Wiktionary talk:Votes/2017-05/Modern Latin as a WDL

Please give feedback on the wording (or anything else)

Pinging a few interested parties (and please feel free to ping anyone else): @-sche, Mx. Granger, Chuck Entz, I'm so meta even this acronym. —Μετάknowledge^{discuss/deeds} 02:39, 3 May 2017 (UTC)

Thanks for creating this vote—it's high time we edit our policies to codify existing practice. Also pinging @Dan Polansky, who may be interested. —Granger (talk · contribs) 02:59, 3 May 2017 (UTC)

Such as?

"Terms in extinct languages only attested after the language went extinct are also treated as well documented languages." Can someone give an example of an extinct language only attested after the language went extinct? Off the top of my head, the only ones I can think of are Sanskrit and Avestan (which weren't written down until centuries after they stopped being spoken languages), but there may also be indigenous languages that were only recorded shortly before they went extinct, and then the results of those recordings were only published after the last native speaker died. I certainly wouldn't want to consider such languages "well documented"! If the intention is to make Contemporary Latin a WDL, this wording isn't going to do it. —Aɴɢʀ (talk) 07:57, 3 May 2017 (UTC)

This sentence has been replaced (see below), but just to respond to a couple of the concerns you mentioned here, I think it would be absurd to date a recording by when it was published and not by when it was recorded. Similarly, the Avesta was composed long before the oldest extant written copies, and quotations of it should use a rough estimation of when it was first composed as the date. —Μετάknowledge^{discuss/deeds} 17:57, 3 May 2017 (UTC)

Sentence doesn't make sense

"Terms in extinct languages only attested after the language went extinct are also treated as well documented languages." So the term is to be treated as a well-documented language? That doesn't make sense.

How about "When a term in an extinct language is only attested after the language went extinct, that language is treated as a well-documented language with respect to that term." It's a bit harder to parse at first glance, but I feel that the policies of a dictionary ought to be written in impeccable English. What do you think? This, that and the other (talk) 11:51, 3 May 2017 (UTC)

The sentence definitely needs work. As seen in the thread above, I interpreted "attested" as referring back to "languages", not to "terms". —Aɴɢʀ (talk) 12:17, 3 May 2017 (UTC)

Thanks for the help! I've replaced that sentence with ISMETA's suggestion. —Μετάknowledge^{discuss/deeds} 17:54, 3 May 2017 (UTC)

Unclear wording?

In my opinion the wording and intention are too vague.

• What languages are "extinct languages"? Middle English, Old and Middle High German, Gothic, Ancient Greek are extinct. But what's with Latin (and maybe Sanskrit in India)? One could argue that Latin is extinct, and one could argue that it's not extinct as it's for example used by the church but is an endangered language.
• Wouldn't the new wording mean that non-authentic words can now be added, if attested by three citations? For example, if enough people translate and publish the paternoster in Proto-Germanic, couldn't Proto-Germanic entries be added in the normal entry namespace? Or if people would write stuff in Gothic, couldn't now new words for Gothic (like the deleted 𐌳𐌰𐌽𐌰𐌼𐌰𐍂𐌺𐍃) be added? I don't have a problem with Neo-Gothic or Neo-Gothic entries, but it shouldn't be allowed to add them as normal Gothic words even if attested by three citations.
• In Wiktionary talk:About Latin#Wiktionary:Votes/2017-05/Clarifying LDL status of extinct languages it was said that the vote "impinges on Modern Latin" and "does not change existing practice". It would only impinge on Modern Latin, if one says that Latin or Modern Latin is an extinct language. Depending on the interpretation, it could change existing practice regarding Latin.
  (a) If one says that Latin including Modern Latin is an extinct language, then a single usage is sufficient to attest a word, and nothing would change for Latin.
  (b) If one says that only older Latin excluding Modern Latin is an extinct language, then existing practice could change. Does one have to attest doubtful inflected forms and genders with three cites too for a WDL? In English this would rarely matter as English is a simple language. In German it is different, and there can be doubts about the gender and the inflection, especially in case of loan words (e.g. Bookmark). Personally, I think that three cites should be required for German to attest such information. While for German it makes sense to attest inflection and gender with three cites, it's a bit different for Latin, including Modern Latin. And inflected forms and genders of Latin words, including Modern Latin, have sometimes been attested with a single cite. For example, the non-classical and un-Greek feminine anōmala of anōmalos was attested by a single cite (the second one in Citations:anomalos, where "?" might be a catholic in the 20th century). And likely there are translations and entries with New Latin words or meanings which got added based on a single usage found somewhere.
• If the intention is to make Modern Latin or more specific contemporary Latin a WDL (compare Angr's "If the intention is to make Contemporary Latin a WDL" above), then it's, in my opinion, not a good idea. While Modern Latin once could have been a WDL, Modern Latin and especially contemporary Latin is not a WDL anymore. And Modern Latin, including contemporary Latin, is different from Neo-Gothic - which isn't a WDL either but could be regarded as a constructed language excluded by WT:CFI#Constructed languages - because of it's history and use for example by the church.

-Slœtel (talk) 20:45, 3 May 2017 (UTC)

'Extinct' may not be the best term. Perhaps 'dead' would be preferable. The idea about Modern Latin being a WDL is very simple, I (and probably you as well) want hamaxostichus to be acceptable to keep, but if someone uses iPhonum in a book of Neo-Latin jokes, that shouldn't be sufficient. —Μετάknowledge^{discuss/deeds} 20:53, 3 May 2017 (UTC)

I think it's a mistake to try to make a general rule for this, and I am currently minded to oppose the proposal as it's now worded. It would be preferable to me to simply add Latin (from Classical all the way to Contemporary) to the list of WDLs rather than trying to fiddle with the wording in such a way as to make Latin (but not Avestan or a recently extinct indigenous language of the Pacific Northwest) fit our definition of a WDL. —Aɴɢʀ (talk) 21:03, 3 May 2017 (UTC)

So you think Latin hapax logomena should be deleted? That's a very different view from that of most of the community. —Μετάknowledge^{discuss/deeds} 21:04, 3 May 2017 (UTC)

@Metaknowledge: I don't see any fair way of including Classical hapaxes while excluding iPhonum if it's published somewhere that the Latin editors have "deemed appropriate as [one of] the only sources for entries based on a single mention". Personally I would rather include both types of hapax than neither. What I don't like is tweaking the wording to make Classical Latin an LDL but all of post-Classical Latin a WDL, all while never actually mentioning Latin by name. —Aɴɢʀ (talk) 06:09, 4 May 2017 (UTC)

For me, both 'extinct' and 'dead' would lead to the same question. In wiktionary, extinct language and dead language are the same, while in w:Extinct language there is a distinction albeit w:Language death implies there is none. I don't know if wiktionary's definition or wikipedia's distinction are universal, but for me a language without native speakers isn't necessarily a dead or extinct language (Latin, constructed languages, programming 'languages', maybe sign languages). Maybe it's a matter of translation, but I've sometimes read that Medieval Latin was a living or not a dead language although it wasn't a native language but a language of the educated people. So with the definition of dead language as "language without native speakers" this would be nonsense.
Well, if someone would do something similar in another LDL, it would be acceptable too. Or as someone else put it (Talk:birotula): "Is modern Latin an LDL? I'd say it is. [...] it's hardly used at all in modern works. I wouldn't be surprised if Irish is used more than Latin is, and Irish is an LDL." The only way to prevent the addition of joke words for LDLs including Latin, would be to require three citatons for every language, maybe with the exception of really dead or extint languages or old sources to allow old hapax legomena. I don't think that is something the majority of users or editors would want. Furthermore, besides "ungood" Neo-Latin words or word forms which one might want to have excluded, there are also "good" Neo-Latin words or word forms with very little attestation which one might want to have included. There are so many words of which I would be glad to find a Latin translation and may it be a single usage somewhere.
Before this vote should start, it should be clarified what would have to be deleted. Considering genders and word forms, the claim that it "does not change existing practice" should be wrong.

• At the moment there are only two citations for the non-classical and un-Greek feminine anōmala, so it would have to go.
• At the moment there is only one citation for the accusative Dzibutum (of *Dzibutum or *Dzibutus), so it would have to go too. Maybe there are other attested place or state names only attested by single citation.
• ununhexium, ununoctium, homophylophilia and venae jejunales in jejunales didn't seem to have any Latin Google Books result and maybe would have to go after a RFV. But well, that's maybe not matter of having one or three citations.
  Esquimensis maybe would have to be changed after a RFV as Google Books seems to have only three results with the plural (Esquimenses, -ium, -ibus) which maybe don't reveal any gender. But well, like above.
• Maybe organa was added based on the reference Neulateinische Wortliste by Johann Ramminger and his source. But Ramminger only mentions a single source. So this term maybe was added without being attested by three durable archived citations, and maybe it would fail a RFV (maybe also because it's not easy to search for that organa).
• Maybe afroasiaticus. As far as I saw there were 0 Latin results for afroasiaticus, -um, -am, -orum, -arum, -os, -as, -is, likely 0 for -a and -i, 1 for -ae ("Species afroasiaticae." in an English and Latin text). There could be a second citation for -ae, but still a third would be missing. There might also be an English book with the title "Pax Afroasiatica? Revisiting Bandung Amid a Changing World Order", but that wouldn't be third citation for the Latin term as it's English and not Latin.
• Harrius by the given etymology. This could also fail because of WT:FICTION.
• One or both genders of Sprea (Spreha) maybe would have to go to after a RFV (some citations which could reveal the gender after searching for Spreha, Sprehae, Spreham, Spreae and partly Spream and in certain combinations Sprea: "[...] Spreae nostrae, Nympharum Numina, alumnae [...]", "Cur novi Sprehae [...]", "Oppidum Lusatae [from *Lusatus = Lusatian?] Spreae", "[...] ad plagam Spreae Lusaticae [...]", "[...] inter Albim & Spream orientem [masculine or feminine, not neuter] [...]").
• Depending on the interpretation of the new text and maybe on the definition of extinct or dead language, Medieval Latin terms would have to be checked too. If one argues that Latin died before or during the Middle Ages, Medieval Latin or parts of it would have to be treated as a WDL, and then the entries in Category:Medieval Latin would have to be checked.

-Slœtel (talk) 06:07, 4 May 2017 (UTC)

I mean 'dead language' to denote after a language loses its community of native speakers. I don't see what genders have to do with this, as nobody else seems to see them as being appropriate for an RFV, nor do I see why you brought up many entries that would fail regardless, like ununhexium. But I am a bit bothered by how I am unable to tell from your walls of text what exactly you would prefer. —Μετάknowledge^{discuss/deeds} 07:19, 4 May 2017 (UTC)

That would make Medieval Latin and Modern Latin dead languages. So by making it a WDL one would have to find three citations for Medieval and Modern Latin.
WT:CFI might not explicitally state it, but inflection and gender has to be attested too. For English it's easy as English is a simple language and a WDL, and for German it's quite easy as German has an article and is a WDL too. For Latin it's sometimes not so easy, and sometimes one could assume different inflections and genders.
I bothered to check for some terms in Category:New Latin how good they are attestable. By making Latin a WDL, for example afroasiaticus could become unattesable for wiktionary. ununhexium etc. did I mention for reasons like:

• If someone wants to add something to that list (and not to check that list), they don't have to look for the terms a second time.
• If someone would bother, he could add a RFV for the terms.
  And obviously attesting them for a LDL would be easier than to attest them for a WDL.

I wanted to know what the rephrased text would mean and how it would affect wiktionary, and I doubt that the claim "it does not change existing practice" is true. If you were asking for my personal opinion, then it is that I don't think that Latin is a LDL.

-Slœtel (talk) 18:23, 4 May 2017 (UTC)

You have still not told me what you would prefer. —Μετάknowledge^{discuss/deeds} 19:43, 4 May 2017 (UTC)

Now I am a bit confused. Are you asking for the classification of Latin as LDL or WDL which I would prefer, or for my opinion regarding iPhonum, or something else?

Short answer: IMO Latin is a LDL, thus iPhonum should be added if attestable (by at least one durable archived Latin usage), although I have some aversion against it too.
(My original longer answer can be seen hereafter in the source code as a comment.)

I now can think of two possible compromises:
(a) One could require to add the template {{LDL}} into the entry if there are just one or two citations. Actually, as far as I know, this is already the rule, but maybe not common practice and not widely known.
(b) Similary of having a "list of materials deemed appropriate as the only sources for entries based on a single mention", there could be an additional list or an additional requirement. For example, one could exclude a single modern print-on-demand (or book-on-demand) or usenet citation as a source. That is, if one finds a word in a printed Nuntii Latini book, it would be sufficient. But if one finds something in the usenet, one would have find other sources. However, if one finds three citations in the usenet, it would pass just like for the WDL English (see e.g. Leftpondia, me too (the verb), agreeingness).
There is a third possibility, but in my opinion not a good one: One could add terms which are attested but unattested for wiktionary in an appendix. So if one says, that contemporary, modern, non-ancient or all of Latin is a WDL, iPhonum could be added into this appendix based on a single citation. But I can't think of any fitting name for that appendix, and this has the problem that appendices aren't easy to find and sometimes it seems that they aren't maintained.

-Slœtel (talk) 11:24, 5 May 2017 (UTC)

Alternative vote

I agree with Angr, we don't need a general rule if all we want is clarify how Latin is treated. I'm not inclined to support this vote, both on that principle and because an overly general rule seems likely to have unintended consequences. (What would it do to American languages that were suppressed but then revitalized by the same peoples who previously spoke them? Or to New York hobbyists' Old English neologisms?) I suggest a vote or even just a WT:T:ALA consensus-seeking poll with options along the lines of:

1. Hold older Latin to the same (LDL-like) requirements as other extinct languages, and modern Latin to the same (WDL) requirements as other conlags.

   This has been the prevailing practice, as seen in various RFVs, and is a compromise position that can be contrasted with how we don't include any words from e.g. Gothic that weren't in the historical corpus, even if they're in three modern books. AFAIK, we haven't had any edge cases that would require us to debate when the cutoff between older and modern Latin is, btw...

2. Hold all Latin, even modern Latin, to extinct-language (LDL-like) requirements (which require only 1 use or mention).

   This seems likely to have unintended consequences, including in distinguishing new coinages that were intended to be sincere (whether by a Vatican linguist or a bumbling hobbyist) from ones that are not necessarily intended to be Latin at all, in the vein of noli illegitimi carborundum.

3. Hold all Latin, even old Latin, to WDL requirements (which require 3 uses).

   I doubt this option would pass.

4. Oppose.

   For whatever reason, e.g. preferring none of the above, opposing having a vote, etc. Users have previously opined that every vote needs an oppose section.

If no option that proposes a rule gets a majority, then the status quo prevails / the situation continues to be unregulated. - -sche (discuss) 07:24, 4 May 2017 (UTC)

That would be fine if it were just Latin. Unfortunately, it's not — RFV has also turned up modern Old English (used by authors like J. R. R. Tolkien) and Ancient Greek (famously used by Jan Křesadlo, but which, like Latin, has Harry Potter translations and other writings). That's why I think we need to rewrite this vote, rather than simply solve the case of Latin. —Μετάknowledge^{discuss/deeds} 07:30, 4 May 2017 (UTC)

As for the 1st point:

• Latin and Gothic are different because of their history and Latin's use for example by the church.
  As for Μετάknowledge's concers regarding Ancient Greek: Was or better is Ancient Greek still used by the Greek church? From what I've read it was more like Ancient Greek vanished in usage and became a really extinct or dead language like Gothic (that is, there were not just no native speakers, but there also was no contemporary usage). As for Modern Old English, Neo-Gothic, New Ancient Greek - or whatever one wants to call it -, one cann argue that they're excluded by WT:CFI#Constructed languages.
• I would like to know where the cutoff would be. I can think of adding Medieval and Modern Latin terms. But if there is a cutoff somewhere which would mean that the terms might not be covered by WT:CFI, I wouldn't bother to add them, as not to vaste my time and as not to (unknowingly) violate wiktionary's inclusion criteria.

As for the 2nd:

• In my opinion Latin is a LDL similar to other LDLs. See above: "Well, if someone would do something similar in another LDL, it would be acceptable too. Or as someone else put it (Talk:birotula): "Is modern Latin an LDL? I'd say it is. [...] it's hardly used at all in modern works. I wouldn't be surprised if Irish is used more than Latin is, and Irish is an LDL."".
• As for the phrase, I first guessed that there are already similar problems and that it's often quite easy to decide. But then I was thinking if I might misunderstand WT:CFI or if there might be different interpretations.
  WT:CFI allows to add terms for a LDL based on a mentioning. If I understand the above argument correctly, then it is that some people could argue for the inclusion of membrum puerile and English noli illegitimi carborundum as Latin terms although it's only attested in languages other than Latin. But the exceptions for LDLs usually do not allow such additions, neither for Latin nor for other LDLs. WT:CFI has this line in it: "the community of editors for that language should maintain a list of materials deemed appropriate as the only sources for entries based on a single mention". By this, Old Prussian words can be entered by the mentionings in the Elbing Vocabulary - but only if this source is deemed appropriate. There is no WT:About Old Prussian, so there one can't find a list of "materials deemed appropriate". I don't know if there is such a list anywhere else, but I'd guess there isn't. And if there is no list, then the Elbing Vocabulary can't be used to add Old Prussian words. As for Latin, only a few old works are deemed appropriate for adding terms based on a mentioning: WT:About Latin#Attestation. So membrum puerile and English noli illegitimi carborundum regardless of there commoness can't be added as Latin entries unless they are attested in Latin.
  If Latin is a LDL, iPhonum and noli illegitimi carborundum could be added based on a single usage conveying meaning in a durable archived Latin text. I can understand the aversion to this. But still I don't think that Latin is a WDL, and there might be "good" words which can only be found once as Latin is rare LDL nowadays. The better way to express the doubts or aversion regarding terms like iPhonum would be to remind people of using the template {{LDL}}. Actually, Latin hapax legomena found in Classical Latin would have to be marked by this template too, or is there any exception? Maybe the pleasingly solution is this: Strictly require {{LDL}} for Modern Latin, but remain by the, as far as I know, common practice of not using it for Old, Classical and Late Latin.

So maybe there should be two different votes:

• How to treat Modern Old English, Neo-Gothic, New Cornish etc. (excluding Latin)? Are they part of the old LDLs, or are they excluded constructed languages, or are they own languages, or are they anything else?
  For example, can the New Ancient Greek Harry Potter (ΑΡΕΙΟΣ ΠΟΤΗΡ καὶ ἡ τοῦ φιλοσόφου λίθος) be used to add Ancient Greek words, if the book is a durably archived source? If it's not a durable archived source, then there is Asterix which is durably archived (side of the publisher, see beneath "ΣΤΑ ΑΡΧΑΙΑ ΕΛΛΗΝΙΚΑ").
  Latin is excluded from this as it's different from Modern Old English, Neo-Gothic etc. While the latter really became extinct, Latin was used over centuries and still is used by a small minority.
• How should the template {{LDL}} be used? Or more specific, can entries of old really dead or extinct languages or language forms like Ancient Greek, Gothic, Old, Classical and Late Latin be added without the template even if the term is or might be barely attested?
  All the entries in Category:Gothic hapax legomena, Category:Middle French hapax legomena, Category:Old French hapax legomena are attested just once and they have a usage note in it, but none of them has the template {{LDL}} in it. In case of Category:Latin hapax legomena it's different and more complicated: Some terms might be hapax legomena in old Latin, but might be better attested in modern Latin (e.g. synaeresis). But anyway there should be Latin hapax legomena for all of Latin without any note or {{LDL}}. The entries in Category:Middle English hapax legomena - it's just one entry - don't have a usage note or the template in it.
  So if there is no such rule, there could be one like "A usage note can replace the template {{LDL}}". And one could add another rule like: "For really dead or extinct LDLs, it's not necessary to use {{LDL}}. In case of living LDLs however, it is required to use them." Regarding modern Latin and iPhonum there could be a line like this: "Modern Latin since X [maybe 1900?] is treated like a living LDL, thus requiring the usage of the template."

-Slœtel (talk) 18:23, 4 May 2017 (UTC)

(@Meta:) has any Old English / Gothic / Ancient Greek term come close to being passed at RFV with only modern citations? As far as I know, everyone (except whatever new user has driven by to add a particular neologism) recognizes that no modern citation in those languages is sufficient (because it isn't a use of Old English / Gothic / etc, but someone's conlang of sorts). This vote would seem to start allowing neologisms if they had 3 uses.

Latin is unlike those languages, in that we do pass terms in it even if they have only modern citations (probably because there've been people continuously using it since the days when it was a vernacular language, and one of the major sources of modern citations is an organization that has used it continuously). As far as I can recall, the only editors who argue terms should pass with just one citation are Dan and the IP/Slœtel (maybe Angr, too?), which is why I'm not entirely sure a vote is needed even for Latin, if we've been resolving things adequately at RFV on the existing wording and understanding of CFI (terms in the non-extinct version of Latin are different from "terms in extinct languages"); codification could be good, but then we might need to nail down a cutoff between old and modern Latin.

This general vote seems like it would allow neo-Old English, Gothic, etc (if they had three cites), exclude various native American languages that were e.g. extirpated for a time but then revitalized, and throw into doubt words in native American and African and Australian languages recalled by people after the languages ceased to have enough fluent speakers to be means of communication. I'm sorry, but I don't think I can support that. - -sche (discuss) 20:37, 4 May 2017 (UTC)

Regarding the comparison to Irish: as I understand it from the various discussions that lead to the LDL and then WDL votes, a language is "well-documented [online]" if (a substantial part of) the language is documented [in places that meet CFI and are findable]. A large part of modern / neologizing Latin is in e.g. Catholic Church works and other works that get published and publicized and made findable, which is different from not just ancient Latin or Greek, of which we know we have only a fragmentary record, but also Irish or Lavi or other languages that still don't have many [findable-by-us] works written in them despite still having speakers.

I'm sorry this turned into a wall of text. - -sche (discuss) 21:04, 4 May 2017 (UTC)

But still it's nowhere said that Neo-Gothic, New Ancient Greek etc. are treated as constructed languages. If WT:CFI would explicitaly say so and people would know it, they likely wouldn't bother to add those terms in the entry namespace.
The proposed text would allow terms in Neo-Gothic etc. and wouldn't necessarily decide the status of Latin.

There are or were other users too, like the one at Talk:birotula from 16:04, 17 November 2013 which I quoted before. And seeing entries like afroasiaticus, there could be more users.

Does contemporary Latin have many online-findable-by-us works? There are church texts and there is Nuntii Latini which can be found. The few translated works like Harrius Potter et Philosophi Lapsi usually aren't to be found online, at least as far as I know. So contemporary Latin doesn't seem to have many online-findable-by-us works. And moreover there aren't many contemporary Latin texts anyway. So even if we could find many works of these few avaible works, it would just be finding few works, making it a non-WDL. As for the uttered concerns, I now can think of two possibilities, using {{LDL}} or to exclude single usenet citations to attest something.
Latin from 1500 (ca. the beginning of printing and of New Latin) till 1850 (works before it should have lost their copyright status in many states and before that Latin was more living) might be well-documentated online. But even for that there are problem too, like besides finding words in works one has to find and verify the meaning. -Slœtel (talk) 11:24, 5 May 2017 (UTC)

• @-sche: Here's a concrete, non-Latin example: I don't want Ancient Greek μικροοδύσσεια (mikroodússeia), used only by Křesadlo, to be added to Wiktionary. Slœtel does. I agree with your viewpoint, but it's just too tenuous as something we bring up every time there's a relevant RFV and that Dan Polansky complains about every time. That's why I want some sort of policy to make it clear. So what if we restricted this vote to Latin, Ancient Greek, and Old English, the only languages where this problem seems to have cropped up? (Oh, and @Slœtel, I'm pretty sure that your belief about Modern Latin on the Internet is just plain wrong. Try looking around a little.) —Μετάknowledge^{discuss/deeds} 17:45, 5 May 2017 (UTC)

@Metaknowledg: As for your "Slœtel does", I've not said that. I always differed between Latin and Gothic, Old English, Ancient Greek etc. as they are different. And as for μικροοδύσσεια, one can argue that it's not permitted by WT:CFI even if one finds three durably archived citations for it. WT:CFI doesn't explicitaly forbid or allow those terms, and there are two ways one could argue: (a) Claim that μικροοδύσσεια is used in an Ancient Greek text (by the language, not by the time and others things). As Ancient Greek is a LDL, μικροοδύσσεια could then pass with one citation. (b) Claim that the text is not Ancient Greek but in the constructed language New Ancient Greek. μικροοδύσσεια then could not be added, even if there were three or a dozen citations for it. According to SIL μικροοδύσσεια wouldn't be part of grc as it is defined as "Ancient Greek (to 1453)". That is, μικροοδύσσεια would be part of another language like New Ancient Greek. So regardless of the opinion I might have and regardless WT:CFI's vagueness, it's more likely that μικροοδύσσεια can't be added to wiktionary at the moment, or only in an to-be-created appendix.
Again, Latin and Ancient Greek, Old English etc. are different and one should differ between it.
As for Latin on the internet, usages on web pages, including boards and wikis are irrelevant for wiktionary. usenet might be a source, but I've never used it and can say anything about it. And regarding usenet I suggested something above. As for Google Books, the common source to find citations, it might depend on the state (or settings, or maybe also internet connection or PC as for Google's instant search?). For example, the Latin Can be Fun source for hamaxostichus doesn't work for me and I don't see anything there not even a search form. If I search for "hamaxostichus" at Google Books myself, I do find some mentionings, but that doesn't qualify as sources. Besides that, the only usage Google Books usually displays to me is in a Latinitas (vol. 55, not 42-43 as in the entry). And there I can't see the whole sentence. As for the source Ursus nomine Paddington, Google doesn't always bring it up to me but it somehow depends on the browser or settings or search. If I search for hamaxostichus or "hamaxostichus", I usually don't see it. If I search for hamaxostichus ursus, I can see it. If I click on the result I can sometimes find or on the provided link in the entry, I see a search form. If I use it, Google gives me a picture with Latin text, but hamaxostichus doesn't appear on it. So if I search for hamaxostichus at Google Books, I can't attest it with three citations and at best only with one incomplete citation. I guess it mostly depends on the state and the copyright, although it might also somehow depend on the browser, the search settings (unlikely), or the PC or internet conection. My assumption is that one can see more at Google Books, if one is in the USA. I'm not from there and not in there. So both could be correct, my statement and yours. If other people would search for hamaxostichus at Google Books, they could help to find the reason why the results are different. If I'm correct, and it does indeed depend on the state, the LDL/WDL distinction might not be universal. That is, contemporary Latin could then be a LDL for an European (maybe one has to differ between Spanish, French, English, German etc.) but a WDL for someone from the USA. -Slœtel (talk) 20:53, 5 May 2017 (UTC)

It looks like no one is trying to add Ancient Greek μικροοδύσσεια (mikroodússeia), and I see no reason to think there wouldn't be strong agreement that it wasn't Ancient Greek if it were added. But this vote would allow it if a few people had used it on Usenet, the opposite of what we all seem to agree we want.
I do think it could be good to migrate away from speaking of "LDL" and "WDL" when discussing these languages, because e.g. Ancient Greek is subject to criteria that govern extinct languages rather than criteria that government LDLs, and it might be more sensible (and more agreeable to some users) to describe conlangs (which at the moment are listed as WDLs, it's true) and modern Latin as subject to "the same criteria as constructed languages" which happen to be the same as WDL criteria, rather than to describe them as WDLs, so as to sidestep debate over whether or not they are "well-documented".
- -sche (discuss) 19:51, 6 May 2017 (UTC)

@-sche: I would prefer to address this preemptively for other languages, but it seems that if I want to be practical about it, I'll have to accept just handling the Latin part. I do like the idea of explicitly linking Modern Latin to constructed languages. If you want, you can try rewriting the vote. —Μετάknowledge^{discuss/deeds} 20:05, 6 May 2017 (UTC)

I think the wording "as subject to the same criteria as WDLs" is better as it is shorter and more direct. And it would avoid the discussion too.

Cornish or more specific Neo-Cornish isn't listed as a constructed language at WT:CFI and not as a WT:WDL, and there are Neo-Cornish words like pellgowsel. Was pellgowsel ever used in a Neo-Cornish text or is it from a word list? Because of Cornish there could be terms similar to a New Ancient Greek μικροοδύσσεια.

One could address it preemptively, and it makes sense to address it preemptively, but still one should differ where necessary. -Slœtel (talk) 22:54, 9 May 2017 (UTC)

Adding contemporary Latin to WDL

I see there is some dislike above of the generality of the proposed wording.

What if we add "Contemporary Latin (1900-present)" to "Wiktionary:Criteria_for_inclusion/Well_documented_languages"?

If there are other languages that need a similar treatment, they can be added one at a time. If they become too many, we may revisit a general wording, right?

--Dan Polansky (talk) 19:05, 8 May 2017 (UTC)

Later: I am not sure which phases of Latin we are talking of. The item to be added would have to be changed to match the intended phases to be 3-attested. --Dan Polansky (talk) 19:22, 8 May 2017 (UTC)

Choosing a date is difficult; the real question is whether we want to include Mediaeval Latin as a WDL. Even if we don't, 1900 seems far too late as a cutoff for New Latin — maybe 1500 would be more reasonable. —Μετάknowledge^{discuss/deeds} 23:52, 8 May 2017 (UTC)

Thanks. So we would add "New Latin and Contemporary Latin (1500-present)", as the first step. Further extension could be made after that. I checked W:Template:Latin periods for names and dates. --Dan Polansky (talk) 17:14, 9 May 2017 (UTC)

@-sche, Angr, Mx. Granger, Slœtel, what are your thoughts on Dan's idea and wording? —Μετάknowledge^{discuss/deeds} 17:52, 9 May 2017 (UTC)

I would support it. —Granger (talk · contribs) 17:56, 9 May 2017 (UTC)

I would support it being the way to word a vote. I'm not convinced I want to separate Latin into LDL-Latin and WDL-Latin at all, but if there are people who do want to do that, then this wording is IMO far preferable to the wording of the current vote. —Aɴɢʀ (talk) 18:11, 9 May 2017 (UTC)

I think that would be a more straightforward vote, only dealing with the issue at hand and thus reducing side effects. I'm still not sure what cutoff date is best. We face much the same problem as when dividing English vs Middle English, but using the same (printing-press-based?) cutoff seems plausible. - -sche (discuss) 18:36, 9 May 2017 (UTC)

I still do not think that Contemporary Latin is well documented on the Internet. That is, it doesn't belong to WT:WDL, unless it's worded like The following languages are subject to the same criteria as WDLs: some Latin. New Latin till 1850 might be well documented, but attestation for it is already difficult as seen in WT:RFV#emodulo from May 2016 (2016, not 2017).

I didn't interpret WT:CFI's "should" as it's advised and preferred but not needed, but some people do as for example seen on Himmelsdéierchen and its talk page. I think there are better possible ways to deal with the shoulds and their implications for New Latin. Why not make New Latin an in-between language which is treated as neither LDL nor WDL?

-Slœtel (talk) 22:54, 9 May 2017 (UTC)

No, Contemporary Latin isn't particularly well documented on the Internet, but I think some people are inclined to treat it as effectively a constructed language and therefore subject to the same CFIs as Esperanto, Lojban, etc. —Aɴɢʀ (talk) 09:44, 10 May 2017 (UTC)

Vote start

@Metaknowledge: The start date has arrived. Shall we start the vote? --Dan Polansky (talk) 11:27, 15 May 2017 (UTC)

@Dan Polansky: Thanks for reminding me. I liked your wording, so I'll move the vote to a more appropriate name, replace it with your wording, and then we can run it. —Μετάknowledge^{discuss/deeds} 18:44, 15 May 2017 (UTC)

@Metaknowledge: Thanks. On the present vote text, I think the two terms in the vote proposal, New Latin and Contemporary Latin, should not link to Wikipedia. That implies that Wikipedia defines what Modern Latin and Contemporary Latin are, whereas in fact, the definition is actually directly in the vote text, by providing the year range. In general, I think CFI and similar policies should rarely or never link to Wikipedia. --Dan Polansky (talk) 19:49, 15 May 2017 (UTC)

@Dan Polansky: The WDL page already links to Standard Indonesian. I think on the whole that it is more informative to link than not to link, and given that the date cutoff is specified, no one could reasonably claim that the Wikipedia page takes precedence. —Μετάknowledge^{discuss/deeds} 19:55, 15 May 2017 (UTC)

You're right: Standard Indonesian is so linked. I don't think it is a good practice, though. However, it looks like a minor quibble rather than something that should make anyone oppose the proposal. --Dan Polansky (talk) 20:03, 15 May 2017 (UTC)

I replaced "sentence" with "item"; please revert if you disagree. --Dan Polansky (talk) 20:05, 15 May 2017 (UTC)

I removed what I think is an unnecessary, potentially controversial statement, nowhere key for the proposal of the vote (diff). I think the intro statement could be removed completely, and for the better. --Dan Polansky (talk) 20:08, 15 May 2017 (UTC)