Wiktionary talk:Votes/pl-2010-03/Placenames with linguistic information are accepted

Questions
Is there a need to specify that sum-of-parts placenames like Hamilton Street, Mount Hamilton, South Hamilton are not acceptable without attributive use? It seems clear anyway since the next CFI section "Names of specific entities" gives New York as an example.

Does somebody want to add "information about grammar such as gender and inflection" as one of the requirements?--Makaokalani 13:05, 30 March 2010 (UTC)


 * Definitely inflection (knowledge of IPA/SAMPA and etymology is often too specialist). I'd also explicitly state that in case of sum-of-parts or multiple-word etymologies the etymological requirement should not apply because that could be completely automated. The purpose of these requirements is, unless I'm not mistaken, to force contributors to focus on relevant content, and not mass-create relatively useless entries. Entry quality prerequisites should be carefully established so to disallow any form of abuse. --Ivan Štambuk 01:03, 31 March 2010 (UTC)


 * I've added inflection. How would you word the etymology rule: "For multiple word placenames, an etymology does not count as one of the justifications"? --Makaokalani 16:40, 31 March 2010 (UTC)


 * One thing, though. The complete entry (in the final stage) should have the gender, inflection and other linguistic thing but sometimes the gender is unknown even by native speakers (rare or foreign words where gender can be unpredictable or can vary by dialect or level of education) or not so easy to create (inflection). These cases can be flagged as needing attention. --Anatoli 04:23, 31 March 2010 (UTC)


 * I can't believe anybody would be mean enough to delete your contributions for that.--Makaokalani 16:40, 31 March 2010 (UTC)

Problems with legalese
I knew I'd get stuck with the wording somehow. Is the reference to the attributive use rule necessary? And the section "Names of specific entities" says this:
 * A person or place name that is not used attributively (and that is not a word that otherwise should be included) should not be included.

It seems to contradict, or anyway confuse, my proposal. Could I include in the vote the change of words here:
 * A person or a multiple word place name that is not used attributively (and that is not a word that otherwise should be included) should not be included. --Makaokalani 16:40, 31 March 2010 (UTC)

After sleeping on it, I've just decided to pretend that sentence isn't there. New York delicatessen seems like a weak reason for including New York anyway. What is the formal reason for excluding Hamilton Street - sum of parts? no citations? not idiomatic? English street names would only pass one of the five requirements of this vote (pronunciation), but street names do have declensions in other languages. If there is a logical gap in the wording of the vote please tell me before I start the vote. --Makaokalani 12:20, 1 April 2010 (UTC)
 * Well, I think that Hamilton Street is usually considered as a name, but not as a word. This is a good reason not to include it. However, some odonyms (street names) are words, such as Strand or Canebière. On the other hand, New York is the name of a town, and such names are considered as words. If, in some languages (maybe German, I'm not sure?), all odonyms are considered as words, I think that they should be allowed for this language, but this should be discussed, because the definition might be an issue. Lmaltier 18:00, 9 April 2010 (UTC)
 * &#x5b;Citation needed ] --Bequw → τ 00:25, 10 April 2010 (UTC)

Couple of issues
I suppose you can pronounce anything, so any entry can have a pronunciation. But that's not a bad thing, just a point. Slightly more problematic is that non-English entries don't have translations, so something like Londres: will have more trouble passing CFI then London:. Mglovesfun (talk) 12:26, 1 April 2010 (UTC)
 * Translations are a major reason for including placenames, and there's no reason to make foreign placename entries more difficult than English ones. Londres now meets the requirement 4.--Makaokalani 09:31, 3 April 2010 (UTC)

Explicit additional requirements
Proposal: replace
 * Placenames are subject to the same criteria for inclusion as any other terms.

with
 * Placenames are subject to the criteria for inclusion specified in the section "General rule", extended with the following additional requirements.

The points is that the replaced sentence is not really true if it is modified in the subsequent sentences. --Dan Polansky 06:48, 2 April 2010 (UTC)

That would mean that "However," should be removed from the subsequent sentence, to the following result:


 * Placenames are subject to the criteria for inclusion specified in the section "General rule", extended with the following additional requirements. A placename entry should initially include at least two of the following: ...

--Dan Polansky 06:58, 2 April 2010 (UTC)


 * Changed. --Makaokalani 09:31, 3 April 2010 (UTC)

Question
Simple question: Does this proposal call for the immediate deletion of virtually all existing place name entries? It seems to, requiring multiple sections of linguistic content, which most entries just don't have. Did I miss something? --Yair rand 06:10, 9 April 2010 (UTC)


 * Of course not. Existing entries cannot be deleted on sight. If somebody decides to rfd them, one by one, there is good time to add the missing info, or to argue about the old CFI. ( Adding the info is probably simpler and faster.) Right now if somebody rfd's a placename that doesn't meet the attributive use CFI, there is no way to keep it. Instant deletion is only recommended for new entries of the type "A town in Texas", with nothing else. Otherwise we could be flooded with automated entries, possibly done by bot. If there is some info - etymology, for example - it would depend on the patroller.


 * Why do you say that virtually all existing entries don't meet the new CFI? Names of important places usually do, or they are easy to fix. The rules are not very demanding, if you look at them closely. A place that has not acquired translations is unimportant, so isn't it fair to ask for an etymology and pronunciation at least? And if you don't know them, why not wait until you do? Also, you can add a foreign name to a translation table even if you don't have more information about it.


 * The purpose of this proposal is to:
 * Give clear rules for the inclusion of placenames.
 * Make placename entries better by concentrating on linguistic information
 * Prevent creation of new, useless encyclopedic entries.--Makaokalani 13:18, 9 April 2010 (UTC)
 * Very many non-English place name entries (probably most, in fact) contain nothing more than a definition line. Whether or not the definition line counts as part of requirement #4 (which is left unclear by this proposal), many would be deleted. Very many English place names only contain one translation and nothing else, or just contain a pronunciation or etymology. With this proposal implemented, a huge amount of useful content would no longer pass CFI. --Yair rand 19:43, 9 April 2010 (UTC)


 * We could add a clause that the new CFI rules don't apply retroactively, if that's a problem. Personally I wouldn't mind if all the entries not passing this CFI would be deleted on sight. There is absolutely no loss at all in deleting such entries, because they are simply mini-wikipedia stubs. The purpose of this policy is to enforce having quality entries with relevant lexicographical data. The sooner we set the bar higher, the better. --Ivan Štambuk 16:57, 9 April 2010 (UTC)
 * As a general rule, I don't like the idea of grandfather clauses in the CFI. (There may be some exceptions I'm not thinking of.) If we think something is inclusionworthy, it is, and if we think it's not, it's not, irrespective of when it was written. &#x200b;—msh210℠ 17:16, 9 April 2010 (UTC)
 * That type of black/white reasoning doesn't scale well to the real-world ~_~ I don't think that anyone is willing to waste time chasing current placename entries not passing the proposed amendments anyway.. Their number is at best a few hundred. Having a clause that will temporarily allow their inclusion until their non-inclusivity is phased out through gradual improvement isn't going to hurt anyone. When (basically) all such entries are eventually improved upon, the exemption clause could be easily terminated, even without a vote. We could even explicitly state that within it (sort of a self-modifying law). --Ivan Štambuk 00:11, 10 April 2010 (UTC)

And demonyms?
An important information about placenames (at least in some countries) is the associated demonym. Why not mentioning it?

This proposal is not clear: it seems to restrict further the inclusion of placenames, while some consider that it allows more placenames. And I don't see why placenames should follow special rules (this includes definitions: obviously, a definition such as a place name is about as useless as an animal for Labrador retriever). Lmaltier 17:51, 9 April 2010 (UTC)


 * I guess that demonyms, as well as other morphologically derived words (e.g. relative or possessive adjectives) are not covered by this because they're 'normal' words. I already mentioned that as an argument for inclusion of placenames: it would be ridiculous to have derived terms of a particular placename (e.g. New Yorker, Chicagoan..), and not the main entries that these are derived forom (New York, Chicago..). Making demonyms and adjectives out of placenames is very tricky in most languages, because these are often non-intuitive formations that need to be memorized by heart, often specific dialectal forms. This way, if we include placenames too, the reader would know where to find the sense he was looking for of 'inhabitant of X', or 'pertaining to X'.
 * Also, note that this proposal is not more restrictive, but is instead less restrictive than the current CFI which only permits attributive usage of a placename. I agree that there shouldn't be some kind of a special treatment, but the big problem is the quality of such unrestricted placename entries which could be easily generated en masse (in terms of hundreds of thousands) by a bot, without adding much lexicographically relevant content, at least not anything that couldn't already be found at wikipedia... We need to add some kind of a restriction to force adding only 'useful' type of content from the perspective of a language learner/explorer or a translator. --Ivan Štambuk 00:02, 10 April 2010 (UTC)

Multi-word terms
Under the proposal how are the following terms to be handled:
 * 1) North Carolina (common name);
 * 2) State of North Carolina (official name);
 * 3) Province of North Carolina (former official name);
 * 4) Tarheel State, Tar Heel State, Old North State (vernacular names or nicknames);
 * 5) West New York (a jurisdiction in New Jersey);
 * 6) East New York (a well-recognized neighborhood not corresponding to any jurisdiction);
 * 7) King's County, Kings County, Borough of Brooklyn (co-extensive entities with Brooklyn, NY)
 * 8) Long Island (island, the 2 counties on the island not part of New York City)
 * 9) Long Island Sound {body of water)
 * 10) Niagara Falls (natural feature, park, city in New York, city in Ontario, electoral district, provincial electoral district)
 * 11) Rocky Mountains (ill-defined area);
 * 12) Rocky Mountain National Park (overlapping with other local places);
 * 13) Port of Baltimore (overlapping jurisdiction, body of water, body of water plus adjacent land, same plus facilities).

Is the specific referent to be a valid definition? If so, then each is ipso facto not SoP. DCDuring TALK 16:04, 16 April 2010 (UTC)
 * Nitpick: Kings County is the county, not King's. And you say Borough of Brooklyn is coextensive with Brooklyn, but then I don't know what Brooklyn is. As far as I know it's only a borough. &#x200b;—msh210℠ 16:30, 16 April 2010 (UTC)
 * AFAICT these are all (if correct) admissible under the proposal. FWIW see also User:Msh210/Placename:New York (which was created in response to not this vote but Votes/pl-2010-03/Placename namespace). &#x200b;—msh210℠ 16:30, 16 April 2010 (UTC)
 * The incorrect ones, if attestable, would be equally includable, either as alternative spellings or misspellings. And presumably eye-dialect spellings such as New Yawk, former names New Amsterdam and City of Brooklyn. Would religious districts be included (eg, Diocese of Brooklyn, Holy Innocents Parish)? I sense that we have just begun to scratch the surface of the possibilities.
 * I don't see why not. The reason I linked above to my subpage was to give an idea of what can be listed s.v. "New York". (That was under the other proposal-vote, but AFAICT would apply here, too, as long as pronunciation of New York and some etymology are given. (The instant proposal doesn't say anything about the criteria (two of etymology, inflection, etc.) being needed for each sense separately, and implies that they're not, since one of the criteria is an additional definition not as a place name, which, of course, does not apply sensewise.)) It has 22 definitions, including a state, a city, a county, some informal senses (synonym of Manhattan Island, of Manhattan, and of NYU), some historical senses (synonym of Teraina, and a province of England in North America), the region with ZIP codes 100xx–102xx, and several ships of the U.S. Navy. &#x200b;—msh210℠ 17:11, 16 April 2010 (UTC)