that this represents any sort of standard system is incorrect.
4. There are many non-admins and anons that I've come to trust, nor am I attempting to flout my administrative privileges. Indeed, I think you will find me significantly more forgiving than some others.
5. Also the reason I like the OLD is because it is very modern (2012) and does not possess the urge that plagues older Latin lexicographers to categorize every word into a nice declensional box. —JohnC5 03:04, 25 March 2016 (UTC)

• Usually, I'm not rude (though maybe sometimes somewhat objective and unemotional, which might be misinterpreted as rude). But I might react rude here, when someone is rude to me. And IMO reasonless reverts were and are rude. The "TR" post was influenced by that, and the typos might indicade that it was written in the heat of passion. But I can't see were that post is "rude beyond belief". The declension here was made up by wiktionarians, wasn't it? So the statement should be correct. Did I insult or attack anyone personally? No. Did I insult or attack anyone? No. Was it rude? Maybe because of the "obviously were made up". But "beyond belief"? I can't see why. Under other circumstances I might have written it rather like "The declension there (especially dative and ablative) seems to be made up." Is that still rude? If so, why and what would in your opinion be a non-rude way to express the same thing?
  According to Wiktionary:Blocking policy a block would just be justified in case of: "Causing our editors distress by directly insulting them or by being continually impolite towards them.". I didn't insult anyone, and I wasn't and I am not "continually impolite towards them". But: (1.) That should apply to admins too, like when admins insult someone or are "continually impolite", they could be blocked too. (2.) The page also says "Some effort should be made to explain to people why their edits are considered incorrect". So a simple block should violate those rules. As for the first point: Reasonless reverts IMO are impolite, and "whom I trust more than you" is an ad hominem. Correctness is not a matter of trust, and as I provided sources it's not a matter of trusting me.
• "For one, all of your sources are wildly out of date and have been superseded." -- Are they really out of date and have been superseded? (1.) In many cases older and younger grammar books mention the same things, like for the "usual" Latin declension. (2.) I've also looked into a few grammar books from the 21st century. They usually don't mention Greek declensions in the same way as those older books do. And my impression is that authors of modern Latin grammar books sometimes fool the reader or have a lack of knowledge in some aspects. For example, in modern grammar books one can find statements like "the vocative only differs from the nominative of some 2nd declension words". That's obviously incorrect as there's e.g. cometes from the first dclension. (Yes, there are also some aspects were modern grammar books are better, like always indicated the vowel length.)
• "I'm not defending what's on the 4th declension page" -- IMHO you did defend it by reverting. If it was reverted because the change was undiscussed, then one could have used an edit summary like "Please first discuss it (at [correct place to dicuss it])".
• "I still see little evidence ...", "This is not a stable declension worthy of a template and certainly not one that should be generalized": Might be. But I didn't create the template and whether or not there should be a template should most likely be discussed elsewhere. Though I can't tell you the correct place where to discuss it (I drink neither tea nor beer).

-84.161.58.137 04:42, 25 March 2016 (UTC)

• I'm sad to say that many people don't particularly care about the blocking policy when it comes to anons. Much is gained by having an actually user page, not the least of which is the ability to contest arbitrary blocks. Then again, people do get banned for being annoying all the time. That's not how it should work. It is, however, what happens.
• We tend to not trust anything that's more older than 1900 with exceptions for extremely important, foundational works. Linguistics has changed massively in the ensuing time to the point that many of the traditional ways of categorizing things have been abandoned as unnecessary or arbitrarily Talmudic.
• You misapprehend how the rollback tool works. When one uses the rollback tool (reverts multiple consecutive edits by a user), it does not allow the admin to give an edit summary. Normally I would go you your user page, but it is not common practice to leave discussion on the talk page associate with an anon, especially one that appeared so recently. I was in the process of examining your other posts to find out where to respond when you posted here.
• That mystic land of which you speak is Wiktionary talk:About Latin. There such matters may be brought up in the sight of those who care. My wish would be that the Greek table would be deleted from the 4th declension appendix and as would the two templates associated with Argo and echo. —JohnC5 05:02, 25 March 2016 (UTC)

Inflections of dialectal forms

Hmm, so it isn't the practice to include inflections of dialectal forms such as φωνᾱ́εις (phōnā́eis) in main entries such as φωνήεις (phōnḗeis)? I sort of assumed all the information would be on the one page, as with dialectal forms that don't have different nominative singulars. Perhaps the the about page should say something about this. — Eru·tuon 00:20, 26 March 2016 (UTC)

What about crases? I moved the inflection of ᾱ̔νήρ (hānḗr) to ἀνήρ (anḗr). — Eru·tuon 00:21, 26 March 2016 (UTC)

@Erutuon: I'm basing this on this discussion I had with Newt a while ago, which extended here. Thoughts? —JohnC5 00:47, 26 March 2016 (UTC)

Hmm... I don't really like the idea of sort of haphazardly putting inflected forms in the alternative form entry when they happen to have a different lemma form (nominative singular, masculine nominative singular, first-person singular present indicative) from the main entry, but keeping them in the main entry otherwise. For Homeric ἀήρ (aḗr), there's a different genitive but not nominative, so the inflection gets put in the main entry. Contracted verbs, on the other hand, have different 1 s. pres. ind., but their present and imperfect forms go in the main entry nevertheless.

Still, maybe readers would be confused if they went to the alternative form entry and didn't find the inflection they were looking for. I like the idea of having inflections in a single location, on a separate page with entries with a particularly large number of forms. Perhaps the solution would be to link to this centralized location in the alternative form entry. — Eru·tuon 01:14, 26 March 2016 (UTC)

That could work. As long as we are consistent —JohnC5 01:23, 26 March 2016 (UTC)

I give each entry its own inflections. It would be a bit strange to have colouring as an inflection of color. This only really makes sense in cases where the paradigms are (mostly) separate, though. When only a small proportion of the forms differs, then it may be better to treat them as alternative forms within a single lemma. However, in that case, we shouldn't be labelling them alternative forms either, but just inflections. In case of a paradigm where the nominative singular is different but nothing else is, something like {{inflection of|xxx||nom|s}}. —CodeCat 01:54, 26 March 2016 (UTC)

Agreed. It's best to put the inflections of δᾶμος on the page for δᾶμος and the inflections for δῆμος on the page for δῆμος, and nothing else. The exception I would prescribe is contraction, as (a) both forms were used in each dialect and (b) it is important to show the relationship between uncontracted and contracted forms. I don't think there were any words where only the nominative singular differed, so that's probably not a problem. —ObsequiousNewt (εἴρηκα|πεποίηκα) 22:54, 26 March 2016 (UTC)

There are cases of the opposite, where only the nominative singular is the same, and the rest of the cases are different: Homeric ᾱ̓ήρ; ἠέρος, Attic ᾱ̓ήρ; ᾱ̓έρος. In that case, there have to be two tables in the same entry.

What irks me is that with forms like δᾶμος (dâmos), the only piece of actual information that will be in the alternative form entry is the pronunciation and the inflections. The rest, the etymology and definitions especially, will be in the other entry. So it's an entry whose only purpose is 1. giving pronunciation, 2. saying this is a dialectal form and 3. providing the inflections associated with that form. It makes more sense to me if the entry is instead made like a "form of" entry and the inflections are moved in with the main entry, where all the other information is. — Eru·tuon 23:13, 26 March 2016 (UTC)

On the other hand, just about the entire entry is duplicated in color and colour, so I probably don't have a case here. Oh well. — Eru·tuon 23:17, 26 March 2016 (UTC)

censor

The same quotation is also used on legio, lecio, magister, macister and consul, FYI. (I don't mind if it's removed from all those places; maybe it belongs as etymology or a reference in lecio and macister, but not elsewhere.) - -sche (discuss) 01:46, 5 April 2016 (UTC)

@-sche: Hmmm, interesting point. I think that a reference for the etymology sections to those entries would be exactly right. I just found it odd to have it under Latin quotes, as you might imagine. Would you change those, or shall I? —JohnC5 01:51, 5 April 2016 (UTC)

I'll change them. - -sche (discuss) 02:05, 5 April 2016 (UTC)

@-sche: Thanks a bunch! —JohnC5 02:08, 5 April 2016 (UTC)

Hebdomecontacometae and Tessarescaedecatitae

Hello JohnC5. What is the reason for changing the declension tables in Hebdomecontacometae and Tessarescaedecatitae to add "masculine Greek type with nominative singular in -ēs"?

• There is no nominative singular, so this addition is irritating, misleading, contradicting.
• Greek words in -ης can have -es and -a in Latin, like there are psalmista, sophistes and sophista, prophetes and propheta.

As there is no singular and as the singular could also end in -a, it makes more sense not to have the "masculine Greek type with nominative singular in -ēs" part. Greetings Boðberi (talk) 15:22, 8 April 2016 (UTC)

@Boðberi Fine. —JohnC5 17:08, 8 April 2016 (UTC)

OK, and thank you. -Boðberi (talk) 23:46, 8 April 2016 (UTC)

analogia

Hello. Are you sure with the length of the alpha? In this copy of Smyth's Greek Grammar it is: "The feminine nominative singular ends in -ᾱ, -ᾰ, or -η" and "Most, if not all, of the substantives in ᾰ are formed by the addition of the suffix ι ̯α or ια". That is, the Greek alpha can be short. As far as I can see, it's a short alpha in the Greek word ἀναλογία. Though I have to admit that some of the sources I use(d) do not properly mark the length of the alpha. -Ikiaika (talk) 05:43, 13 April 2016 (UTC)

@Ikiaika: Yep, I'm quite sure! The -ῐᾰ (-ia) being discussed in that passage (the inherited PIE *-yós + collective suffix *-h₂) is closely related to but not the same as -ῐ́ᾱ (-íā) which is being used here. —JohnC5 06:01, 13 April 2016 (UTC)

Ok and thanks.

Does ā̆ mean it is both long and short in Latin or that the length is unknown? From the context I'd guess it's unknown, but just seeing ā̆ I'd guess it can be both. So if the length is unknown there maybe should be a text saying that the length is unknown. -Ikiaika (talk) 06:24, 13 April 2016 (UTC)

@Ikiaika: Yeah, we don't have a great way of showing unknown vowel length around here. I find the concept of analogiản humorous, but it's a bad idea. I'll make a note.

Scripts

I just want to thank you for filling all the missing scripts in pages. I have no idea how to write any of these, my knowledge is limited to Latin, Greek, Cyrillic and Gothic. It's a great help! —CodeCat 14:27, 20 April 2016 (UTC)

@CodeCat Thanks! My Avestan stuff is mostly from AP:Avestan script, and I'm thinking of adding it to the edit tools with transliterations so that people can use it more easily. Hittite and Old Persian Cuneiform are a nightmare and take me forever to work with. —JohnC5 14:35, 20 April 2016 (UTC)

Regent

Hello JohnC5, regarding the form Regents:

• de.wikt mentions it related to wine, so it's different from the entry here.
• google book results use it in other senses, but I'm not sure about it's meaning. E.g. "Der Senat ist in zwei Klassen oder Häuser, in die Regents und Non-Regents abgetheilt. Jene bilden das Ober- oder White-Hood Haus [...] Die Non-Regents consituiren das Unter- oder Black-Hood Haus [...]" seems to refer to (older) English politics. Compared with Meyers Großes Konversations-Lexikon the word could be "Regent m., genitive Regents, plural Regents", from English regent with the meaning "A member of a municipal or civic body of governors" or "A member of governing board of a college or university".

So in both cases, Regents is not a form of Regent meaning ruler. Greetings, Ikiaika (talk) 09:01, 24 April 2016 (UTC)

@Ikiaika Makes great sense! Could I ask you to add those other meanings with the plural Regents? —JohnC5 15:43, 24 April 2016 (UTC)

I'll try to add the meaning related to politics and universities. I'm most likely not going to add the meaning related to wine (lack of knowledge, and de.wikkt's entry seems to be quite complicated or detailed). If I don't add it, I'm going to add it here: Wiktionary:Requested entries (German). I hope that's fine too. Greetings, Ikiaika (talk) 16:13, 24 April 2016 (UTC)

@Ikiaika Thanks. And if it is any help, deWikt is referring to w:Regent (grape). I can add that once you're done making changes. —JohnC5 16:30, 24 April 2016 (UTC)

As for the wine-related meaning:

• The genitive should be Regents (found 2 examples at books.google), not Regenten which de.wikt mentions (found 0 examples at books.google). I'm not sure if there is a plural. As a Sorte (de.wiki), that is sort, kind, it maybe has none. Maybe there are transferred senses like "a bottle with certain wine" which can have a plural. But I found no plural at books.google.
• I don't know the etymology. By the declension, I'd guess it comes from French régent or English regent. There I'd guess it comes from French, as France is often associated with wine.
• Regarding the definition, I'm not sure if that's enough. Maybe more information should be added, but maybe it then would be too encyclopedic.

I'm still working on it, but maybe I'm only going to add an example for this meaning (if I find a good one).

As for the other meaning:

• The singular might be rarer.
• The definition needs to be improved, but I'm working on that too.
• Examples for this meaning will certainly be added.

-Ikiaika (talk) 17:30, 24 April 2016 (UTC)

It took a while (I fell asleep yesterday, sorry), but now it should be done. A few notes:

• Regent (from English): Maybe it also refers to politics ("a member of a municipal or civic body of governors") and not just to universities, but I haven't seen such a usage.
• Regent (grape): Carsten Sebastian Henn, Vino Diavolo. Eifel Krimi, Hermann-Josef Emons Verlag, 2013, e-book edition without page numbers has the genitive and could be cited, but that doesn't seem to be a good example. Also the genitive (and all other cases as well) can be paraphrased by adding the word Rebsorte (variety of grape, type of grape), that is nominative and accusative die Rebsorte Regent, genitive and dative der Rebsorte Regent. As for the plural, the grape type Regent maybe has none, and for wine made of Regent there is the word Regent-Wein with the plural Regent-Weine.

Greetings, Ikiaika (talk) 04:04, 25 April 2016 (UTC)

@Ikiaika Thanks for all your help; this looks good to me. Also, were you that IP with whom I was discussing Latin declensions before? I realize I was somewhat of a dick, and I'd like to apologize if your the same user. —JohnC5 04:12, 25 April 2016 (UTC)

Might be, we certainly discussed some Latin (like above with analogia).
Well, IMHO it's a general problem anyway. Some IPs obviously are bad, and then sometimes long time users and even admins react in a bad way towards other IPs by misjudging them; and then those IPs might react unfriendly too. Out of experience at the German Wikipedia I know that IPs can even be blocked for valid edits for which a normal user wouldn't be blocked, because an admin might simply misjudge edits of an IP user. Greetings, Ikiaika (talk) 05:03, 25 April 2016 (UTC)

Latin schĕma

You wrote in the etymology of σχέμα (skhéma) that it is supported by Latin schĕma. But schema#Latin only has schēma. Also, what you mean by "alternative reading"? An alternative reading to me would mean an alternative reading of the same written word, but ΣΧΕΜΑ and ΣΧΗΜΑ are written differently. --WikiTiki89 15:18, 25 April 2016 (UTC)

@Wikitiki89: To be honest, I did not do a great job on this entry. The information is based primarily on this LSJ entry. I was having trouble finding evidence of schĕma as well. Feel free to alter the entry as you see fit―I had just grown weary of seeing this word in WT:WE. —JohnC5 15:21, 25 April 2016 (UTC)

Template:ine-decl-noun-n and collectives

I have held back on fully implementing this template because of the question of collective nouns. Specifically, which nouns had collective plurals and which had regular plurals? Ringe mentions explicitly that -men- nouns had collective plurals, but he also reconstructs a few other nouns with collectives, including athematic *wédōr ~ *udnés (“water”) and thematic *kʷekʷléh₂ (“wheels”). The latter is a thematic collective, which had a rightward shift of the accent relative to the base noun (*kʷékʷlos in this case). Ringe also states that neuters probably had no plurals at all in early PIE, but only collectives, but that it is not reconstructable whether this situation was maintained in later PIE and to what degree. He reconstructs several neuter nouns with regular plurals, in any case. —CodeCat 18:54, 25 April 2016 (UTC)

@CodeCat: I had been thinking about asking this for a week now. You are very astute (or did I ask somewhere and have just forgotten?). Is it possible to make the collective be an option (or mention both options?) The only collectives I've seen are the *r/n-heteroclitics and *-mn̥-stems as you mention. —JohnC5 19:19, 25 April 2016 (UTC)

According to Fortson (page 118), in Anatolian, Greek and Old Avestan, neuter plurals take a singular verb. This agrees with the statement that neuter nouns originally had only collectives, since collectives were gramatically singular. Fortson also mentions that animate nouns could also have collectives, such as AG μηρός (mērós), collective μήρα (mḗra). It seems, then, that collectives were just derived nouns that came to be used as the plural of neuter paradigms, because they lacked one originally. The question, then, is whether we should list collective forms as part of the inflection table, or as separate derived nouns. A difficulty with showing them in inflection tables is nouns like the Greek one, which have a plural and a collective. We could decide to list both, and include a note for neuter nouns that the plural may be anachronistic. But I'm not sure. —CodeCat 19:47, 25 April 2016 (UTC)

@CodeCat: I've created Ancient Greek entries for μηρός (mērós) and μῆρα (mêra) — the latter of which I took you to mean by *μήρα (mḗra) — so would you like to add your information about Proto-Indo-European collectives to either or both of those entries' etymology sections? — I.S.M.E.T.A. 20:39, 17 May 2016 (UTC)

@CodeCat: I think listing both with a note would be my preference, but only when we have evidence of collectives. In the absence of a collective for a particular paradigm, list the theoretical plural with a note. —JohnC5 19:53, 25 April 2016 (UTC)

This can be made to fit nicely into the existing n= parameter. This parameter takes a bunch of letters that indicate the numbers, the default is n=sdp which means singular, dual, plural. I could easily add a c to that, so that for a neuter noun with no plural, you'd specify n=sdc (for some reason, neuters did have duals). There would probably need to be some extra parameters to specify the collective form, though, as it's not as simple as adding an ending (accent shifts, too) and beyond the kind of complexity that I want to add into the module. —CodeCat 19:58, 25 April 2016 (UTC)

@CodeCat: That sounds very good. I concur that the collective main and oblique stems will probably need to be specified by hand, since collectives like that of *sóḱr̥ may be rather exotic. —JohnC5 20:02, 25 April 2016 (UTC)

I've now added collectives to the templates, and added collective forms to *kʷékʷlos and *sóḱr̥. The collective of the latter was actually incorrect; collectives of athematic nouns are amphikinetic, and have full grade of the root in the nominative. —CodeCat 20:40, 25 April 2016 (UTC)

@CodeCat: Concerning *sóḱr̥, I will point out that Beekes, Kloekhorst, (and Klimp) all prefer *sḱṓr with Kloekhorst explicitly rejecting *séḱōr for Hittite /zakkar/ and mentions that similarly Hittite /widār/ ("water") must come from *udṓr (whence also ὕδωρ (húdōr)). Klimp actually goes into the existence of Ø-ṓ ~ Ø-Ø-ó type collectives in some detail. I can send it to you if you'd like. I was also wondering whether we should move *pléwmō to the collective of *pléwmn̥ given that lungs are normally referred to in the plural and the existence of क्लोमन् (klóman) as a neuter singular. —JohnC5 23:34, 25 April 2016 (UTC)

The Leiden model

I've been reading up a bit more about the "Leiden" model of nominal inflection. I'm not sure if you're familiar with it (I wasn't very much, until now). The most striking deviation is the collapse of the traditional hysterokinetic and amphikinetic types into a single type with 3-stage ablaut (termed "hysterodynamic"):

• Nom: é-Ø-Ø
• Acc: Ø-é-Ø
• Gen: Ø-Ø-é

The later types can be derived from this assuming levelling and a change of unstressed e to o at some point:

• An intermediate type (Beekes's type 2): extension of the suffix full grade in the nominative, followed by its change to o
• Later amphikinetic (Beekes's type 3): from type 2 by extending the accusative stem to the nominative
• Later hysterokinetic (Beekes's type 4): extension of the accusative stem into the nominative, either directly or with type 2 as intermediate

I think these ideas are certainly interesting, but I'm hesitant to include it in entries as it's not exactly mainstream (yet?). As a more practical point, our tables don't support the 3-stage ablaut yet. —CodeCat 21:58, 16 May 2016 (UTC)

Also, as an additional point, Beekes says that the root grade in type 3 can be zero, and indicates it has its own subtypes. So that would at least allow for a collective *skṓr like you suggested above. However, it's also pretty much guaranteed that such a form had *sékōr as an earlier preform, because that's the only way an "o" could make its way into the nominative that I can think of (i.e. by the process of type 3 noted above). Kloekhorst goes with an alternative chronology, where "o" arose in the accusative first, by extending the nominative root grade into the accusative, and then this stem was extended back into the nominative to give type 3 later. This would eliminate type 2 as an intermediate for type 3, but the type that Kloekhorst reconstructs as the intermediate doesn't seem to be attested anywhere? —CodeCat 22:10, 16 May 2016 (UTC)

@CodeCat: Thanks for this research. I had seen reference to the Leiden model but had not taken the time to examine it fully. We could include Leiden model alternative forms, but we are only increasing the complexity for the uninitiated reader (I'll admit all the short descriptions I've read of the Leiden model have been somewhat mind-bending due to concision and lack of examples). Are there any good sources on this matter you would recommend? I would enjoy reading up on it if I ever finish my back log of Beekes and Ringe. :) I've really appreciated the discussion we've had over the last few days―you've lavished great time and effort on these templates that are mostly for your and my pleasure, and I do appreciate it! —JohnC5 22:23, 16 May 2016 (UTC)

libertus and liberatus

I am sure I am misunderstanding how these templates work, or possibly I am misunderstanding the word formation rules. It seems to me the -tus suffix in libertus should be a noun-forming suffix because libertus is the noun formed from the adjective liber with the sense `one who is manumitted'. Similarly, for verutus, I called it adjective-forming because the sense of -tus in this case is `armed with.' The adjective usually is construed to derive from the participle and is liberatus. If this doesn't seem wrong to you I will add this adjective also. I apologise that some of my recent edits are a bit of a mess. I am trying to be cleaner. Thank you for watching over some of them.

Edit: on further thought, I think I understand your argument. Libertus is second declension and a concrete noun, not an abstract noun. Therefore it was originally an adjective (libertus) formed from an adjective (liber) from which an identical noun (libertus) was subsequently derived, and this is why we say the etymology of the noun involves an adjective-forming suffix? Thanks. Isomorphyc (talk) 03:29, 23 May 2016 (UTC)

(Post edit-conflict) @Isomorphyc: No problem, and I should have left more explanation. This is actually an interesting problem. Generally, -tus is an adjective-forming suffix and not a noun forming one (ignoring the third and fourth declension suffixes that are spelled the same), and if you get a noun, it would be a substantive use of an adjective, as you said above. I'd also say that līberātus is the participle formed from līberō, not directly from the adjective līber. What's really tricky about this one, however, is that this noun may have been formed as Proto-Italic *louðertos or Proto-Indo-European *h₁lewdʰ-er-tos (see more at *h₁lewdʰ-). If we say that this word comes from līber + -tus in Latin, that represents a pretty surface analysis and synchronic interpretation, but in fact that etymology should state that lībertus is inherited from an earlier ancestor. At some point this was an adjective which became a noun, but I'm not sure we know at what stage (though I'd prefer PI as opposed to PIE). Shall I change the etymology? —JohnC5 03:31, 23 May 2016 (UTC)

You make an good point. There's the potential argument in ἐλεύθερος, which is both substantive and adjective, that this happened prior to Proto-Italic, though I am not in a position to make it. If you wanted to expand or clarify the etymology, I would welcome that, but in retrospect I think most readers would have realised what was meant. I was abnormally confused because I was making some superficially similar edits at the time, but I ought to have looked a bit closer. Isomorphyc (talk) 03:54, 23 May 2016 (UTC)

@Isomorphyc: I very much appreciate your work! It's always nice to have thoughtful editors helping out. —JohnC5 03:58, 23 May 2016 (UTC)

Thank you; you kind to say that, and it is a great pleasure to read what you write here also. Isomorphyc (talk) 05:57, 23 May 2016 (UTC)

The suffix of lībertus is certainly the same as the one found in participles, due to its declension. But in PIE, it's normally found affixed to roots, not to derived verbs and certainly not to another adjective like līber. So something unusual is going on here, but I don't know what. —CodeCat 15:48, 23 May 2016 (UTC)

So, can we say that this must be an odd PI formation, since it is attested in PI and cannot be PIE? —JohnC5 15:51, 23 May 2016 (UTC)

I'm not sure. An entirely different (?) adjectival suffix, also *-tos, was also used for adjective comparisons in several languages. So there may be yet other functions of this that we're not aware of. —CodeCat 16:09, 23 May 2016 (UTC)

Has anyone written about the ablaut patterns of this? I could envision *h₁léwdʰ-r̥-tos > louðertos > lībertus if we went back to PIE. Also, does this adjectival suffix normally affix to other adjectives or was *h₁lewdʰ-er-os a noun? —JohnC5 17:45, 23 May 2016 (UTC)

Osthoff's law

Just to remind you, this law didn't apply to Indo-Iranian! —CodeCat 15:46, 23 May 2016 (UTC)

@CodeCat: Why do I keep having so much trouble with this dang rule? *sigh* —JohnC5 15:47, 23 May 2016 (UTC)

MediaWiki:Edittools

I hope you know that you can test this from the preview without having to save every time. --WikiTiki89 15:47, 25 May 2016 (UTC)

@Wikitiki89: Really? I thought I had tried and it didn't work. Maybe I just didn't notice once and then never tried again. Regardless, in the Combining diacritics section of Greek all the diacritics are scrunched up together and unclickable for me. Any idea how to fix that? —JohnC5 15:50, 25 May 2016 (UTC)

Has this fixed the problem? — I.S.M.E.T.A. 11:12, 16 June 2016 (UTC)

Yes, it seems to have fixed it. Maybe it would work for the Greek ones too. --WikiTiki89 14:10, 16 June 2016 (UTC)

@Wikitiki89: John fixed the Greek ones. Thanks, John! :-)  — I.S.M.E.T.A. 13:04, 18 June 2016 (UTC)

Reversion of Reconstruction:Proto-Indo-European/-ós

[2]. Why? I acknowledge that it's a noun deriver, but the category is not created and there is not link. The one with word is there and can be linked though. Thanks.

• I read about uninteresting letters/characters above (Æ in English, January 25th). What's that? I may help. Sobreira (talk) 07:23, 27 May 2016 (UTC)

@Sobreira: Sorry for the revert. I'm not entirely sure now what the situation is. CodeCat Is apparently switching to the |id= param as opposed to |pos=; though I don't yet understand the difference (I'm sure there is one). Currently the descendants listed are not *o-grade, which is incorrect according to the article. To be honest, I dunno. Again, sorry for the confusion on that front.
The uninteresting letters functionality has already been implemented somewhat, as seen in English (en) and French (fr) code blocks here. If we could get this for more languages, that would be useful. —JohnC5 16:23, 27 May 2016 (UTC)

As far as I can tell from their usages in {{affix}} and {{suffix}}, |pos= specifies the nature of the word so formed, whereas |id= specifies the nature of the morpheme (or whatever) involved in that word's formation. Furthermore, whereas |pos= changes words in derivations categories (for example, changing {{suffix|la|verbum|ō}} to {{suffix|la|verbum|ō|pos=verb}} changes Category:Latin words suffixed with -o to Category:Latin verbs suffixed with -o), |id= adds parenthetical comments to derivation categories (so that changing {{suffix|la|verbum|ō}} to {{suffix|la|verbum|ō|id2=verb}} changes Category:Latin words suffixed with -o to Category:Latin words suffixed with -o (verbs)). This has the functional benefit of allowing multiple derivations categories both with and without POS-specificity to be treated within the same template ({{affix}}, {{suffix}}, or whatever) and is, therefore, an improvement, in my opinion. Does that make sense? @CodeCat, please confirm or correct my statements. — I.S.M.E.T.A. 10:55, 16 June 2016 (UTC)

The id= parameter also ties in with {{senseid}}, so clicking on links to the affix takes you to the right sense. —CodeCat 13:09, 16 June 2016 (UTC)

@CodeCat: Oh, cool; I didn't know about that. A question: Would the |id= system be able to handle Category:Ancient Greek words prefixed with ἀ- (alpha privativum), Category:Ancient Greek words prefixed with ἀ- (alpha copulativum), Category:Ancient Greek words prefixed with ἀ- (alpha intensivum), and Category:Ancient Greek words prefixed with ἀ- (alpha euphonicum) for the four Ancient Greek prefixes at ἀ- (a-)? I was at a loss to think of a good way to distinguish them under the |pos= system… — I.S.M.E.T.A. 12:52, 18 June 2016 (UTC)

Can we not give them clearer names that everyone can understand? —CodeCat 13:28, 18 June 2016 (UTC)

@CodeCat: Like what? — I.S.M.E.T.A. 14:09, 18 June 2016 (UTC)

Really? How about English, to start off with? —CodeCat 14:18, 18 June 2016 (UTC)

@CodeCat: I was just using the names given in the entry. Would you be so kind as to come up with better names than those that are established in Ancient Greek grammars? Forgive me, for I lack the imagination. — I.S.M.E.T.A. 14:50, 18 June 2016 (UTC)

Compare Category:Latin words prefixed with in-. The subcategories use the meanings as ids. So there'd be Category:Ancient Greek words prefixed with ἀ- (not), Category:Ancient Greek words prefixed with ἀ- (same), Category:Ancient Greek words prefixed with ἀ- (intensive). However, it looks like the suffix for the meaning "same" is actually ha- so that should be used for the category. —CodeCat 14:55, 18 June 2016 (UTC)

@CodeCat: Thank you. Something like that would be fine. BTW, I was surprised that, during this reform (from |pos= to |id=), that you had not undertaken to change Category:Language words affixed with affix to Category:Language terms affixed with affix, per the character and derivation categories. I would think that that would be necessary, given the existence of terms like United Statesian (United States + -ian). — I.S.M.E.T.A. 15:06, 18 June 2016 (UTC)

Recent edits to PIE pages by User:Nima Farid

This user is new so I don't know how good their historical-linguistic knowledge is. If you know anything about Persian, could you check their recent additions? —CodeCat 12:40, 27 May 2016 (UTC)

@CodeCat: So the edits to PIE entries look correct except they are categorized incorrectly. For Iranian languages besides Old Persian and Avestan, we tend to add the verbs as descendants of the the base root. This user incorrectly claims once that the -dan ~ -tan suffix is from *-tós (I don't remember whence it stems, but Fortson talks about it in the section of Persian). Instead, we consider these verbs to be just descendants of the root itself. I would also ask Vahagn to take a look. The only one that seems very suspicious is PI *gaub- and Persian گفتن. I also can't confirm the addition to *kh₂em-. —JohnC5 16:15, 27 May 2016 (UTC)

Greek LSJ Links

Hi JohnC5, I noticed you reverted a small subset of the edits I made with my test robot OrphicBot. I couldn't discern a pattern, except that they were from early in the run; of course they are almost all in error for putting External links after References, and there are three or four other problems I have discerned since as well. Would you like me to undo all of them now that I have run my test and have the information I wanted? Obviously I don't want to leave the Greek section temporarily a mess; and I don't want you to feel I have created work for you in anticipation of your return. Of course if there is a pattern, or you have any advice, including, possibly, that I ought to wait a good while before trying something like this, it's something I would appreciate and respect as well. Thanks, Isomorphyc (talk) 23:01, 5 June 2016 (UTC)

@Isomorphyc: Sorry for the lack of explanation. I reverted those because they happened to be words that I had edited. My main issue with them was that they added LSJ under an External links section for pages that already had LSJ under the References section. It makes no sense to me to added LSJ as an external link if it is already a reference. It also seems like it should always fall under a References section anyway, but I don't know. —JohnC5 23:36, 5 June 2016 (UTC)

No need to be sorry-- I'm grateful for your help. Since I have been more active in Latin than Greek, I think you may be the only regular Greek contributor with whom I have corresponded at all. The reason the reference was duplicated is that I did not know until my robot posted that the template LSJ was a redirect for the template R:LSJ, so I didn't test against its presence. I had been hesitating between References and External links. My thought was that References should be used bibliographically, while External links can be used without implying a link is being used as a source. At least, this is the practice followed from time to time in the German section. I am trying canvass for opinions about this in a thread in the Beer Parlour, and I would be very grateful for any ideas or desiderata you would have. If I do make these changes, I'd want them to be broadly useful, of which I am in myself only a partial judge. I do wonder from time to time why the Greek section is a good bit underdeveloped compared to Latin, and think widely available working links could make it slightly friendlier. Isomorphyc (talk) 00:15, 6 June 2016 (UTC)

Edit: I need to add something-- I could be wrong, but I think only Gorgon and Gadara of the four the items you reverted has the redundant link to LSJ. Is there a different or subtler issue with the other two that I am not seeing? Again, no need to apologise for not leaving a comment -- I shouldn't have left so many errors behind. It was not a very good first run. Thanks. Isomorphyc (talk) 02:03, 6 June 2016 (UTC)

@Isomorphyc: It's possible that I misread the other ones. Regardless, I am grateful for the work you are doing. We also have many pages containing [[w:LSJ|LSJ]] that should be replaced with {{R:LSJ}}. Indeed, a great many AG entries are antiquated: using {{grc-ipa-rows}}, using the old declension and conjugation templates, not using {{R:Strong's}} and {{R:Woodhouse}}, etc. —JohnC5 02:35, 6 June 2016 (UTC)

This is very helpful. So it would not be too much clutter if I linked everything to LSJ, Strong, and Woodhouse, where available? I'm open to other suggestions, but my procedure at this time will be to replace any generic dictionary reference without bibliographical details in References with a headword link, but to put new headword links in External links. For grc-ipa-rows: Is it reasonable to add these with mechanically generated arguments from the polytonic spelling in addition to retaining any user-entered pronunciations on separate lines? Anything with both could be added to a category for manual clean-up. If this doesn't have obvious problems, the template itself could be rewritten to derive its arguments from the page name, as is the current procedure in Latin. For the stemming: I would have to do some verification to make sure the transition from the template system to the module system can be made mechanically without errors, but I am certainly open to trying this. Most likely I would approach these three in order, and the latter two will be slightly iterative, since they are I think ascending complexity. I will probably start a robot vote in a few days, so anything else you can think of before or after will be very welcome. Isomorphyc (talk) 03:52, 6 June 2016 (UTC)

I'm gonna ping Newt on this one. I'd also say, when available, add {{R:DGE}} too. It is quite extensive for the sections that have been written thus far. —JohnC5 03:57, 6 June 2016 (UTC)

Thanks -- I felt a bit at a loss because I didn't know whom else to write to for opinions. I've regularly found DGE very helpful too. If the first iteration looks ugly with too many links, something more compact can always be devised later. Isomorphyc (talk) 04:08, 6 June 2016 (UTC)

For what it's worth, I've always used the References header for LSJ, because that seems to be what everyone else was using. Since we don't rely on authoritative references for sourcing (except for etymologies), the References section tends to be more like the Further Reading header on Wikipedia rather than strictly bibliographic. Chuck Entz (talk) 05:15, 6 June 2016 (UTC)

Thanks, everyone, for commenting here. Here are my current thoughts on these three (separate) robot tasks. I'm only going to do the pronunciation and inflection template switches once everyone is comfortable that the dictionary hyperlinks worked out well, since that is much more basic.

I concede the point about External links. Excluding my edits, currently 99.3% of the references to LSJ are in the References section. In retrospect, it was absurd of me to want to follow the German model rather than what already exists in the Greek entries. What I was trying to do was retain a distinction between existing and robot-generated hyperlinks, but I agree this is unnecessarily fussy if nobody sees a problem mingling them.

For the dictionary links: My new proposal is to add the templates where valid in the References section. This amounts to adding 1450 new instantiations of the LSJ template, in addition to the 2606 existing ones, so that 4056 entries will have LSJ templates (out of 7498 total entries). Overlapping Wikipaedia links will be removed, but the 237 references to print editions will remain in case they are intended to be edition-specific. I will flag the few hundred lemmas without LSJ entries for manual arguments, because these usually have a different spelling in LSJ than Wiktionary, but sometimes are Byzantine words and do not. I will also undo my old External links changes.

For Strong's, Woodhouse, and DGE: I will see if I can alter the Strong's and Woodhouse templates slightly today. Depending on some Unicode issues I sometimes have, It should be possible to generate arguments more automatically. To the extent this is possible it will be worthwhile to use these templates more widely.

For the pronunciations: I actually did not quite realise grc-ipa-rows is deprecated, because I see it so often. Apparently it is used 4251 times, compared to 1267 for grc-IPA and 845 for grc-pron. It seems to me there ought to be a one-to-one mapping between grc-ipa-rows arguments and grc-IPA arguments; one would just have to add the breves to the unicode in the right places. I will test to see if this actually works, though I would consider this a separate project. grc-IPA with no arguments ought to be a drop-in replacement for words involving no ambiguous vowels, but I would want to do some testing to be quite sure.

For the nouns: I'm finding 316 instances of grc-decl and 4352 instances of the older tables. I think this change is also feasible; most likely the robot would just have to verify all the forms are the same before making the change. I would prefer to also test all the forms against the Perseus stemmer, but because of some Unicode issues I cannot tell now how difficult this will be for me.

The numbers in this note are based on a few month old Wiktionary xml file and may have drifted a bit since then.

Isomorphyc (talk) 19:22, 6 June 2016 (UTC)

grc-ipa-rows conceivably could be converted to grc-IPA, but it's definitely easier to just use the page title and add length marks as appropriate. grc-ipa-rows usually lacks said length marks anyway. That's the question I'm assuming you wanted me to answer? —ObsequiousNewt (εἴρηκα|πεποίηκα) 04:46, 9 June 2016 (UTC)

Hi ObsequiousNewt, thank you for stopping by; I have enjoyed reading over your work here very much. Some of the robotic tasks discussed in About Ancient Greek : Talk and on ancillary pages make me a bit unsure; but a subset again is unambiguous string substitution. After changing some of the old reference templates I would probably consider approaching this in stages. My Greek is mainly Koine and Attic prose, really mostly Plato, and a bit of drama, and not so comfortable as my Latin; at a minimum where archaic dialects and alternative forms are concerned I would have more than a bit to learn. My sense is that About Ancient Greek : Talk is the best place for me to put new discussions, if they are necessary. I think few Hellenists are in the Beer Parlour. Isomorphyc (talk) 20:05, 9 June 2016 (UTC)

Strong's Concordance

Thank you for reading and cleaning up my template/module edits. I was afraid my template especially was unreadable. If you are interested in the Hebrew, I have the full array here: https://en.wiktionary.org/wiki/User:Isomorphyc/Strong%27s_Numbers_Hebrew . But I think it is missing a curly brace around the zeroth entry; the script is not playing well with the toolchain I am using for this. I also noticed what you meant about a deprecated version of Woodhouse; I located the 1959 instances of grc-wh-page which should be R:Woodhouse. Isomorphyc (talk) 04:23, 7 June 2016 (UTC)

@Isomorphyc: Shouldn't this array be formatted ["<hebrew_entry_name"] = <number> and not the other way around? —JohnC5 04:31, 7 June 2016 (UTC)

It should. That's how I output it in Python but it prints this way instead, even though the file is laid out correctly. I wonder if the display scheme is alternating between forward and reversed for the Hebrew and non-Hebrew bits. I don't know if a CSV or something else would be any better for you, if you had enough experience with this to want to make an attempt? Isomorphyc (talk) 04:45, 7 June 2016 (UTC)

@Isomorphyc: You're going to need someone else's help with this. @Wikitiki89? —JohnC5 04:55, 7 June 2016 (UTC)

It's definitely due to the Hebrew text- strange things happen when you mix left-to-right and right-to-left scripts on the same line. There's an invisible left-to-right-formatting character that could fix that, but you don't want invisible characters in your data. Chuck Entz (talk) 08:08, 7 June 2016 (UTC)

Yes, Unicode has some weird rules about the directionality of numbers within bidirectional text, that is what is causing the strange display. But as far as trying to automatically find the Strong's number based on the page title, I don't think it's gonna work very well for Hebrew. The lemmas and spellings that Strong's concordance uses are not the same as the lemmas and spellings we use. --WikiTiki89 14:09, 7 June 2016 (UTC)

I think I have it working for about 50-70% of cases, with an error message for when it will not work. I am trying to see once the collisions and misses are removed, how significant are the number of false results. I don't have any Hebrew however; I would appreciate if you could take a look when I have finished? Of course it may be one will just have to generate a graceful error message should anyone try to call this template without arguments from a Hebrew page. Isomorphyc (talk) 14:33, 7 June 2016 (UTC)

Let me give you an example. The page שרה contains four different words that link to four different Strong's numbers; three of these words even have the exact same nikkud. On the other hand, the verbs גָּדַל (“to grow/be big”), גידל / גִּדֵּל (“to grow/raise/magnify something”), הִגְדִּיל (“to make something big”), and הִתְגַּדֵּל (“to magnify oneself”), which are different binyanim of the same root, all have the same Strong's number. --WikiTiki89 15:20, 7 June 2016 (UTC)

Thanks for the example-- I will have it output an error message, if anyone calls it from Hebrew without arguments. Isomorphyc (talk) 15:27, 7 June 2016 (UTC)

Yeah, I think this will have to do only for AG. Poor Hebrew will have to get its numbers directly. —JohnC5 15:33, 7 June 2016 (UTC)

@Isomorphyc: On an unrelated note, do you own User:Serenedijppitty? I would be careful about using multiple accounts because it could be construed as sockpuppetry. —JohnC5 15:43, 7 June 2016 (UTC)

I do; I only use it to avoid cluttering up my Contributions page with Sandbox entries. I was going to ask to coalesce it with OrphicBot, partly because the name is so idiotic. Unfortunately my Lua and template skills are so deplorable I need to recompile and run a lot to get anything done. I will put a note on the user page to this effect for the time being. Thank you for mentioning this. Isomorphyc (talk) 15:46, 7 June 2016 (UTC)

Hi JohnC5, sorry to keep bothering you about Greek stuff; I don't know whom else to ask. I edited the documentation slightly but significantly for Template:grc-IPA. It implied the module interprets all a,i,u as ambiguous, but I looked over the code and tested a few examples and am nearly certain it automatically treats circumflexes correctly as being long. I think the operative place this happens is line 394. I know you have worked quite a bit on this module; does this seem correct or incorrect to you offhand? If you're not sure, that's fine -- I'll double check. Partly I wanted to let someone know I had done this. Thanks. Isomorphyc (talk) 23:28, 24 June 2016 (UTC)

@Isomorphyc: Have no fear of pestering me! In AG, the acute accent represents a vowel or diphthong stressed on the final mora regardless of length (so ά could be á or a͜á), whereas a circumflex represents a vowel or diphthong stressed on the first of at lease two morae (so ᾶ must be á͜a and cannot be short). Therefore all vowels with circumflex accents are unambiguous. —JohnC5 23:57, 24 June 2016 (UTC)

Thank you! I was nearly sure, but wanted to be sure I was not overwriting something important. Isomorphyc (talk) 00:14, 25 June 2016 (UTC)

Hi JohnC5, a couple more questions; thanks for being open to this. 1) Would there be any reason for me not to add the usual macron/breve exclusions to Wiktionary:About Ancient Greek#Diacritics_and_accentuation? That is, omega, eta, omicron, epsilon, circumflex, iota subscript, i/u terminated diphthong. I happen to think the rules are a trifle exigent, and they are indeed honoured more in the breach than the observance. But it seems to me it must have been intended that only genuinely ambiguous vowels be marked. 2) I should probably know this, but what is the difference between the bracketed length annotations and those given in the main text of LSJ?

@Isomorphyc: 1) You are quite right—please add this information. Could you add it in some nicely formatted list of some sort? I think that would look good! :). 2) I assume you are dicussing a situation like Γάδαρα? In the LSJ, they tend not to put the length marks on the headword when they would collide with any other diacritics (and they sometimes put the information at the end of the entry if there is more complex discussion). Perseus does not render them very well, and if you have any doubt, I'd advise looking at this version for clarification. Also, they frequently leave out the length of α when it is the thematic vowel of the first declension since the length may be derived from declension. Is that what you were asking? —JohnC5 17:15, 26 June 2016 (UTC)

Ah, the bracket is just for typography. Very helpful. I was hesitant to include it in case it represented some kind of disagreement. I'm aware of the first declension alpha in LSJ, which is unambiguous. Indeed, a large plurality of such substantives in Wiktionary follow the LSJ practice. I was going to suggest adding these macrons mechanically, though if that's unwanted I still do know the vowels are long. The largest other omission is that I think unmarked ambiguous diphthongs in LSJ should probably be assumed to be both short. There are a small number of cases were one or both vowels are long, as noted in Vox Graeca, but I think these are usually marked with macrons. Since you linked TLG -- do you think they would mind if we linked Cunliffe like we link LSJ? I've had the tables necessary to do it for some time but I hesitated since it is copyrighted and I believe they have had server issues in the past with too much traffic. I do think it is much less unwieldy a resource for readers of Homer than LSJ. By a list in About Greek, were you thinking of something with bullets and examples? You do not think that would be too much detail for so short an article. Isomorphyc (talk) 17:49, 26 June 2016 (UTC)

@Isomorphyc: Please mark first declension alphas mechanically. I'd love to know your method for determining which are long or short. Is it merely testing for proparoxytonic stress?
Could you give me an example of what you are describing with the ambiguous diphthongs?
I had already created {{R:LBG}}, which has only recently come back online (If you'd also like to set up a table for that one, be my guest). I have multiple times in the past run into TLG's browsing limits due to server issues. It is, however, such an excellent set of resources that I use them as much as I can.
I bulleted list with examples would be perfect! —JohnC5 18:03, 26 June 2016 (UTC)

I actually test against paroxytone and augment it with double consonant or sigma in the stem and genitives in eta, and exclude the small list of exceptions in Smyth; the reason is that I am not entirely willing to trust my syllabation function yet. The result is unappealing but it seems to work. Also, thank you for mentioning the non-adjoining [^] or [_] notations; I was indeed omitting this. (Fortunately I do not have to worry about the Perseus rendering as I am using the underlying xml file-- which is itself quite retro looking xml). By diphthongs, frustratingly, Perseus fails to give a result for any of the examples I can think of. Here is one elsewhere: [3] The alpha in the diphthong is long. Allen mentions a few very esoteric examples but points out that long-alpha iota is surprisingly common. I feel all of these will be annotated, so I do not need to worry about diphthongs otherwise. I will add the tables and the bulleted list. Isomorphyc (talk) 23:33, 26 June 2016 (UTC)

el-adj

{{el-adj}} you asked "Is there a reason we're not using the built in headword?" - please can you expand.   — Saltmarsh^{συζήτηση-talk} 05:00, 16 June 2016 (UTC)

@Saltmarsh: Of course, and I hope you don't mind my editing this template. What I meant was that all of the functionality that {{el-adj}} possessed (displaying genders, displaying different gendered forms, adding categories) can be done by the {{head}} template itself. This also means that it will have a consistent look with other languages' headwords. To my mind, adding this material after the headword looks strange. I was hoping to change all of the headword templates for Greek over, if there were no objections. What do you think? —JohnC5 14:48, 16 June 2016 (UTC)

I should probably have started doing this some time ago. I'll go through the various POS, experimenting to see if there is anything that cannot be coped with - and come back to you.   — Saltmarsh^{συζήτηση-talk} 05:02, 17 June 2016 (UTC)

@Saltmarsh: Cool! Tell me if I can help! —JohnC5 05:25, 17 June 2016 (UTC)

• Having thought for longer I'm having second thoughts: (1) using {{head}} has the downside of requiring more keystrokes, a disadvantage even if boilerplates are used; (2) my main interest is modern Greek entries, with a el-specific HWL template I can create a temporary trap. I note that para:3 of Template:head/documentation talks about language specific templates. I would hate to do away with the el- ones and then discover another downside.

So I agree that common layout is highly desirable (but I've been here long enough to see the problems of getting consensus!). So I would like to hold fire for a while - meanwhile I will try to ensure that all the el-templates filter their data through {{head}}. And maybe revisit this later.  — Saltmarsh^{συζήτηση-talk} 09:02, 17 June 2016 (UTC)

@Saltmarsh: I think there may have been a miscommunication on my part. Did you think I meant we should get rid of the {{el-*}} HWL templates and just use the bare {{head}} template in the mainspace? If so, this was my fault for not explaining better. I would never espouse such an idea for reasons you've listed and others. I want to keep all the Greek HWL templates with the exact functionality, just a change in the backend and sometimes the appearance. What I am espousing are edits like this one to take the extra Wiki markup formatting after the call to {{head}} and integrate as much as possible into the headword template itself. Another example is {{el-adj-comp}}: the call to {{g|m}} and the formatting of the feminine and neuter declension parts should be passed into the head template not written out explicitly with afterwards. Does that make more sense? —JohnC5 12:10, 17 June 2016 (UTC)

:) yes - I had misunderstood you. I'll go through them over the next few weeks - the exercise will do me good. (Shame really, because changing to {{head}} would have made me update About Greek which is long overdue!)   — Saltmarsh^{συζήτηση-talk} 04:57, 18 June 2016 (UTC)

iuxta, De Vaan

Why should just De Vaan's opinion matter and what length does he mention anyway? The etymology section which is sourced with De Vaan mentions the Latin word "iūgera". When one assumes that "iuxta" and "iūgera" are related, wouldn't it be more likely that the vowel length is the same in both words?
Well, if De Vaan has a short u in "iuxta" and as it might be more common with a short u in English dictionaries, how about moving the form with a long u into a usage note like "Several sources state that the u is short.[references: Lewis & Short, Gaffiot, Oxford Latin Dictionary (1968), Langenscheidt] However, some sources state that it is long.[references: Georges, Pons, Allen and Greenough]"? Greetings, Poskim (talk) 16:37, 27 July 2016 (UTC)

I don't happen to have my De Vaan in front of me at the moment, but his opinion matters because his is the best respected work of Italic etymology currently in existence. Otherwise, I was merely asking for source which showed a long ū, which you have now provided. —JohnC5 17:09, 27 July 2016 (UTC)

But couldn't there be different opinions like another linguist having another opinion? (I couldn't cite any linguist for any length of the u, but maybe someone else could.)

Review of Michiel de Vaan's Etymological Dictionary of Latin and the Other Italic Languages mentions: "It is impossible to produce a book of this size without a few lapses, editorial or otherwise. There are typos (278, "distrubtion"), stylistic infelicities (10, "included into" for "included in"; 13, "shortly" for "briefly"; 585, "excrements"), and missing macrons (126, bidens for bidēns; 326, lanx for lānx)." (That doesn't say anything about his entry for iuxta.)

Walde's etymological dictionary (1910) has "juxtā", but also "ū?", that is he's not sure about the vowel length. Maybe the length is still unknown and maybe there are different opinions?

BTW: Other sources which have a long u: Allen and Greenough (1903), Lateinische Grammatik (2015), Comprehensive Etymological Dictionary of the Spanish Language, Latina Mythica II, latinlexicon.org. Well, searching for "iuxtā" instead gives a few more results, often English books, sometimes ebook editions. But if one would just go by the number of results, then both forms would exist.

-Poskim (talk) 17:54, 27 July 2016 (UTC)

@Poskim: Again, you seem to have misunderstood my position here. I'm fine with the claim that there could be a macron. Of the sources mentioned on the page at the time, only De Vaan's opinion I didn't know. I welcome the opinion of other dictionaries (references which you should add to the entry), but when this discussion began, no reference with a long u was mentioned. Please please please, add this discussion to the page. De Vaan is a great source, but certainly not the end-all-be-all. You just seemed very skeptical of why it was relevant, so I answered. Also, I still don't know what De Vaan's opinion is anyway because I don't have it in front of me. I was just curious what De Vaan says since it's already on the page. —JohnC5 18:06, 27 July 2016 (UTC)

Ok, thank you very much.

I added some references. If PONS is not ok (it's an online dictionary, but there are also printed dictionaries by PONS), feel free to remove it.

As it seems that more sources have a short u, it would be ok for me if "iūxtā" would be moved into a usage note. (I personally don't favor any length. I just had the impression that "iūxtā" is quite common, so that is worth mentioning.)

-Poskim (talk) 19:03, 27 July 2016 (UTC)

@Poskim: It looks fine as is, though I may fiddle around with the references a bit. What does interest me is how we acquired this knowledge of the length of iū̆xtā. Metrics wouldn't tell us the length of the vowel in a closed syllable, so I'm not sure how we would know. —JohnC5 19:48, 27 July 2016 (UTC)

Hi @Poskim, JohnC5: hope you'll excuse the intrusion. For what it is worth, Gildersleeve favours the long u, but De Vaan does not mark it. I could be misreading the entry, but he seems to make the derivation from the Proto-Italic *jougVsto- via an intermediate form: *jūgVsto-, implying the diphthong became a long vowel before the syncope occurred, whence iūgerum. Given the care with macrons in this entry, a typographical error seems less likely. The consonant cluster seems to be the only difference between these two words, and at a minimum it makes a discussion of the length of the u in iuxta difficult, and perhaps there is an implication it was responsible for shortening. If you are interested, his bibliography is: Walde-Hoffman I (1930-54): 737; Ernout-Meillet 1979: 328; Pokorny 1959: 508-510; Cowgill 1970: 125; Lexikon der indogermanischen Verben: *ieug-. -> iungo. The only journal article cited is Cowgill, which is: 1970 Italic and Celtic superlatives and the dialects of Indo-European, in: Indo-European and the Indo-Europeans, ed. G. Cardona e.a., Philadelphia: University of Pennsylvania Press, 113-153. Hope this is some help. This is an intriguing question for the indirectness involved in answering it. Isomorphyc (talk) 01:47, 28 July 2016 (UTC)

@Isomorphyc: Hello Isomorphyc. Of course I don't mind the "intrusion" (how you called it) and I'm thankful for your input.
Is it Basil Lanneau Gildersleeve (d. 1924)?
I don't know Walde-Hoffman, just the dictionary by Alois Walde (d. 1924). He has "juxtā" but together with "ū?" in his Latin etymological dictionary (Lateinisches etymologisches Wörterbuch, 2nd edition, 1910).

Walde: "juxtā, juxtim (ū?) „dicht daneben“: aus *jugistā, adv. zum Superlativ eines adj. *jūgos „eng verbunden“ ([Reference]) oder *jŭgos (eventuell ū durch g bewirkt, vgl. āctus? die roman. [= romanischen] Sprachen sagen über die Quantität des u nichts aus, s. [= siehe] [Reference]); nicht zum -es-St. *jeuges- von jūgerum, jūmentum."
I'm not sure what his abbreviations "adv." and "adj." are supposed to mean. Usually that should stand for adjectives or adverbs, but I have the impreasion that it stands for nouns here.
A rough translation: "juxtā, juxtim (ū?): from *jugistā, adverb to a superlative of an adjective *jūgos or *jŭgos (maybe ū because of the g, compare āctus? the Romanic languages don't reveal anything about the quantity/length of the vowel u); not from an -es-stem *jeuges- of jūgerum, jūmentum.".

Dirk Panhuis has a long u in his grammar (or at least in a German translation of his grammar). Here is a description of the English translation together with some information about the author and with a few comments.
-Poskim (talk) 12:39, 28 July 2016 (UTC)

@Poskim, Isomorphyc: Ok, having read this discussion and looked at De Vaan, I think that the ū seems very likely. If De Vaan's reconstruction of *jougVst(ād) us correct, then that would yield iūxta (You may find useful this chart I've been working on for a while. Please feel free to add to it). His other proposal *yéwg-os > *yéwg-s-to- > jougst(ād) would also lead to ū. Even in the bizarre of chance that something like *yug-s-tó- occurred, then it would give *jugst(ād) which would be subject to w:Lachmann's law, which would still give ū. —JohnC5 16:13, 28 July 2016 (UTC)

@Poskim: For Gildersleeve, it is indeed; please see here, for example: https://books.google.com/books?id=jo1iAAAAMAAJ&pg=PA269#v=onepage&q&f=false . @JohnC5: Your suggesting of applying Lachmann's Law makes some sense given De Vaan's attractive adjective/participle formulation in something like -stus/-stum/-sta from the comparison with oxen, but this is not a wholly clear case because: 1) the syncope alone may be adequate to weaken the consonant and alter the vowel; 2) the diphthong is intrinsically long so this is not a case of lengthening the /u/, but monophthongisation; 3) Lachmann's Law is more of a modest propensity, than a law, that is easier to apply in more controlled circumstances such as true participles. The analogy with iugerum is adequate to accept a long vowel result, given the PIE diphthong, which is short in the PIE derivation of iugum, explaining that difference in quantity. To me the question is whether it was shortened after that, and when, or contrariwise, if it even makes sense to talk about a long vowel in a syllable that already has a semivowel and three consonants. I suspect it may not. The argument for a short vowel, if I had to guess, is something like this, in skeletal form: /u:/ in Latin is usually preserved in the Romance languages, for example: fru:x -> es. fruto, fr. fruit, it. frutto ; lu:x -> es. luz, fr. (X), it.luce. But short vowels get altered at more points. So crux -> es. cruz, fr. croix, it. croce; nux -> es. nuez, fr. noix, it. noce. (A counterexample is dux, which is short, and perhaps usually preserves the /u/ due to continual reborrowing from Latin for so important a term; but note it is doge in Venetian, as nux -> nos, but fru:x -> fruto in Venetian). For these heuristic arguments, since we regularly find /ou/ in French, I think it would have been reasonable to assume, prior to the PIE reconstructions, that the /u/ was short in Late Latin or Vulgar Latin. How much earlier it may have been short is hard to say without other types of evidence. Very enjoyable chart, by the way. Isomorphyc (talk) 23:10, 28 July 2016 (UTC)

Thanks for seeing my delete request

Thanks for deleting the module I commented. Unfortunately there's neither a template to mark modules for deletion, nor a deletion `thank' link, as these are not precisely edits. Isomorphyc (talk) 16:33, 31 July 2016 (UTC)

@Isomorphyc: No problem! I try to keep an eye on all the modules in which I'm interested. —JohnC5 16:50, 31 July 2016 (UTC)

The strange -s in French alors

Where does it come from? --kc_kennylau (talk) 17:02, 2 August 2016 (UTC)

@Kc kennylau: According to the TLFi (it's the very last full sentence), there is an adveribal -s added to *lor from illā hōrā. —JohnC5 17:29, 2 August 2016 (UTC)

are there any other examples of adverbial -s in French ? Leasnam (talk) 17:33, 2 August 2016 (UTC)

Thank you for your response, but where does the adverbial -s come from? --kc_kennylau (talk) 18:02, 2 August 2016 (UTC)

ailleurs, ains (< ante). --WikiTiki89 18:14, 2 August 2016 (UTC)

There are a lot more that end in -s but already ended in -s in Latin. Perhaps that -s was generalized to other adverbs? --WikiTiki89 18:16, 2 August 2016 (UTC)

How about toujours ? Superficially it looks like tous + jours ("days", pl), but could it possibly be tout + jour + -s (i.e. all (the) day (long)/all the time, adv) ? Leasnam (talk) 18:28, 2 August 2016 (UTC)

It seems in OF it was toz jorz or tousjours, so it is plural. --WikiTiki89 18:36, 2 August 2016 (UTC)

@Kc kennylau: This is a bit unsystematic, but I think the so-called French adverbial -s comprises several small classes of common adverbs. Those derived from 1) Latin adjectives in -us (spissus -> épais, pressus -> près, (LL.) bassus -> bas); 2) Latin adverbs in -is or -us ((iam)magis -> (ja)mais, subtus -> sous, minus -> moins, foris -> hors); 3) Latin comparatives (peior -> peius -> pis); 4) French plurals (jadis, entretemps, toujours); 5) hypercorrections: umquam -> onc/onques/oncques; alio loco -> allieurs ; voluntarie -> volontiers ; n'a guère -> naguères/naguère ; ad hora -> lors/lorsque/alors. I do not know of any French sound change which systematically turned vowels into /s/. Isomorphyc (talk) 19:02, 2 August 2016 (UTC)

You can't call it a hypercorrection, because the -s used to be pronounced, and still is in certain derived terms like lorsque. --WikiTiki89 19:06, 2 August 2016 (UTC)

@Wikitiki89: It is still a hypercorrection if the hypercorrection is pronounced. --kc_kennylau (talk) 03:14, 3 August 2016 (UTC)

I would say that the "hypercorrection" would be this elusive "adverbial -s" which is an amalgamation of the other cases. I'd also ask whether Frankish possessed genitive-cum-adverbializer -s similar to the one in English that could have influenced the development of Old French. —JohnC5 19:22, 2 August 2016 (UTC)

Since it is largely unattested, it is hard to say for sure if the Frankish language had this feature, though we can speculate roughly due to other closely related languages such as German, Dutch, and English which all developed adverbial -s from the genitive case, so it's a possibility. Leasnam (talk) 19:40, 2 August 2016 (UTC)

Do any other Romance languages outside of traditionally Oïl-speaking areas show this tendency ? Leasnam (talk) 19:41, 2 August 2016 (UTC)

@Leasnam: One word just occurred to me: Latin ante > Spanish antes. --kc_kennylau (talk) 03:17, 3 August 2016 (UTC)

Is the Spanish -s a suffix ? Are there any other cases of it ? Leasnam (talk) 14:26, 3 August 2016 (UTC)

@Leasnam: There are después (de ex post) and quizá/quizás (qui sapit). Isomorphyc (talk) 16:46, 3 August 2016 (UTC)

How did the "s" change into "z"? --kc_kennylau (talk) 17:47, 3 August 2016 (UTC)

The gemination in Italian chissà and the cedilla in Portuguese quiçá both indicate that something extra was going on there at the word boundary. I don't know enough to say any more. --WikiTiki89 20:12, 3 August 2016 (UTC)

@Wikitiki89: The fact that "pit" was dropped in all three descendants tell us that quizás~chissà~quiçá may not have come from qui sapit. "sapit" became "sabe", "sa", "sabe" in those languages. (Alright, Italian dropped the "pit" also) --kc_kennylau (talk) 05:47, 4 August 2016 (UTC)

Thank you ! Isn't the s on después just from the s in Latin post though ? Leasnam (talk) 16:57, 3 August 2016 (UTC)

You're right. Isomorphyc (talk) 16:59, 3 August 2016 (UTC)

@Wikitiki89, Leasnam, Isomorphyc: Strangely enough, the adverbial -s is also present in English (Old English, Middle English, Modern English), with example words being "towards, once, always, and unawares". --kc_kennylau (talk) 06:08, 18 August 2016 (UTC)

With English, though, it is purely native, and as you say is easily demonstrated in the historical record of English Leasnam (talk) 15:50, 18 August 2016 (UTC)

Weird Etymologies

In view of your aspersions as to some of my Entry edits, I am about to revert those two of mine that are unsourceable; ie. fun (that I shall transfer to its Talk Page), and that of bras. Normally I take great care to ensure accuracy of all my edits so as not to mislead Users or Readers. My Main Entry edits are but a handful, and generally of a very basic nature, all sourceable, mostly from O.E.D., Professor Skeat, or Wiktionary itself - I would have added the sources (as in Wikipedia) if I had known that this was required. By deleting most of my edits on the worth page the Wikipedian also deleted a blatant error; so that page is more accurate than it had been for a long time! All edits of Entry pages that I had made are clearly visible on the Watch List, and all the rest therein are quite reliable. I have had to learn a deal more about etymological rules during the past 10 months in order to present true etymologies. To present an example of a lexeme that should demonstrate that I am not biased by Celtic origins; it is that of loop, where other etymologists connect it with Celtic lub (bend) - and I always had some difficulty about that derivation, without intermediate meanings - but it only took seconds for me to accept the Main Entry form, related to leap, whence "leap knot", that is clearly Germanic. I have had to patrol my Talk Page edits since I found a few inaccuracies therein. My due apologies for any inaccuracies in formatting, as well, and for the inconvenience caused! Even if most of the Entry edits are correct, uncorrected edits (after being pointed out) still left or unreverted, would be a blockable offence. Andrew H. Gray 12:21, 4 August 2016 (UTC)Andrew

@Werdna Yrneh Yarg I'm sorry for all this, and I mean you no ill will. Another suggesting for concerning talk pages is the existence of the "+" button between the "Read" and "History" buttons at the top of the page. This allows you to append a new discussion with title to a discussion page without interfering with any previous topics. Also, I advise using the "Show preview" button early and often, and proofreading before saving in order to prevent multiple edits to the same page in short periods. —JohnC5 14:56, 4 August 2016 (UTC)

@JohnC5 Thank you for this information. Andrew H. Gray 15:02, 4 August 2016 (UTC)Andrew

Reconstruction:Proto-Indo-European/h₂ŕ̥tḱos

Can you please explain this revision? Sir Nicholas de Mimsy-Porpington (talk) 16:10, 14 August 2016 (UTC)

My apologies, I should have probably left a better explanation. I arrived at the page referred above from ours, which under ours#French says: "From Middle French ours, from Old French urs, from Latin ursus, from Proto-Indo-European *h₂ŕ̥tḱos ‎(“bear”)." What do you think? Sir Nicholas de Mimsy-Porpington (talk) 16:23, 14 August 2016 (UTC)

@Sir Nicholas de Mimsy-Porpington: Of course. The etymology for the French ours is certainly correct. As for *h₂ŕ̥tḱos, the PIE *s stem *h₂rétḱ-os ~ *h₂rétḱ-es- (formerly represented on that page as *h₂rétḱes) would give the Sanskrit lemma form of रक्षस् (rákṣas) (remembering that the lemma form of Sanskrit nouns are meant to give declensional information, not necessarily show a form that actually exists. Cf. मनस् (mánas) < *ménos). You are correct that *h₂ŕ̥ would give Sanskrit ऋ (ṛ́), but *h₂ré gives र (rá). I'm more worried about the issue that, since the advent of laryngeal theory, the link between *h₂ŕ̥tḱos and रक्षस् (rákṣas) has been contested, with Kloekhorst, De Vaan, and Beekes all ignoring it or rejecting it. —JohnC5 16:37, 14 August 2016 (UTC)

Thank you for the response. Sir Nicholas de Mimsy-Porpington (talk) 09:58, 16 August 2016 (UTC)

DodgeBow rollback

JohnC5, regarding https://en.wiktionary.org/wiki/dodge_bow

This entry is an instance of copyright infringement as it is referencing a service without recognition to the trademark holder. Following, https://en.wiktionary.org/wiki/Wiktionary:Copyrights, Individual words are not copyrightable, although some may be trademarks, but this does not prevent [Wiktionary] from using these words, as long as [Wiktionary] is not producing a product to compete with the trademark holder.

Please consider your rollback and submit upwards for review if needed. If page must be removed, let this be the compromise.

Unfortunately for you, no company or copyright laws may control the natural development of language. If speakers of a language decide to use this word in this way, no organization or government may stop them. —JohnC5 20:28, 25 August 2016 (UTC)

I agree with you, but it is fair to give recognition to the existence of the trademark and not inadvertently put the trademark owner's registration in peril. Example of recognition: https://en.wiktionary.org/wiki/Nike - can you please consider the same for DodgeBow?

Adam M.

Why should we be responsible for a trademark owner's registration? They should advertise their own trade mark if they feel a need to, Wiktionary is not going to do it for them. —CodeCat 21:33, 25 August 2016 (UTC)

"Wiktionary does not [except when etymologically relevant] include information about whether or not a term is or has been trademarked. Wiktionary entries are not intended to provide a legal opinion as to the trademark status of words, or the legitimacy of any claim to rights in a word." per WT:TM. - -sche (discuss) 21:48, 25 August 2016 (UTC)

The purpose is not to advertise the trademark, but to protect it from genericization by giving credit where due. So much time and effort has been put into marketing the DodgeBow brand and promoting the activity; it is more than likely the reason why an entry "Dodge Bow" found itself on Wiktionary. The mark is registered and active in 29 countries. That is noteworthy. Would it be unfair to ask that an entry like "A brand of combat archery games and equipment" be added? Doing so, without making any reference to the company who owns the mark, would already be enough to protect its intellectual property. If not, why would that not be acceptable? I can readily provide all necessary the references to support such an entry. -Adam M. 01:36, 26 August 2016 (UTC)

It's not Wiktionary's responsibility to protect trademarks from genericization, nor to give anyone credit. It's not up to us to protect anyone's intellectual property. We document words, not IP. —CodeCat 01:48, 26 August 2016 (UTC)

If that is the case, can you please then explain the logic behind the double standard for many other brands such as Nike, and Adidas, for instance, which all contain a trademark reference on Wiktionary https://en.wiktionary.org/wiki/Nike - https://en.wiktionary.org/wiki/Adidas -Adam M. 14:19, 26 August 2016 (UTC)

One issue is that those terms are actually identical to the brand name. The name DodgeBow is not the same as dodge bow in capitalization or spacing. —JohnC5 14:39, 26 August 2016 (UTC)

If that is the only issue, it is incorrect to say that Wiktionary does not give anyone credit for trademarks, by simple evidence of the two examples links I gave above. Further, "dodge bow" is an entry made by someone who actually played or saw DodgeBow - a brand that provides Combat Archery Games and Equipment. The activity itself is Combat Archery, not "dodge bow". This entry is not correct. DodgeBow (or dodge bow, as entered by the original entry) is the name of a brand that provides Combat Archery activities. This is no different from the Adidas entry. - Adam M. 17:36, 26 August 2016 (UTC)

So, the pun "dodge bow" (a blend of dodgeball and bow) is not so obscure and fantastically creative to prove without a doubt that the pun arose only once. Indeed, there have been several Minecraft mods called dodgebow which represent the same premise (playing a dodgeball-like game with bows and arrows) and which were created 2 years before the copyright for DodgeBow was filed. This does not prove that this generic term for the real life game was created separately or before the copyright, but it does demonstrate how this leap could be made separately and that your claims that DodgeBow is necessarily the origin of the term dodge bow are hard to prove. As such, we could mention in the etymology section the possibility that dodge bow might be based off of the brand name, but we would need to add the caveat that the name had arisen separately before the copyright. —JohnC5 17:52, 26 August 2016 (UTC)

We've removed the trademark references at the other entries, which should not have been there. As for dodge bow, I've taken it to WT:RFV, because I see no evidence of use that meets our Criteria for inclusion. This is really the best way to deal with the issue, since such usage would be more damaging to the trademark than any dictionary entry, but lack of it is the only justification need for deletion. Attempting to turn a dictionary entry that's going to be deleted into a trademark disclaimer is just plain silly. Chuck Entz (talk) 18:27, 26 August 2016 (UTC)

Thanks Chuck, I appreciate that. If it's of any use at all, the original entry was made August 10 2015, several weeks after we incorporated the startup "DodgeBow Sports". Even more evident, the IP address of the entry is from Vieux-Saint-Laurent, a borough of Montreal. This is actually the city in which DodgeBow Sports is headquartered and where we first started our activities. - Adam M. 18:37, August 26 2016 (UTC)

Slavic Interior rollback

Hello, John. Since yesterday I keep trying to post a true information and enrich Wiktionary but it seems that censorship is fine here.

In Polish "henteros" changed to "ątr". "-os" dropped because Slavic masculine nouns and adjectives lost this ending. Greek "Αλεξανδρος" is "Aleksander", Latin "Marcus" is "Marek", "hyacinthus" is "hiacynt". So: henteros -> ątr. "Ą" comes from "en". This letter is pronounced like "on" in French "mon". It is nasal. After some time a prefix "wn" meaning "in" appeared. In Slavic languages (Baltic too) prefixes change the ending of the word. Usually words get "-e" that palatalize the stem: za + rzeka = zarzecze, po+jezioro=pojezierze. In Polish there is a phenomenon called Lechitic umlaut. According to its rules "ą" followed by soft consonants changes to "ę".

Finally: prefix adds a suffix to the word, suffix palatalizes the stem, due to the palatalization an umlaut appears: wn + ątr -> wn-ątr-e -> wn-ątrz-e (palatalization) -> wn-ętrz-e (umlaut).

Wnętrze means literally "inside the interior", in Greek it could be εν-εντερος.

Many Slavic languages lost their nasal vowels so the changes are even deeper like in Russian where "enteros" changed to "vnutrenniy" (vn + utrenniy, "u" comes from "en" but it's not nasal anymore).

The word "ątr" is extinct. Examples with prefix "wn/vn/u/вн" are present that is why I post them. They come from this stem, it was said by the professor of linguistics Jan Miodek. So what have I to do if I want to post this info? Does Holy Ghost have to reveal it and an evangelist write about it? If "wnętrze" does not come from "enteros" what it comes from then?

Here's Russian etymology: Происходит от др.-русск. нутрь ж. «внутренность»; ср.: укр., белор. нутро́, словенск. nȏtǝr «вовнутрь». Из *vъn-ǫtrь «вовнутрь», которое воспринималось как *vъ-nǫtrь. Связано с утро́ба, внутрь, ятро. Ср.: греч. ἔντερα мн. «внутренности», др.-инд. ántaras «внутренний», antrám «внутренности», авест. аntаrа- «внутренний», лат. interus — то же. Отсюда нутре́ц «плохо кастрированный жеребец». Использованы данные словаря М. Фасмера; см. Список литературы.

And Polish: Uświadamiając sobie ogólnosłowiański zasięg opisywanego dziś słowa, warto wiedzieć, że jego ostateczne korzenie wiodą do czasów praindoeuropejskich. Jest ono spokrewnione i ze staroindyjskim wyrazem antra „wnętrzności”, i z greckim enteron „wewnętrzny”, i z łacińskim internus „wewnętrzny”, interior „bardziej wewnętrzny, głębszy” (oczywiste pokrewieństwo etymologiczne z takimi formami, jak interna, internista, internować czy internat). Bo też prymarnym rdzeniem, z którego wywiedzione są słowiańskie i niesłowiańskie warianty, jest indoeuropejski enteros//onteros – „wnętrze”.

@N7-Vanyaquetta: As I mentioned at Reconstruction talk:Proto-Indo-European/h₁énteros#Considerations, I had placed some other reconstructions at *h₁én. I will also state again that PIE *eN > Proto-Slavic *ę, not *ą and furthermore that the sound *ą does not exist in Proto-Slavic. I worry though that you won't see this message since you didn't see the last one. Ah well, ever onward.
Here are a few more issues. To your point about censorship of "true information", you have made two errors. One mistake is your belief that anyone has the right to have their views shown in an entry—this is not the case. We only show that information that is accepted by the scholarly community, relevant to the article, and correctly formatted. In this we have not censored you because it is not your right to put whatever you like. This normally would not be an issue except that you also happen to be putting information in the wrong place. I have no doubt that several, if not all, of those forms descend from *h₁én and some ablaut form of *-ter- ~ *-tr-, but not *h₁énteros. When we say a term is a "descendant" of another, we make a claim that the descendant shares the exact form with regular sound changes applied along the way. This is not the case here. Not only have a Proto-Slavic prefix and suffix been added later to the PIE form, but the stem is not *-tero-, it is *-tr- which does not go under *h₁énteros. *h₁én was quite productive in its formations using *-ter- ~ *-tr- (including *h₁(e)n-tér, *h₁en-tr-om, and *h₁ón-tr-om), and it is from *h₁ón-tr-om that these forms eventually come. It is honestly baffling why you are on the warpath for your view that these terms go under *h₁énteros (which the evidence does not support), as opposed to one of its near relatives. —JohnC5 06:25, 4 September 2016 (UTC)

Passi/Passus

Good catch! Isomorphyc (talk) 21:14, 7 September 2016 (UTC)

@Isomorphyc: No problem! It's like when people come to me claiming that the plural of ictus is “icti”. :P —JohnC5 04:31, 8 September 2016 (UTC)

"ictus ‎(plural ictus or ictuses or ictūs)" — ObſequiousNewt — Geſpꝛaͤch — Beÿtraͤge 04:49, 8 September 2016 (UTC)

@ObsequiousNewt: Did I compliment you on your new signature? —JohnC5 04:51, 8 September 2016 (UTC)

Maybe. Surely, though, "ictūs" is not actually used in English? — ObſequiousNewt — Geſpꝛaͤch — Beÿtraͤge 04:53, 8 September 2016 (UTC)

I pronounce them differently, but I doubt I'd ever write it down with a macron... —JohnC5 04:54, 8 September 2016 (UTC)

I wanted to also change the username to VnterwuͤꝛfigesMolch, but I figured it would be better not to obfuscate myself. — ObſequiousNewt — Geſpꝛaͤch — Beÿtraͤge 05:25, 8 September 2016 (UTC)

I wish my name had an instance of a long S, R rotunda, or an umlaut, but no such luck. Maybe JohannCFuͤnf? —JohnC5 05:33, 8 September 2016 (UTC)

Re: "... it would be better not to obfuscate myself". That sounds like it could be quite painful. I have enough trouble with self-deprecating humor, which, as I'm sure you know, is quite messy. It's almost as bad as engaging in persiflage with unsuspecting bystanders. Chuck Entz (talk) 08:02, 8 September 2016 (UTC)

Module:sac-decl

What did you do, and why? — ObſequiousNewt — Geſpꝛaͤch — Beÿtraͤge 20:30, 11 September 2016 (UTC)

Okay, I see you've fixed the module error. I'm still curious, though. Also, the default of the wa parameter shouldn't be wa, it should just be nil. — ObſequiousNewt — Geſpꝛaͤch — Beÿtraͤge 20:36, 11 September 2016 (UTC)

@ObsequiousNewt: Yeah, sorry I broke it for a second, but now it's fixed. I used the parameters module which enforces parametric rules. So now, it will automatically treat use the first parameter as the stem (or if none is specified, use the pagename without the final a), set the |noem= parameter as a Boolean variable that defaults to false, and just catch whatever is in |wa=. Any other parameters (at the moment) will be rejected. Are there any other ways to form the stem in Fox besides by removing the final a? Also, what are all the possible values of |wa=? I was also trying to get rid of the extra line that is showing up after the table. Any insight on that? I hope you don't mind my helping out! —JohnC5 20:41, 11 September 2016 (UTC)

The wa= parameter overrides the contraction of CwaC > CôC. By default this occurs always except after /k/, so setting wa=ô causes it to happen even after /k/ and setting wa=wa causes it never to happen. — ObſequiousNewt — Geſpꝛaͤch — Beÿtraͤge 20:43, 11 September 2016 (UTC)

@ObsequiousNewt: Can we always detect the gender based on the ending (a or i)? —JohnC5 21:06, 11 September 2016 (UTC)

@JohnC5: Yeah, we can. I'll get to it accordingly. — ObſequiousNewt — Geſpꝛaͤch — Beÿtraͤge 19:47, 3 October 2016 (UTC)

Maybe discuss this sort of thing?

(cur | prev) 04:44, 16 September 2016‎ JohnC5 (talk | contribs)‎ . . (47,637 bytes) (+216)‎ . . (undo | thank) (cur | prev) 20:58, 15 September 2016‎ JohnC5 (talk | contribs)‎ . . (47,421 bytes) (+855)‎ . . (undo | thank) (cur | prev) 18:50, 15 September 2016‎ JohnC5 (talk | contribs)‎ . . (46,566 bytes) (+40)‎ . . (The Lexicon Leponticum (http://www.univie.ac.at/lexlep/wiki/Category:Word) does actually have some support for the v. The x is not attested.) (undo | thank)

Maybe discuss this sort of thing?

UtherPendrogn (talk) 19:31, 18 September 2016 (UTC)

Not that your edits were wrong, you're right. But even so, don't forget I made those edits for a reason (and in this case, unfortunately, it was me confusing the x's... And the wikipedia article for Lepontic said "v" was not used a letter, but it turns out the wikipedia article is wrong. The Lepontic Lexicon indeed has the correct script. UtherPendrogn (talk) 20:24, 18 September 2016 (UTC)

@UtherPendrogn: Yeah, I was gonna' write you a long response, but then I had to go to a wedding, so I thought I'd wait till you got unblocked. But to be clear on one point: Do we now agree that z and ś are represented by different characters in Lepontic? —JohnC5 02:06, 19 September 2016 (UTC)

Yes, the Wikipedia article on Lepontic Script has an erroneous form which I should have cross-referenced and double-checked. UtherPendrogn (talk) 06:18, 19 September 2016 (UTC)

Cool, cool. So does this also mean you are prepared for the use of the Ital Unicode block (which is named "Old Italic" but is designed for other languages as well) to encode inscriptional Lepontic lemmata? —JohnC5 15:02, 19 September 2016 (UTC)

Etymology of Old French autrui and nului

(Notifying Ungoliant MMDCCLXIV, Wikitiki89, Vahagn Petrosyan, Wikitiki89): Where might the -ui come from? I suspect alterum > *altru > altrui but I am not certain, as final -u seems to have neutralized long before Old French. However, the environment -tr- might have kept it. --kc_kennylau (talk) 13:11, 19 September 2016 (UTC)

@Kc kennylau one theory is that they come from the genitives alterius and nullius. Another theory for autrui is that it comes from alter-huic. — Ungoliant ^(falai) 01:01, 16 October 2016 (UTC)

Greek verb headword line template

When you edited {{el-verb}} back in June you introduced two (there may have been others) new arguments "f2nolink" and "f6nolink" - please can you tell what they refer to. Thanks — Saltmarsh^{συζήτηση-talk} 06:09, 28 September 2016 (UTC)

@Saltmarsh: Of course! It was meant to make it so that, if |past=- or |passive=-, the resulting ― would not link to anything. I seemed to have failed last time, but now it is working correctly. Does that make sense? —JohnC5 06:35, 28 September 2016 (UTC)

Thanks - it makes perfect sense, I hadn't come across "f-nolink" before. Cheers — Saltmarsh^{συζήτηση-talk} 04:47, 29 September 2016 (UTC)

Erroneous diphthongs

Could you help me go through Category:Kenny's testing category 3 correcting the erroneous ei/eu diphthongs, thanks. --kc_kennylau (talk) 17:28, 3 October 2016 (UTC)

@Kc kennylau Of course. For the moment, we're just adding . between the offending vowels? —JohnC5 17:31, 3 October 2016 (UTC)

Yes. --kc_kennylau (talk) 22:34, 3 October 2016 (UTC)

@Kc kennylau: Status report: I've done A-C. There are a fair number that exist in New Latin/taxonomic entries. When I find these, I just remove the pronunciation template all together, since they are not Classical or Ecclesiastical. I've only found a few where the eu diphthong is correct. —JohnC5 04:23, 4 October 2016 (UTC)

What about cretaceus? --kc_kennylau (talk) 13:41, 4 October 2016 (UTC)

@Kc kennylau: Here are some general pointers. If L&S or Gaffiot have -ĕus, this should be understood as -ĕ.ŭs. This includes all adjectives in -āneus, -āceus, and most native words in -eus. The only examples words I've found thus far (A-E) fall into two categories:

• Borrowed Greek diphthongs: arce͜uthinus, Caphāre͜us, Die͜uchēs, e͜ugenēus (note that the first eu is a diphthong and the second is not because of the hiatus caused by long ē)
• Short words that experience poetic contraction: dĕĭn, de͜in (dĕĭnde, de͜inde, dĕĭnceps, de͜inceps)

It seems to me that this first category does not occur with ei as Greek ει becomes Latin ē or ī and the second category does not occur with with eu based on the evidence of dĕŭs and mĕŭs. Not having check everywhere, I do not know this for certain yet. I might even propose that we should either throw a warning when "eu[ms]$" occurs that one must explicitly say whether the vowels are a diphthong or separate. Or better mark all "eu[ms]$" and "ei" as separate unless a diphthong is specified somehow (perhaps {{la-IPA|caphāre=us}}, {{la-IPA|de=inceps}} or something like that). Of course we still need to go through this list to check anyway. —JohnC5 15:30, 4 October 2016 (UTC)

There's evidence that the possessive adjectives were either 1 or 2 syllables in free variation. In several Romance languages, the accusative of these adjectives appears with final -n, such as French mon, French ton, French son etc. Final -m was preserved as a real consonant and did not become a nasal vowel after stressed syllables, compare rem > French rien. The fact that the possessive adjectives have final -n in some descendants shows that they were also monosyllabic at some point. —CodeCat 16:01, 4 October 2016 (UTC)

@CodeCat: Good to know! Do you agree, however, that "eu[ms]$" and "ei" should be made set to be disyllabic unless otherwise specified? —JohnC5 20:36, 4 October 2016 (UTC)

I suppose it makes more sense to use as a default, but while we have a notation . to force breaking up a sequence, we have no notation to force to combine it. —CodeCat 20:43, 4 October 2016 (UTC)

We could use =, split on "eu[ms]$" and "ei", then remove ='s and have the module progress as normal. —JohnC5 00:16, 5 October 2016 (UTC)

Hmm, I noticed this discussion in my watchlist. = would run into problems because it's used in template syntax. What if an underscore _ were used, for its similarity with the undertie? — Eru·tuon 01:16, 5 October 2016 (UTC)

Oh of course = can't be used! Yeah, underscore looks fine to me. —JohnC5 01:24, 5 October 2016 (UTC)

@Kc kennylau, CodeCat: Ok, I'm done sorting through the list. Those terms which remain in the list are true diphthongs or, more often, false positives. May we now discuss ways of automating this? I would then probably like to remove all the .'s I've just put in. —JohnC5 15:07, 5 October 2016 (UTC)

I don't think it is possible. Consider Caphareus which has an eu diphthong. Usually we would expect -eus to not be diphthong. Also, deinde may have the ei diphthong or not, and both are valid. --kc_kennylau (talk) 17:02, 5 October 2016 (UTC)

I think it's unlikely that Caphareus had a diphthong. In all the other case forms, the -e- is a syllable of its own, so for it to differ in the nominative would be odd. —CodeCat 17:05, 5 October 2016 (UTC)

@Kc kennylau: I have a solution for this problem, but it's not perfect. Before anything else, you have two rules that split "eu[ms]$" and "ei" with a ., and then you have a line that removes _. If an _ is inserted in the call, it will prevent the vowels from being split and then create a diphthong. —JohnC5 19:32, 5 October 2016 (UTC)

@kc_kennylau, JohnC5, CodeCat, Erutuon: What do y'all think of using the character tie in our headword lines, much like we currently use macra? The OLD has “Caphāre⁀us” (well, strictly, “Caphēre⁀us¹ ⁓eī, m. (Caphār-).”), for example. — I.S.M.E.T.A. 17:43, 5 October 2016 (UTC)

Ew. —CodeCat 17:46, 5 October 2016 (UTC)

I'm not loving the way the tie displays on top of a space instead of over the eu, as it should. — Eru·tuon 01:12, 6 October 2016 (UTC)

That's a font issue. --WikiTiki89 14:12, 6 October 2016 (UTC)

@CodeCat, Erutuon, Wikitiki89: "Ew" is right. Do you know how we can fix the font issue, Wikitiki? — I.S.M.E.T.A. 13:29, 19 October 2016 (UTC)

Sorry, I misread. Actually, you're using the wrong character. You are using U+2040 (CHARACTER TIE), when you should be using U+0361 (COMBINING DOUBLE INVERTED BREVE). This would give: Caphāre͡us. Now, the tie displays over the wrong characters (for me, at least), and that is a font issue. How can you fix that font issue? By getting a job at Microsoft and working your way into the team that develops Arial Unicode MS. --WikiTiki89 14:51, 19 October 2016 (UTC)

@Wikitiki89: My mistake! Thanks for the correction; that now displays properly (at least for me). @kc_kennylau, JohnC5, CodeCat, Erutuon, Wikitiki89: So, what do y'all think of “Caphāre͡us”? — I.S.M.E.T.A. 15:22, 19 October 2016 (UTC)

@I'm so meta even this acronym: It looks fine to me now. Perhaps my browser doesn't use Arial Unicode MS? (I use Google Chrome and Windows 10.) — Eru·tuon 15:44, 19 October 2016 (UTC)

@Erutuon: Thanks of the feedback. @Wikitiki89: I use Mozilla Firefox with Windows 8.1 and my default font is Quivira. — I.S.M.E.T.A. 16:36, 19 October 2016 (UTC)

Is there a very small and narrow version of the mid-dot? Whatever it is should be almost invisible, but still make it look like the two letters are separated by something. Chuck Entz (talk) 03:20, 6 October 2016 (UTC)

@Chuck Entz: Do you mean that for instances of (disyllabic) -ĕŭ-? — I.S.M.E.T.A. 13:29, 19 October 2016 (UTC)

You know, if you are going to change the format, you should have told me before you went through every page... --kc_kennylau (talk) 10:58, 6 October 2016 (UTC)

@Kc kennylau: Yeah, I didn't know whether it would be possible until I was part of the way through, and then I wanted to do all of them just to be sure. Thanks for the change! Could we now track everything that has manual syllable breaks so that I can see what's unnecessary? —JohnC5 17:05, 6 October 2016 (UTC)

Already done. --kc_kennylau (talk) 09:50, 7 October 2016 (UTC)

Ahem!

You haven't checked CAT:E since you edited Module:grc-decl yesterday, have you? Just askin'... Chuck Entz (talk) 02:03, 11 October 2016 (UTC)

@Chuck Entz: Fixed. — ObſequiousNewt — Geſpꝛaͤch — Beÿtraͤge 02:12, 11 October 2016 (UTC)

@Chuck Entz: I'm honestly flattered that you think I'm clever enough to be the person in charge of mod:grc-decl! —JohnC5 02:16, 11 October 2016 (UTC)

Silly me. I saw you and the Newt exchanging pleasantries and overlooked the fact that it was your "pied à wiki" and not his. Fortunately he was paying attention and it worked out in the end, etc, etc. Now excuse me while I slither into the hole I just dug for myself and make a sincere, but pathetically unsuccessful attempt to remain unnoticed for a year or two... Chuck Entz (talk) 06:01, 11 October 2016 (UTC)

*ḱank-

@JohnC5, do you think you could create a *ḱank-/*ḱak- (“branch”) PIE entry for me? It would save me the trouble from having to write out all the cognates. I believe *ḱank- is the nasalized form of *ḱak-. If not, I'll just take a crack at it myself and you can correct my work. ;-) --Victar (talk) 16:53, 17 October 2016 (UTC)

You realise a is a highly dubious vowel in PIE, right? You should try to find another reconstruction. —CodeCat 16:59, 17 October 2016 (UTC)

@Victar: I had intended to mention this, but yes, anything with the vowel *a is normally considered foreign or highly irregular. I'll try to look into this later. —JohnC5 17:34, 17 October 2016 (UTC)

I'm going off Mallory's reconstruction. Matasociv cites *ḱonk-, *ḱok- for PCelt, but I don't know if that works for all its descendants. --Victar (talk) 17:53, 17 October 2016 (UTC)

Also, I forget where we discussed it before, but the leap to *ganskyos still seems very speculative to me. —JohnC5 18:19, 17 October 2016 (UTC)

Sigh. You can find that here, but there are plenty of *ḱank-, *ḱonk- descendants to warrant an entry, regardless. --Victar (talk) 20:05, 17 October 2016 (UTC)

Special:Diff/41144809

Hi John. Thanks for fixing that link. Do you have any idea why that's necessary? I.e., what changed to break usage like {{R:LSJ|tri/braxus|w=τρίβρᾰχυς}}? — I.S.M.E.T.A. 13:42, 19 October 2016 (UTC)

Oh, never mind; it turns out that it's not necessary after all. Whatever was broken must've got fixed. :-)  — I.S.M.E.T.A. 14:03, 19 October 2016 (UTC)

Yeah, I was confused about that before. I'm glad it's working again! I bet it was this. —JohnC5 14:10, 19 October 2016 (UTC)

You might well be right. Lua code's as clear as mud to me! — I.S.M.E.T.A. 14:32, 19 October 2016 (UTC)

Yep. I requested @Isomorphyc's help at Talk:ἅλς, where {{R:LSJ}} was not working. I also have trouble understanding that Lua code, even though I know what the and and or do in Lua, and so on. I think the logic needs to be simplified somehow. But I would be afraid to do anything lest I break one of the many instances of the various templates depending on the module! — Eru·tuon 00:45, 22 October 2016 (UTC)

@Erutuon: Thanks for flagging up the problem. & @Isomorphyc: Thanks for fixing the problem. — I.S.M.E.T.A. 14:07, 22 October 2016 (UTC)

*ǵʰéyōm

Can we be really sure about all the forms, in particular the nominative dual and plural forms? Keep in mind that this is the 3-stage ablaut pattern of the so-called "Leiden model", which is far from universally accepted. —CodeCat 23:48, 21 October 2016 (UTC)

Yeah, I was wondering about this. It just seems weird to have the nominative, genitive, and accusative listed without giving a table. Honestly, our reconstructed lemma form takes a fairly strong stance on the declension, even if by implication. Perhaps we should remove the table, but it seems odd to list all the critical forms without committing to a table. I do agree that the dual and plural are very speculative, and I'll admit that I didn't look up all the different proposed reconstructions like I normally do. Would you like me to do a survey of the usual sources? —JohnC5 23:58, 21 October 2016 (UTC)

I'm curious if all three ablaut grades are actually attested. —CodeCat 00:23, 22 October 2016 (UTC)

So LIN has three base inflection types, *ǵʰéy-om-, *ǵʰey-m-en-, and *ǵʰey-óm-, and a whole bunch of derivatives. *ǵʰéy-om- seems to be the primary. *ǵʰéyōm might be seen in χιών (khiṓn). Locative *ǵʰy-ém-i in Hittite [script needed] (giemi, “in winter”). Genitive *ǵʰimés in Old Avestan 𐬰𐬌𐬨𐬋 (zimō). Accusative *ǵʰyḗm in Young Avestan 𐬰𐬌𐬌𐬄𐬨 (ziiąm). That hits most of them. The weakest is definitely the nominative. —JohnC5 00:59, 22 October 2016 (UTC)

The Ancient Greek form has an -i- which is a bit odd. -eyō- would be expected to just lose its y and become -eō-, like in the causative verbs. At the same time, *ǵʰyō- would have palatalisation, giving Greek *zōn instead. Maybe if the i was syllabic in Pre-Greek times already, it could have given the attested form. —CodeCat 01:34, 22 October 2016 (UTC)

Beekes calls for *ǵʰiōm for the nominative of Greek and Armenian ձիւն (jiwn) and *ǵʰiōms for Avestan 𐬰𐬌𐬌𐬃 (ziiā̊). He takes the Latin hiems from accusative *ǵʰi-em-m. He makes no attempt to explain the lack of palatalization in Greek and also seems to ignore several PIE phonological rules in his reconstructions. —JohnC5 02:36, 22 October 2016 (UTC)

As per our discussion about Sievers' Law, I realize that Beekes recocnstructions could be accounted for with Lindeman's Law. —JohnC5 18:30, 1 November 2016 (UTC)

How do you know?

You always seem to be aware of every movement in the Celtic conlangs. Is there somewhere I should click a watch button? I keep missing every discussion, including ones that actively concern me. Thanks for pinging me the other day though. UtherPendrogn (talk) 22:09, 2 November 2016 (UTC)

@UtherPendrogn: You use your Watchlist, right? The watch button for a page is the star between the "History" and "More" tabs in the top right corner. I watch every page I edit (you can set that in preferences). Also, make sure you watch the requests and main discussions pages for every month. You can watch categories to see when pages or added or subtracted (I watch things like Category:Proto-Celtic lemmas). Does that make sense? —JohnC5 03:42, 3 November 2016 (UTC)

Is it a prefix or a separate word?

What do these words have in common? introduco, introeo, introfero, introgredior, introiugus, intromitto, introrumpo, introspicio. --kc_kennylau (talk) 05:16, 6 November 2016 (UTC)

I believe the normal joke response to this one would be “yes”. I'd say that it is a word that has be lexicalized as a prefix. Normally, I'd go with a prefix, but I like affixes more than I should. :) —JohnC5 05:20, 6 November 2016 (UTC)

@kc_kennylau, JohnC5: Interestingly, the Oxford Latin Dictionary has no entry for *intrō- and its etymologies for all those words claim their derivation from the adverb intrō. It also notes that many of the words so formed can be written as phrases, namely intrō currō for intrōcurrō, intrō dō for intrōdō, intrō dūcō for intrōdūcō, intrō eō for intrōeō, intrō ferō for intrōferō, intrō mittō for intrōmittō, intrō vocō for intrōvocō, and intrō vorsus for intrōvorsus (there's also intrō spectō for introspectō, though that one derives from the verb intrō). — I.S.M.E.T.A. 08:14, 8 November 2016 (UTC)

De Vaan has it, in the entry for inter, as an adverb deriving from that preposition. Isomorphyc (talk) 12:12, 8 November 2016 (UTC)

Palatal approximant gemination

When words containing opening diphthongs + vowel were borrowed from Ancient Greek to Latin, does the palatal approximant of the opening diphthong become geminated? E.g. Troia, Cāiēta. --kc_kennylau (talk) 05:47, 6 November 2016 (UTC)

Well, one of the origins of intervocalic iota (or subscripted iota) in AG is from Proto-Hellenic *yy (e.g. *ōyyón). —JohnC5 06:29, 6 November 2016 (UTC)

What about Troia and Cāiēta? --kc_kennylau (talk) 07:41, 6 November 2016 (UTC)

Much to my annoyance, this element of Latin comparative phonology is unknown to me. Nor do I know where to look. Maybe @CodeCat, Isomorphyc might know? —JohnC5 07:47, 6 November 2016 (UTC)

I believe these were geminated in Ancient Greek already, so it's reasonable that Latin would borrow that feature, especially as it generally occurred in Latin words too. —CodeCat 13:09, 6 November 2016 (UTC)

So you're saying that the spelling ⟨V[ij]V⟩ in Latin represented /Vj.jV/ as well? What is the evidence of this out of curiosity? —JohnC5 18:40, 6 November 2016 (UTC)

w:Latin spelling and pronunciation has more on it. We have it in our entries too, like cuius. —CodeCat 18:44, 6 November 2016 (UTC)

Thanks! —JohnC5 18:57, 6 November 2016 (UTC)

From a historical point, all /j/ were lost between vowels in the Proto-Italic period, so that it only remained word-initially and after a consonant. Thus, all Latin instances of /j/ must be traced back to either of these two possibilities. Since cuius is not a compound of some word with another starting with /j/, the first explanation can't work, so that implies that it must be the second: there was once a consonant before the /j/. It is assimilation of this consonant (in the case of cuius it was /s/, < PIt *kʷosjo) that produced a geminate /jj/. —CodeCat 19:03, 6 November 2016 (UTC)

@Kc kennylau: Is this information sufficient? —JohnC5 19:08, 6 November 2016 (UTC)

Edited: I am sorry for the confusion I left earlier. My results with scansion have been somewhat inconclusive and are not worth presenting. I think it suffices to say that diphthong followed by vowel is extremely rare in Latin and true oe/oi diphthong followed by vowel only exists in Greek borrowings. The borrowed words in idealised Greek are geminate, with some caveats; the descendents in Italian are geminate; and the most analogous native words are geminate. I agree with others there is some gemination, but although Allen is emphatic that gemination is the norm in Latin, he does not address the point of gemination of diphthongs before vowels in Latin directly, though he does address it in Greek. Isomorphyc (talk) 03:53, 7 November 2016 (UTC)

@CodeCat: How about compound words? --kc_kennylau (talk) 10:22, 7 November 2016 (UTC)

Flood flag

90% of my watchlist screen is taken up with your AWB changes. I note that you have the flood flag. Does that not work with AWB? It should and I thought it did. DCDuring TALK 15:02, 8 November 2016 (UTC)

Interested third party here. I have not been able to flag an edit as "bot" through the API unless the account is already flagged as "bot" or "flood" (which is just an alias for bot as far as an edit is concerned). If there is a way to flag an individual edit as bot or flood I would love to know about it. If not I think it would be a good bug to submit for a change to the API. - TheDaveRoss 15:08, 8 November 2016 (UTC)

@DCDuring: I'm sorry for the inconvenience. Flood flag is primarily for things like Special:RecentChanges. If you have watched a page, the flood flag won't prevent you from seeing changes. @kc kennylau and I should be done with this first pass soon. Unfortunately, there's going to be a second pass probably later today. Again, sorry for the annoyance, but this has been a long time coming. (Only 90%? You need to edit more Latin! :P) —JohnC5 15:09, 8 November 2016 (UTC)

Apparently I caused my own inconvenience. I wasn't hiding bot changes on my watchlist. I think that covers the offending changes by all parties. Sorry, but I didn't really associate AWB with bots, though I should have. DCDuring TALK 15:14, 8 November 2016 (UTC)

Would you like me to inform you when you are done so you can add bot edits back to your list? —JohnC5 15:17, 8 November 2016 (UTC)

That's more than I want to ask for. I just need to get in the habit of looking at the boxes when watchlist loading seems slow. That would cover all such users. In any event, if my diagnosis below is right, there is a tradeoff between speed and the amount of page clutter. DCDuring TALK 18:14, 8 November 2016 (UTC)

Does this make sense? It seems that, when there are a lot of bot entries (or other filtered-out entries) AND I have selected the "grouping changes" preference, watchlist takes longer to process than:

1. when there are fewer filtered-out changes
2. if fewer items are filtered out.

? DCDuring TALK 15:26, 8 November 2016 (UTC)

I don't know the answer to that question. —JohnC5 15:29, 8 November 2016 (UTC)

Re Special:Diff/41493763

Thank you very much for your part in getting rid of that literal waste of space. :-)  — I.S.M.E.T.A. 23:50, 8 November 2016 (UTC)

@I'm so meta even this acronym: It's been driving me crazy for so long. The fringe benefit is that now I am watching every Latin nominal currently in existence on Wiktionary. @Isomorphyc, kc kennylau, going through all this, I noticed some serious issues that could be rectified with bots/bulk research. One is that many words in long -ō are listed as short -o due certainly to L&S and Gaffiot's annoying habit of ignoring final vowel length. It is nearly impossible for a word to end in short -o without borrowing (I honestly can't think of any examples off the top of my head), and a list of the terms with short -o on here would be nice. Also, there are a fair number of capitalized nouns that are listed as nouns and not proper nouns. If we would get a list of that, it might be nice to clean them up. —JohnC5 18:06, 9 November 2016 (UTC)

Thank you very much, to both of you, for fixing this template! Here is a list of (probably) missing macrons: User:OrphicBot/Sandbox/Latin Short o Terminated Entries (236 items), and here are the proper nouns labelled common: User:OrphicBot/Sandbox/Capitalised Latin Titles without Proper Noun Entries (4647 entries). A great many of the latter were mechanically inflected with the less specific heading. In the former list, at least one of the headwords, for octavodecimo, is not actually missing the macron, but is misspelled. Isomorphyc (talk) 20:30, 9 November 2016 (UTC)

@Isomorphyc: false positive? --kc_kennylau (talk) 10:08, 10 November 2016 (UTC)

Not quite-- it is misspelled: trīsissimō, hence, as with octavodecimo, it wasn't a match. Isomorphyc (talk) 11:09, 10 November 2016 (UTC)

@Isomorphyc: Could you make a list of misspelt entries? Thank you. --kc_kennylau (talk) 13:16, 10 November 2016 (UTC)

@Kc_kennylau: Thank you for fixing the adverbs mechanically. Do you mean misspelt headwords in general, over all of Latin? I can probably do that. I wouldn't recommend doing it for this list alone because there are only 121 items and they should probably be handled manually since other macrons may be missing. A bit of a question about the adverbs: a large number are marked as not comparable which I would not particularly think of as not comparable. For example, imperiose or localiter. Is this a semantic distinction, or one based on attestation? Isomorphyc (talk) 13:50, 10 November 2016 (UTC)

@Isomorphyc: I think it is based on attestation. By the way, imperiōsius (imperiōsē) is attested but I cannot add it at the moment due to technical reasons. Will add it as soon as I can. --kc_kennylau (talk) 13:55, 10 November 2016 (UTC)

@kc_kennylau Here is a list of misspellings, including punctuation and html mismatches between the headword and the title, and a one or two other related errors. User:OrphicBot/Sandbox/Potential Latin Misspellings. This list only encompasses lemmas, from a set of about 23,000 applicable entries. If there are non-lemma headword misspellings I will post a list soon; but I do not have it yet. I'm also not totally comfortable with the policy of using comparability to mean attested comparability, because this creates misleading results for almost all rare adverbs. Moreover, attestation would often have to be manually verified because the comparative shares its form with several other possibilities, and it is not usually possible to create an unambiguous parse tree. That said, I don't really have a better idea other than letting users guess comparability on a sematic basis, and perhaps writing unattested forms in plain text (rather than linked). Isomorphyc (talk) 17:12, 10 November 2016 (UTC)

@Isomorphyc: What about including them in the template, but prefix them with *, the established marker for hypotheticality / lack of attestation? — I.S.M.E.T.A. 12:36, 23 November 2016 (UTC)

@I'm so meta even this acronym: I think some way to make the distinction between grammatical type and known attestation is important. Out of our 814,262 Latin forms, I am finding attestations for approximately 150,000. I suspect if we used Wikisource as a corpus the number might be 140,000. A list of such clearly attested strings could conceivably be stored in modules to annotate headwords in some way. I would not be in favour of using the asterisk, because it broadens too much the idea of reconstruction. But a footnote might not be a bad idea; it might read: `No attestation for this form has been found in Wikisource.' The major caveat is that by doing this we would likely be excluding most of the Mediaeval and later corpus; the minor caveat is that we will need to implement a way to handle texts which use consonantal u and i, because our convention is to ignore these forms. Also, thank you for the whitelist nomination. I will work on the spellings soon. I was very exigent in the BP, so I was able to get the editing permissions a day or two earlier. But I technically still did need the nomination, per the vote procedures, so it is much appreciated. Isomorphyc (talk) 15:00, 23 November 2016 (UTC)

@Isomorphyc: Re “I would not be in favour of using the asterisk, because it broadens too much the idea of reconstruction.”, why do you think this? — I.S.M.E.T.A. 17:47, 29 November 2016 (UTC)

@I'm so meta even this acronym: Wiktionary differs from paper dictionaries in that we include entries for inflected forms, not only for lemmas. As a result, almost all of our entries are non-attested in highly inflected languages on a per-form basis. I feel the asterisk should be used only for material ambiguity, not mere lack of attestation, because the latter criterion cannot be usefully generalised. If application to non-attested forms is generalised, the mark will lose much of its force for phonetic reconstructions and similar situations of more importance. On the other hand, other users have suggested word frequency templates in the past. If others are interested it would be easy to make something like this for Latin or Greek, both for lemmas and for individual forms; such a template would mark non-attestation of forms in a more useful context I think. Isomorphyc (talk) 18:00, 29 November 2016 (UTC)

@Isomorphyc: I probably wouldn't've agreed before, but I think I do, at least a little bit, by now. See WT:RFV#Forms of the Latin synaeresis. — I.S.M.E.T.A. 14:32, 6 December 2016 (UTC)

Thanks so much! I'll look into this soon. I'm also going to try to remove unnecessary parameters from the remaining headword templates (e.g. {{la-suffix-noun}}). —JohnC5 21:47, 9 November 2016 (UTC)

@Isomorphyc, I'm so meta even this acronym, kc kennylau: if people would be interested in helping, Category:John's testing category 1 currently contains adverbs needing to change over to the new format I have created. Place the full lemma with macra in first param. If the adverb is regular and comparable, the template will detect the type automatically. If the adverb is not comparable, put - in the second param. And if the adverb is irregular, put the comparative in the second param and superlative in the third. I would really appreciate some help because this process is a bit painstaking! —JohnC5 07:39, 10 November 2016 (UTC)

@JohnC5: It looks like I missed that party; that category now just has fourteen German terms. — I.S.M.E.T.A. 12:39, 23 November 2016 (UTC)

Some of the typos may be prevented by an edit to Module:headword. This edit would add a check to see whether the given headword, when diacritics and links are removed, matches the page name. —CodeCat 21:18, 10 November 2016 (UTC)

@CodeCat: I'd support such a change. —JohnC5 22:59, 10 November 2016 (UTC)

Someone needs to unprotect the module then. —CodeCat 23:23, 10 November 2016 (UTC)

@CodeCat Would something like this work around line 404?

function no_diacritics(x) return mw.ustring.gsub(mw.ustring.toNFD(x),'[^%w]+',"") end
if type(data.heads) ~= "table" and no_diacritics(data.heads) ~= no_diacritics(mw.title.getCurrentTitle().text) then 
  table.insert(tracking_categories, "head tracking/headword pagename spelling mismatch")
  require("Module:debug").track("head tracking/headword pagename spelling mismatch")
end

Thanks, Isomorphyc (talk) 16:06, 18 November 2016 (UTC)

Edit: I know this is wrong, at minimum because it is not removing links. It is a little bit difficult to test, but I will try to add this soon. I also just fixed a parenthesis error. Sorry this is so ad-hoc. Isomorphyc (talk)

Whatever code we use, it should come after preprocess is called on line 408. This function provides default values to various parameters, including the headwords themselves. That way, you don't need to check if data.heads is a table, because it always will be. In fact, the code could be placed within the preprocess function, after line 220 perhaps. Something like this:

for head in ipairs(data.heads) do
	if mw.title.getCurrentTitle().prefixedText ~= require("Module:links").getLinkPage(head, data.lang) then
		require("Module:debug").track("headword/pagename spelling mismatch")
	end
end

—CodeCat 16:32, 18 November 2016 (UTC)

That looks vastly cleaner than mine; thank you. Do I understand correctly it is unnecessary to add to the local tracking_categories table in show_headword_line(), as I did? Isomorphyc (talk) 16:49, 18 November 2016 (UTC)

My code uses tracking templates. They're more convenient when working with modules. —CodeCat 17:06, 18 November 2016 (UTC)

I thought what I was doing was slightly redundant. Thank you so much for your help. Isomorphyc (talk) 17:08, 18 November 2016 (UTC)

I'd also make this slight emendation:

for head in ipairs(data.heads) do
	if mw.title.getCurrentTitle().prefixedText ~= require("Module:links").getLinkPage(head, data.lang) then
		require("Module:debug").track("headword/pagename spelling mismatch")
        break
	end
end

So, do we intend to do this? —JohnC5 18:19, 18 November 2016 (UTC)

The break-statement seems like a good change to me. I would normally do a lot more testing than I am able, since I am not very familiar with this module. I'm sorry to impose on both of you like this, but if everything seems all right with @CodeCat, would you mind committing the update? Isomorphyc (talk) 18:29, 18 November 2016 (UTC)

Also, where is the output of the debug module? I've never actually found it before. —JohnC5 18:35, 18 November 2016 (UTC)

They are subpages of Template:tracking, in this case Template:tracking/headword/pagename spelling mismatch. It's customary to put the name of the module or template first before the slash. —CodeCat 18:38, 18 November 2016 (UTC)

I've made the change to the module. Could you set up the tracking page (it's still a bit unclear to me. Why are there so few tracking pages when I see this module all over?)? —JohnC5 18:44, 18 November 2016 (UTC)

Please replace: for _, head in ipairs(data.heads) do —CodeCat 18:45, 18 November 2016 (UTC)

Fixed. *sigh* —JohnC5 18:48, 18 November 2016 (UTC)

Currently false friend is showing up as a transclusion of the tracking template. I think the cause is the embedded wikilinks, they need to be stripped first. So try this:

local m_links = require("Module:links")

for _, head in ipairs(data.heads) do
	if mw.title.getCurrentTitle().prefixedText ~= m_links.getLinkPage(m_links.remove_links(head), data.lang) then
		require("Module:debug").track("headword/pagename spelling mismatch")
		require("Module:debug").track("headword/pagename spelling mismatch/" .. data.lang:getCode())
		break
	end
end

This way the links are stripped, and there's also a second tracking template to indicate the language (which eases searching). —CodeCat 18:52, 18 November 2016 (UTC)

It would be so much easier if I could do this myself. So much for "temporary" desysopping then. —CodeCat 18:55, 18 November 2016 (UTC)

Yes, your desysopping is very frustrating to me, in part because I am lazy and like getting you to do things for me. I still don't understand this debug module. The page Template:tracking/headword/pagename spelling mismatch still does not exist ,and the only pages I see under Template:tracking are:

• Kordofanian
• compound
• compound/prefix
• compound/suffix
• head tracking
• head tracking/singular category
• module override
• pt-adj no posit1

—JohnC5 19:04, 18 November 2016 (UTC)

Unfortunately, I've been having exactly the same problem not finding the category with false friend in it. You are both much faster than I am; I'm watching this but I have not able to say anything before it has already been said. Isomorphyc (talk) 19:12, 18 November 2016 (UTC)

Oh, it is here: [4] Isomorphyc (talk) 19:16, 18 November 2016 (UTC)

Interesting... Is this the intended behavior? If so, that is a bit silly. I'd prefer a category. —JohnC5 19:22, 18 November 2016 (UTC)

Feel free to make it a category. Not worth all the effort IMO. —CodeCat 19:23, 18 November 2016 (UTC)

Okey dokey. Now that I understand the function of the template, I see why it's done this way, I suppose. —JohnC5 19:24, 18 November 2016 (UTC)

Well, it took me several hours to find a subset of the misspellings from the XML files because it takes a lot of template expansion to find the actual headword module invocations. So to the extent others were concerned about misspellings, this is quite an improvement. Isomorphyc (talk) 19:34, 18 November 2016 (UTC)

@CodeCat: If you don't mind me asking, why aren't you a sysop anymore? --Fsojic (talk) 17:44, 23 November 2016 (UTC)

The results

Now that the transclusions have filled up a bit, I'm already seeing lots of them that we probably don't really want listed. For example, Netherlands is included because the proper noun specifies "the Netherlands" as the headword, and there's many more like that. On mi, the Zulu term adds a hyphen to the headword, but this is legitimate because many Zulu lemmas are just the bare stems (they're not affixes). —CodeCat 19:34, 18 November 2016 (UTC)

I'm wasting a lot of both of your time. I'm so sorry about this. Would either of you mind nominating me for the template editing whitelist? I'll relinquish it when I am done with this module. Isomorphyc (talk) 19:41, 18 November 2016 (UTC)

@Isomorphyc: Done, I think. — I.S.M.E.T.A. 12:46, 23 November 2016 (UTC)

Etymology of dubius

Is *duhibius or any similar words attested? --kc_kennylau (talk) 15:12, 10 November 2016 (UTC)

@Kc_kennylau: Not that I can find, but there are a lot of fragments and inscriptions which are hard to search. De Vaan does not mention this etymology, and it sounds quite whimsical to me. He mainly prefers a variant on *dwóh₁ + *bʰuH-, reconstructing the laryngeals somewhat differently than we do, and also offering several alternatives.

Edit: I have actually been having a slightly similar problem. I wonder if you would be able to help me at a glance? This etymology [5] suggests there is a variant on broccus/brocchus in Plautus, although this word is not in L&S, but only in Du Cange. It seems moderately unlikely, but not impossible, that Plautus would use a rare word of Gaulish origin. But I also don't know how to make sense of the citation: (Plaute, Sitell. frg. 2 dans TLL s.v., 2202, 63). It seems to imply Sitell. is a fragment, but I don't recognise it as one, and I cannot otherwise find any attestation for the word. Isomorphyc (talk) 16:32, 10 November 2016 (UTC)

I would interpret that as being a fragment of Sitellitergus as referenced by TLL (w:Thesaurus Linguae Latinae?) in the entry for the word. Chuck Entz (talk) 20:34, 10 November 2016 (UTC)

@Chuck Entz: Thank you so much! I was apparently unaware of Sitellitergus, but I was able to find my citation in Loeb's fragments volume with the title. Isomorphyc (talk) 21:06, 10 November 2016 (UTC)

@Isomorphyc: TLFi's claim seem to reflect the view preferred by Walde. The Celtic etymon, if correct, is well attested. —JohnC5 22:57, 10 November 2016 (UTC)

Thank you for checking in Walde and Matasović; it would be fascinating to see a list of Celtic origin words in Plautus. They must be from cis-Alpine Gaul if they are not remnants of one of the wars. Isomorphyc (talk) 14:47, 11 November 2016 (UTC)

As to the original subject: given the headword for the entry, I find the somewhat autological nature of the whole thing a little amusing. It reminds me of the old practical joke of telling some naive person "did you know that gullible isn't in the dictionary?". Chuck Entz (talk) 00:25, 11 November 2016 (UTC)

I thought that was pretty droll also. Isomorphyc (talk) 14:47, 11 November 2016 (UTC)

Pronunciation of "adj-" in Classical Latin

Do you think they would be pronounced as /ad͡ʒ/ or /adʲ/ or /ajj/ instead of /adj/ in Classical Latin? For example:

• Latin adiūdicāre > Old French /a.d͡ʒyˈd͡ʒjer/
• Latin adiūdicāre > Italian /ad.d͡ʒu.diˈkaː.re/
• Latin adiunctus > Aromanian /aˈd͡ʒum.tu/?
• Latin adiungō > Aromanian /aˈd͡ʒuŋ.ɡu/?
• Vulgar Latin *adiuxtāre > Old French /a.d͡ʒowsˈter/
• See Latin adiūtō, adiūtor, adiūtōrium.

--kc_kennylau (talk) 14:24, 12 November 2016 (UTC)

@Kc kennylau: This is considered a Vulgar Latin change (see w:Romance languages#Palatalization). —JohnC5 18:04, 12 November 2016 (UTC)

@JohnC5: Is there any evidence that it is not a Classical change? --kc_kennylau (talk) 00:33, 13 November 2016 (UTC)

@Kc kennylau: None of which I am directly aware except that ⟨dj⟩ served to create heavy syllables in poetry, which would imply that the underlying segments remained heterosyllabic. Also, it seems typologically strange to only have /d͡ʒ/ or /dʲ/ and no other affricates or palatalized consonants in the phonetic inventory. It makes more sense for this process to be part of the later wave of palatalization which occurred in Vulgar Latin. Again, I'm going to ping @CodeCat for commentary. —JohnC5 04:00, 13 November 2016 (UTC)

Descendants of Proto-Germanic

I changed the order of the entries in accordance with the list on this article: https://en.wiktionary.org/wiki/Wiktionary:About_Proto-Germanic --Theudariks (talk) 23:19, 13 November 2016 (UTC)

@Theudariks: Ah, I see. Sorry for the confusion. I've started a conversation about this at Wiktionary talk:About_Proto-Germanic#Germanic descendant order. —JohnC5 01:15, 14 November 2016 (UTC)

Skipped eddits

You skipped several edits I made. As a rule, I prefer discussions on the relevant pages, not on your page. As they were supported by relevant references, they are apt to discussion there, not to mere skipping, which I consider destructive behaviour.Otto S. Knottnerus (talk) 16:50, 14 November 2016 (UTC)

Questions

Hi there, I'm pretty new to Wiktionary and I'm still learning a lot of MoS and template stuff. I'd like to keep on learning and ask a few questions based on the edits you made to flagritriba: based on your example, should I leave the entry without inline citations? And what are those templates in the reference section? The Cassell's I understand, but I don't know what L&S and Gaffiot are. Thanks, Icebob99 (talk) 03:18, 17 November 2016 (UTC)

@Icebob99: Howdy, and welcome! Unless there's a particularly interesting piece of information that needs citation, we tend to put all the references at the bottom of the language section. I'm a little confused by what you are asking about L&S and Gaffiot. They are both Latin dictionaries (Lewis & Short and Gaffiot's dictionaries) similar to Cassell's. You can see a list of all the reference templates we have for Latin here. —JohnC5 03:52, 17 November 2016 (UTC)

Thanks! That makes sense and it seems efficient too. So you found the word in those dictionaries and added the templates? Should I add a template for every dictionary in which I find the word in question? Icebob99 (talk) 14:41, 17 November 2016 (UTC)

@Icebob99: I did indeed find it in those and add them. To your question of adding all the dictionaries possible, this is an interesting debate. For Latin and Ancient Greek, we tend to have a policy that we add any out of copyright (reputable) dictionaries we can find. L&S and Gaffiot are both great resources because they are out of copyright, fairly easy to use, annotated with good citational information, and available online. We don't add Cassell's as often because it does not have the lemmatized interface that the other two do, but feel free to add it! —JohnC5 14:57, 17 November 2016 (UTC)

Well this pretty clearly displays my novice status, because I don't know what a lemma is. I looked up the definition and it apparently is a proposition in a mathematical proof? Icebob99 (talk) 18:49, 17 November 2016 (UTC)

See definition #2 at lemma. --WikiTiki89 18:54, 17 November 2016 (UTC)

And to be fair, I still don't know that MoS stands for in your initial post. —JohnC5 19:41, 17 November 2016 (UTC)

I believe it's w:Wikipedia:Manual of Style. Chuck Entz (talk) 03:40, 18 November 2016 (UTC)

I suspected it was something of that nature. Good to know. —JohnC5 05:35, 18 November 2016 (UTC)

Āl

Dear JohnC5, I have to state that the past etymology here and that of anneal are amongst the very few incorrect etymologies left. I had spoken to someone with a full degree in ancient languages, who rejected the idea of a fabricated "ailijaną" as Proto-Germanic root, deriving from the P.I.E. root. That is the root of Greek "aitho", our "oast" and numerous other analogies; but the excrescent "l" needs explanation. Certainly the Nordic forms derive from the same root; but that they are Germanic is mere conjecture, since a few lexemes have been borrowed into Germanic at an earlier date from all these dialects, and more so in the Old Norse than into English! The Old English ǣled answers completely to Breton oalad (fireplace) and so do the Norse forms - only more contracted. I presented the etymology as from the modern Cornish dictionary for cognates. If lexemes akin to these without the additional 'd' were present, then the Germanic root would be entirely applicable; but as one stated to ChuckEntz, we want the truth in the end. You are always welcome to comment constructively on etymologies which I had presented, and especially the rules on my User Page - it is not my page, of course, any more than it is yours - it belongs to Wiktionary! I do not even believe that this lexeme is even Celtic, just because it is found in Celtic. CodeCat sorted me on that straight away - she put "Don't assume". That is part of my charter! If you find a water-tight explanation for reverting any of my presentation, that is all I need! We want the truth, or at least the nearest to it that is citable! Regarding an Estonian connection that I wondered about; it had absolutely nothing to do with it at all! The Ancient Greek forms are only applicable if they are of Phœnician origin and then, that Argentine Spanish alúa (glow worm) is of ancient Punic origin. Kind Regards, Andrew H. Gray 16:54, 9 December 2016 (UTC)Andrew

(Since Chuck and Code (who is a woman) were part of previous discussions, I shall reping them.) @Werdna Yrneh Yarg: Given the wide variety of Germanic reflexes of *h₂eydʰ- (*aidijaną, *aidaz, *aisǭ, *ailidaz, *ailijaną, *ailą), the fact that all the Celtic descendants listed are only Brythonic (Breton oelad, Cornish oeles, Welsh aelwyd), the fact that Kroonen provides a perfectly plausible explanation (PIE *h₂éydʰ-lom > PG *ailą > *ailjaną > *ailidaz > OE æled > PB), and that Breton oelad and Cornish oeles seem to show one Brythonic vowel, *oɨ, while Welsh aelwyd shows another, *aɨ, I feel that there's more evidence for Germanic inheritance with Brythonic borrowings and not the other way around. —JohnC5 00:22, 10 December 2016 (UTC)

Dear JohnC5 Thank you for the corrections. Have just seen one or two etymologies of Kroonen (who is certainly extremely learned and clever), but at least two I came across link logically with others that have been cited by him; and I am seriously concerned for their accuracy that is less so than that of Meriam Webster'sǃ He does cite "ailida" (asterisked) as origin of ǣled, which is, of course, entirely logical - if such a lexeme existed. Sorry, the "oaled" form may be something else that Breton - Dr. Ken George just presented C.B., and I may have wrongly assumed the Breton lexeme; so my due apologies for that. He also states that the forms in the Brythonic dialects were borrowed from early Anglo-Saxon - this assumption (which is very unlikely) just confirms a Germanic biasǃ Regarding your statement as to only one common vowel, (there were no common vowels earlier anyway - compare Old Cornish "oilet" (fireplace, hearth), that itself shows the futility of such an assumption) you know better than I that the older the language the more vowel and sound differences are likely to evidence themselves. I shall leave the entry edit as it is for now, since the sub-root form includes the excrescent "l" and it is sourceableǃ Sorry for the reversion; however, I cannot agree with any of those further origins of these related lexemes. I have to state here and now that if I had made that etymology of ̪"*āl" at the time when it was made I should have struck (stricken) out all the forms beyond the Old English ones; and stated its origin to be "uncertain". The classic assumption still continues that every Old English lexeme must be older than any infiltrated Celtic forms; and that if lexemes are found in Brythonic or Goidhelic dialects, that they must be Brythonic or Goidhelic. While it is true for most of them, this is certainly not the case with all; and there are quite a number of even Old Cornish words that are borrowed from Anglo-Saxon.̴̴̴̣̣̣ Kind Regards. Andrew H. Gray 13:07, 12 December 2016 (UTC)

I need to add to my improper reply. To begin with, it was just as well that you did revert my edit because two errors do not make cancel one another - as you fully know, Proto-Germanic is not derived from Proto-Celtic, although words may be borrowed from it; so to leave it as I had was upon reflection very confusing, to say the least. The other point I need to state is that it would have been practically impossible for Old English to have borrowed lexemes from Breton, nor the other way round, due to the fact that their native speakers moved over from Cornwall before Anglo-Saxon could get any hold on the Cornish language. What is not ascertained is whether the term 'fireplace' was used in Celtic beforehand. Only if not, would the Welsh and Cornish have borrowed from Old English, but not Breton! Nor do I believe that there is sufficient evidence that they derive from Proto-Celtic. It was just that this fabricated latter root was suitably sourced, but I should not have presented it if there is no evidence to confirm it! The Nordic forms are too fossilised to confirm a Germanic root, nor are there any present analogous forms in Dutch or German. However, it is also untenable that at least three forms should be extant or retained in Old English without extant cognate forms in ancient Spanish (apart from one - alúa) and in another Punic dialect. For the Ancient Greek forms, ἐιλη and ἀλεα to be relevant, they would have had to be of Phœnician origin, as I tried to make clear. My guess is that its real origin is cognate with that in Lappish, but no Estonian cognates are found; and those in Basque are prefixed by "kisk-", so "kisk" needs to mean something for that to be in any way relevant, otherwise such remoteness is merely sidetracking. Andrew H. Gray 11:54, 22 December 2016 (UTC)Andrew talk

@Werdna Yrneh Yarg: I'm sorry, what do "ancient Spanish" and Punic have to do with this question? —JohnC5 17:06, 22 December 2016 (UTC)

@JohnC5: Your question might be asked by practically any etymologist, because any assimulated words before the Celtic invasion are so scarce, and so little is known - you may know more - about the Iberian language akin to Punic that was spoken widely in Spain before the Roman invasion. However, one example of such a form that is found in Cornish is "ot!" = 'behold!', akin to Spanish "otear" (to pry). It has taken years of interest and research to sort out some of the pre-Celtic connections, during the 45 years of my interest in etymologies and dictionary comparisons. You clearly did not get the point. I trust that I did not imply that Brythonic derived from Iberian Punic, because it is not; just that a tongue related to Punic would have been spoken in Great Britain during the Bronze Age. Also Basque is largely akin to Lybian, except for a short vovabulary related to Tartar. However, there is no chance of any word being borrowed by the old Cornish who emigrated to Breton, since a late borrowing would in any case have been impracticable. We still just assume that these Old English related forms of āl are Germanic - that must be established first, I believe. It seems as if general teaching on etymology disregards the possibility - and indeed fact - of a few lexemes being carried forward from more ancient dialects! If this is not understood, it seems pointless to continue this discussion. However, I absolutely agree with your first two sentences below; and the period of Old English. Kind Regards. Andrew H. Gray 18:02, 22 December 2016 (UTC)Andrew talk

I will point out that the absence of a cognate is not proof to the contrary. Also, Brythonic is not immediately descended from Celtiberian and unrelated to Spanish. The discussion here is whether Old English (5th cen. to late 11th cen.) could have lent a word to Proto-Brythonic (1st cen to 6th cen.). It would have been a late borrowing, which could account for the differences in the Brythonic reflexes, but it is still very possible. The goings on in Iberia are unrelated. —JohnC5 18:41, 22 December 2016 (UTC)

Ok, there is no support to the claim that either “a tongue related to Punic would have been spoken in Great Britain during the Bronze Age” or “ Also Basque is largely akin to Lybian, except for a short vovabulary related to Tartar”. Modern linguistics does not support either claim in any way. There is a fair amount of comparative evidence that Old English āl is of Germanic descent (as stated in this discussion and on the relevant content pages). As to your comment about linguistics disregarding the possibility “of a few lexemes being carried forward from more ancient dialects”, this is literally the majority of what historical linguistics does, The main issue here seems to be that your notion of the relationship of different language families is very confused and not based in modern scholarship. I apologize if my statements sound rude, but some of the theories you espouse are just so outlandish! —JohnC5 20:26, 22 December 2016 (UTC)

@JohnC5: Thank you for your responses and am sorry for the way my last paragraphs may have seemed high-handed. That I certainly do not mean since I have respect for you and the other administrators. However, I now see where one of your theories originates. There are two major points here:

1/ There must be no confusion in the relationship/non-relationship of the different language families, as you importantly stress. I am, of course aware that Germanic, Brythonic and Goidelic are separate branches of the Indo-European family; whereas Iberian is reputed to be Semitic, like Punic and Phœnician. Also, Lappish, Estonian, et cetera, belong to a separate language head - Finn-Ugrian, and the grammar of Basque is related to a dialect in the Caucasus Range, standing as it does, on its own. The understanding of the movement of the ancient peoples in Europe (from the Readers' Digest atlas) is vital to present clues as to history of languages in Britain and in Europe. Regarding the Javanese language in Spain; many of their characters found were apparently similar to that of Greek, hence the conjecture. The map clearly depicts the Megalithic New Stone Age tribe encircling Spain and arriving in Biscay, as well as in parts of West Ireland and Britain. Any further knowledge on this subject is welcomed.

2/ "Historical" "linguistics disregarding the possibility of a few lexemes being carried forward from more ancient dialects". If you believe that - and I am not suggesting that you do believe such a statement - I think you have been very polite, and I sympathise completely with your concerns as to my etymological edits. I may as well state now that highly intelligent and more intellectual professionals than I condemn such a theory, since it is not based on fact. There will be no perfecting of etymologies of the oldest European lexemes until this is realised and accepted. They would best keep to their linguistic talents, but out of the etymological field until they rescind what is glaringly wrong, because that statement that you helpfully brought in highlights the whole problem, which is still wrecking accuracy, and should save your valuable time and that of the other Administrators, by bringing it to light. This explains why some of the etymologies go wildly off track, with one or two modern etymological dictionaries and why, people in general trust the multi-volume editions of the Oxford Dictionary that keeps to what they know and that can be trusted. Thank you again for bring that issue to light. You are fully welcome to shift all these paragraphs on to my Talk Page, so as not to clog up yours; but please not on the User Page. Kind Regards and hope that you enjoy the holiday break. Andrew H. Gray 14:04, 23 December 2016 (UTC)Andrew talk

re: your first point: There are all kinds of claims about Iberian, and most of them (especially those of a century or more ago) are utter nonsense. Also, there's a region called Iberia in the Caucasus that has nothing in common with the Iberian peninsula except the name- be careful not to confuse the two. As for Basque, there have been attempts to link it with just about every language group out there- again, mostly utter nonsense. The same holds true for Sumerian and a few other mysterious and iconic isolates. There are languages with similar grammar to Basque all over the planet, so that aspect isn't very helpful in determining relationship. Linguists can't even demonstrate a relationship between the language families in the Caucasian region, let alone a relationship with Basque. As for "Javanese", that's the name of a language spoken in Java, so you need to clarify what you mean by that. It reminds me of a name for the Greeks in some ancient languages of western Asia (probably related to Ἴων (Íōn)), so there may be some confusion here.

re: your second point: there are no doubt survivals from ancient languages unknown to us, but there are also words that people have made up over the millenia that coincidentally resemble others by random chance- and there's simply no way to tell them apart. It's fun to speculate, and for some people it's a way to pretend there's support for their pet theories- but anything based on such speculations is bound to be nonsense. As the old thought-experiment says, an infinite number of monkeys at an infinite number of keyboards would eventually type out all of Shakespeare's works- but also an infinite amount of random gibberish. If you pay too much attention to those monkeys, you may get something like the phrase from an old Bob Newhart routine on the subject: "To be, or not to be, that is the gezorninplatz". Chuck Entz (talk) 22:09, 23 December 2016 (UTC)

Chuck Entz Thank you very much for the main points which you put. The number of made up words is likely to be of very small and of almost insignificant proportion. However, great care is required in linking lexemes that seem to be pre-historic or of pre-Celtic origin, since the element of uncertainty is exponentially increased. I strongly believe, rightly or wrongly that bird is Germanic, although no reliable cognates are known and the older O.E.D. seems to discount it as a derivative of brēdan (to breed). Also I thought that Old English low (flame) is Germanic, because of its counterpart cognate in German and Old Norse, but although the English form is certainly derived from those equivalent forms, in those Germanic dialects their two lexemes are not necessarily Germanic. I fully accept that indeterminate conjectures from extinct languages in any area must be avoided on Entry pages. Regarding ancient Iberian, its script characters have hardly anything (and that only in a few), in common with those of ancient Greek, so there is not enough evidence for a Phœnician connection there. The Greek form, Ἴων (Íōn), is certainly from JAVAN; and the splinter dialects in Caucasus are quite independent of one another. However, the grammar in one of them is, whether significantly or coincidentally, akin to that of Basque. More recent Pictish is Brittonic; the oldest forms are totally unrelated and belong to a non-Indo-European family. You are fully welcome to move these sections to my Talk Page, as I do not want to clog up JohnC5's, nor yours. Hopefully this has led to an agreed conclusion now, since I do NOT want a "Nemzag" style argument page; that is, what the administrators put up with him! Kind Regards. Andrew H. Gray 13:44, 29 December 2016 (UTC)Andrew