Wiktionary talk:Votes/2021-02/Expanding CFI for place names

Attributive use
When I read “three citations of attributive use”, my first thought was that this referred to the attributive use of a noun phrase as the modifier of a compound noun, such as “the Paris scene” and “a scenic Rhine River cruise”. Perhaps replace “attributive use” by “figurative use”? --Lambiam 16:09, 11 February 2021 (UTC)
 * Good call, I've updated the wording. Ultimateria (talk) 17:32, 11 February 2021 (UTC)

LDLs
The three-figurative-citations criterion makes sense for well-documented languages, but what about limited-documentation languages? I'd think they should have a lower threshold, or perhaps just a different policy, like including all attested place names (given that even the names of obscure manmade structures can serve as significant evidence if appearing in the corpus of a sparsely attested extinct language). — Vorziblix (talk · contribs) 18:47, 11 February 2021 (UTC)
 * I can't believe I overlooked that! I've opted to remove the requirement entirely. Let me know if it needs to be changed in any way. Ultimateria (talk) 22:37, 11 February 2021 (UTC)
 * After reading this I have a question—What is the criterion for fictional locations (and characters, etc.) for LDLs (e.g., Middle English)?—Is it the same as for WDLs (i.e., WT:FICTION)? J3133 (talk) 07:24, 12 February 2021 (UTC)
 * I imagine it's just one cite. Number of citations is surely a higher-level policy than Fictional universes. Any updates to FICTION should make this explicit. Ultimateria (talk) 08:05, 12 February 2021 (UTC)

I am not seeing a good rationale for this statement: “The figurative use requirement for place names not listed above does not apply to limited documentation languages. All place names in these languages shall be included if they fulfill attestation requirements.” If obscure bridges, roads, etc., are undesirable to be included for well-documented languages, then they are similarly undesirable for LDLs. This part needs to be redrafted to ensure that obscure places are not inadvertently included by this statement. — SGconlaw (talk) 06:40, 14 February 2021 (UTC)
 * LDLs simply have less documentation, so the notability criterion would exclude entries that we do in fact want. With this rule, it is inevitable that a few truly obscure bridges and roads will be included, but we will avoid having to exclude truly notable ones, like 🇨🇬. —Μετάknowledge discuss/deeds 07:39, 14 February 2021 (UTC)
 * In that case can't the policy be worded more explicitly? Perhaps something along the following lines: place names not listed above [in the policy], including manmade structures, may only be included either (1) if there are at least three citations of figurative use; or (2) there is consensus that they are historically significant. Then the policy can apply to both LDLs and well-attested languages. — SGconlaw (talk) 07:58, 14 February 2021 (UTC)
 * The rationale is now on the vote page. What does "consensus that they are historically significant" look like? An RFD debate? That's the outcome I'm most hoping to avoid with this vote. (The language I'm most eager to remove is "other significant natural geographic features" and "there is currently no definition of 'significant natural geographic features'".) I'd rather have something concrete in CFI, and I don't think it's an issue to give LDLs more freedom in attestation; it's a modicum more of the special treatment we've agreed to give them. Ultimateria (talk) 18:44, 14 February 2021 (UTC)
 * well, yes, I was thinking that in such cases there should be an RFD discussion ... I don't really know what the situation is with LDLs – maybe your proposed policy won't be a problem because non-notable places in such languages are unlikely to be included? — SGconlaw (talk) 18:58, 14 February 2021 (UTC)
 * Exactly. Ultimateria (talk) 19:36, 14 February 2021 (UTC)

Neighbourhoods
What's the rationale for including neighbourhoods without figurative use, while excluding e.g. streets? I know there are some famous neighbourhoods and districts like Queens, but for every one of those there are probably a dozen neighbourhoods in smaller places which have been mentioned three times by different journalists in local or nearby papers. I don't feel strongly for or against including them, though, since they will often be bigger than very small villages, and I suppose a size cutoff (for villages or for neighbourhoods) is probably unwise or unwieldy in practice. (I almost wonder whether people should vote on which individual bullet items they want to be subject to the looser vs the tighter requirements, but I realize that'd get unwieldy.) - -sche (discuss) 04:59, 12 February 2021 (UTC)
 * One argument for them is that they tend to have names that are not readily interpretable; just within the borough of Brooklyn (itself under this rule), there are neighbourhoods like Williamsburg, Coney Island, Bensonhurst... coming across these out of context, how would someone know those weren't towns (or islands, as the case may be)? Their unique names also mean that they potentially carry useful lexicographical information (etymology, pronunciation). With Atlantic Avenue and Fulton Street, you may not know they're in Brooklyn, but you know one's an avenue and the other's a street, so the chance someone would look them up seems far lower. —Μετάknowledge discuss/deeds 07:32, 12 February 2021 (UTC)
 * That may be the case for street names in English-speaking territories, but it is not so clear, perhaps, that ', ' and  are street names. --Lambiam 23:04, 15 February 2021 (UTC)
 * Same with in Brooklyn, I suppose — but many of these exceptional streets are so notable that we could find figurative use, and for those that aren't, I won't be losing any sleep over excluding them. —Μετάknowledge discuss/deeds 23:14, 15 February 2021 (UTC)
 * One possible rationale could be that neighbourhoods are more likely to be used attributively, like seen in “a Bronx accent”, or Erdoğan calling himself a “Kasımpaşa man”. We might also consider expanding “figurative use” to cover such cases, where the sense is that of some quality – not necessarily made explicit – ascribed not so much to the named place itself, but to entities stemming from that place. Or is that already implied in “figurative use”? --Lambiam 23:36, 15 February 2021 (UTC)

All place names not listed above
So there are place names not listed above which are not manmade structures. What would it be? I doubt we intend to have a numerus clausus for unforeseen constellations that requires that all such place names, which perhaps nobody can imagine, can only be included on figurative use. It appears to me that the sentence “All place names not listed above, including manmade structures, shall be included if they have three citations of figurative use that fulfill the requirements listed under the Attestation section of CFI.” should be removed altogether, as repetitive to the following sentence; this second sentence already says “may only be attested through figurative use [scilicet under the usual attestation requirements]” and clearly lists kinds of place names for which this applies; the first sentence says the same but additionally relates to a broader group (which is dark) and incorporates the three-durable-uses requirement which should not be referred to at this point at all but one should refer dynamically to the attestation criteria in force at any point (anticipating possible amendments of them, so this toponym-specific text doesn’t have to be changed when the attestation criteria are expanded). Fay Freak (talk) 19:39, 14 February 2021 (UTC)
 * If we can't imagine a future category of place name, I don't know how we can let it be allowed immediately under CFI while also being concrete about what we will and won't include now. The potential for future categories is precisely why we need a catchall to allow them even on a limited basis. And as I've said in the rationale, I've organized the list to make it easier to add future categories.
 * I've removed the clause "including manmade structures" which became redundant after several rounds of revisions. However I've kept "that fulfill attestation requirements" (after simplifying the wording). It's important to avoid ambiguity here even if it seems redundant. Read it again; I haven't referred to durable archiving, but broadly to "attestation requirements", which is dynamic. Ultimateria (talk) 20:09, 14 February 2021 (UTC)
 * The text now runs a bit more smoothly, but you have little changed in its ambiguous scope. I have reread it. So now I notice I could understand “all place names not listed above shall be included if they (have figurative use)” as formulating a sufficient condition when something is includable. The words “three citations of” are still redundant, because those are the attestation requirements, right, and the current ones, so the reference is not dynamic.
 * It is a truism though that something is includable when figurative and not only toponym. Likely one will still understand this sentence as formulating a necessary condition. Maybe with this text  and are not includable because they are not in the bullet list (for “it only contains places on earth”) nor manmade structures nor figuratively used. So you see why I reckoned that you should only positively formulate, one group which can be included if attested and another if attested figurative, about the rest be silent. What you are trying to tell us with “All place names not listed …” you do not even know yourself. If you don’t know then don’t explicitly allow or forbid. Fay Freak (talk) 20:27, 14 February 2021 (UTC)
 * I acknowledge that you find it redundant, but it's only three words and it's worth avoiding ambiguity.
 * The only other place where attributive use applies is WT:FICTION. It's not a truism if it's not our existing policy. The names of planets are already not explicitly allowed by CFI. I would like to rectify that if this vote passes, but I don't want to attempt too many changes at once. Perhaps the ideal situation is to list every possible place name (not just mountains, but mesas, plateaus, buttes et al...). However I don't find this approach feasible at all. We would surely forget some items and have to add them in votes, and disagreement over granular specifics would almost certainly keep any proposal from passing. There's a reason CFI is not full of extremely detailed sections. Ultimateria (talk) 21:03, 14 February 2021 (UTC)
 * An example of a not explicitly listed category of features of natural origin is formed by named ravines, such as . --Lambiam 23:40, 15 February 2021 (UTC)
 * It would be included as a geological landform, bullet point 3. Ultimateria (talk) 00:42, 16 February 2021 (UTC)