Wiktionary:Votes/2023-11/Ordering of etymologies within an entry

Ordering of etymologies within an entry
Voting on: adding text to Entry layout to provide recommendations to editors on how multiple etymology sections within an entry should be ordered. In particular, we propose that the following text be added at the end of the section:


 * When multiple etymologies are present, we recommend that - whenever possible - they are ordered by descending commonness in the language, so that the most common etymology is at the top. This may differ from the order given in a native monolingual dictionary. We recognize that editors won't always have sufficient information or context to order etymologies that way, so we encourage the application of this guidance on a best-effort basis.


 * We also recommend that etymologies for lemmas precede etymologies for non-lemma forms. Languages on Wiktionary include non-lemma forms to varying degrees (if at all); however, it's a universal goal to increase the coverage of lemmas. In narrow cases, non-lemma etymologies can precede lemma etymologies, such as when a very common non-lemma form happens to have a rare or obsolete lemma homograph. Additionally, a lemma derived from a non-lemma form can follow that non-lemma form. Further deviations from this recommendation are allowable subject to editor consensus, reached on an entry's talk page, Beer parlour or another forum.

Note: ordering of POS sub-sections within the same etymology section is out of scope for this vote. This vote is strictly about the ordering of multiple etymology sections relative to each other.

Schedule:
 * Vote starts: 00:00, 20 November 2023 (UTC)
 * Vote ends: 23:59, 19 December 2023 (UTC)
 * Vote created: Chernorizets (talk) 00:05, 13 November 2023 (UTC)

Discussion:
 * [[Image:Wikt rei-artur3.svg|20px]] 
 * [[Image:Wikt rei-artur3.svg|20px]] Wiktionary talk:Votes/2023-11/Ordering of etymologies within an entry

Support

 * 1) . Makes definitions look tidier. CitationsFreak (talk) 01:13, 20 November 2023 (UTC)
 * , as proposer. Chernorizets (talk) 01:49, 20 November 2023 (UTC)
 * : looks good, no objections from me. This way makes a lot of sense and will look much nicer overall. LunaEatsTuna (talk) 12:44, 20 November 2023 (UTC)
 * : Surprised this is not already a thing honestly. AG202 (talk) 14:10, 20 November 2023 (UTC)
 * 1)  in general though I think the wording could be a little tighter (e.g. "should normally" rather than "we recommend"). —Al-Muqanna المقنع (talk) 19:48, 20 November 2023 (UTC)
 * @Al-Muqanna I chose the "we recommend" wording for a few reasons:
 * with over 4,300 languages on Wiktionary, I didn't want to assume that what made sense for some languages (e.g. those discussed in the BP thread) would necessarily make sense for all languages. I don't have examples OTOH of languages where my proposed ordering would be unnatural or otherwise problematic, but then again I'm only really familiar with a handful of languages.
 * many of those 4,300 languages are worked on by a single editor, who is often not a native speaker, and who may simply not have the resources available to make judgments like which etymology is more common. I don't want those editors to feel like their entries are non-compliant with WT:EL through no fault of their own and despite doing their best.
 * since my proposal requires human judgment, it's not bot-enforceable. Since, anecdotally, many entries have etymologies ordered by some other principle (e.g. etydate), it is likely not practically human-enforceable either. I hesitated to imply that all existing entries whose etymologies are not in most-common-first order would be non-compliant, per my first two points, and because I don't have numbers.
 * If you feel like those concerns aren't at odds with a tighter wording, I'd be happy to get your thoughts. Chernorizets (talk) 20:53, 20 November 2023 (UTC)
 * Those concerns make sense, but I think the problem with "we recommend" is that it's vague enough to just be ignored in practice. A wording like "normally" allows some room for flexibility while still making clear that there should be good reasons, or a consensus within the relevant community, to depart from the norm. I support this in any case though. —Al-Muqanna المقنع (talk) 21:18, 20 November 2023 (UTC)
 * I also agree, especially as, as we've seen many times, people will take any loose wording to purposefully disregard the policy entirely. AG202 (talk) 03:07, 21 November 2023 (UTC)
 * @AG202 @Al-Muqanna would you mind proposing an alternative wording, perhaps on the Talk page, keeping the issues we've collectively raised in mind? Also, I'm not sure when it is appropriate to change the wording given that people have already voted on the current wording. What's our usual practice? Chernorizets (talk) 03:53, 21 November 2023 (UTC)
 * 1) . I concur with @Al-Muqanna. I also think we should do the same for the order of definitions and the order of POS headers. —Caoimhin ceallach (talk) 20:21, 20 November 2023 (UTC)
 * 2) . However, I suggest replacing "commonness" with "recognizability" (which is not exactly the same thing). Ioaxxere (talk) 21:08, 20 November 2023 (UTC)
 * 3) . -- Huhu9001 (talk) 02:35, 21 November 2023 (UTC)
 * 4) . ―Svartava (talk) 04:00, 21 November 2023 (UTC)
 * 5) . --Saranamd (talk) 08:07, 21 November 2023 (UTC)
 * 6) . Whoop whoop pull up Bitching Betty ⚧️ Averted crashes 01:51, 22 November 2023 (UTC)
 * 7)  as long as it keeps the wording reflecting that these are general rules of thumb, not strict and inflexible obligations. are is a nonlemma form and are is a lemma, but they should not switch places. —Mahāgaja · talk 13:20, 22 November 2023 (UTC)
 * 8)  Theknightwho (talk) 20:17, 22 November 2023 (UTC)
 * 9)  Vininn126 (talk) 23:51, 22 November 2023 (UTC)
 * 10)  Anarhistička Maca (talk) 09:04, 23 November 2023 (UTC)
 * 11)  lattermint (talk) 17:48, 24 November 2023 (UTC)
 * 12)  — Vorziblix (talk · contribs) 19:49, 6 December 2023 (UTC)
 * 13)  — Fenakhay ( حيطي · مساهماتي ) 14:02, 9 December 2023 (UTC)
 * 14) . &mdash; excarnateSojourner (talk &middot; contrib) 03:29, 10 December 2023 (UTC)
 * 15)  — well thought through and nicely written. John Cross (talk) 21:04, 12 December 2023 (UTC)
 * 16)  — I like that is allows for exceptions, and it is generally well stated. Cnilep (talk) 03:47, 15 December 2023 (UTC)

Oppose

 * 1)  this proposal as it is currently written. I agree with the recommendation that "etymologies for lemmas precede etymologies for non-lemma forms" (as exemplified by 🇨🇬p.l. and 🇨🇬,p.l. ,p.l. ,p.l. and p.l.). I also agree with the recommendation that "a lemma derived from a non-lemma form can follow that non-lemma form" (as exemplified by ,p.l. both Russian and Ukrainian). I disagree with the recommendation that "non-lemma etymologies can precede lemma etymologies  when a very common non-lemma form happens to have a rare or obsolete lemma homograph" (as exemplified by 🇨🇬p.l. listing the non-lemma  before the lemma ). I disagree with the entire first paragraph of the proposal. I believe that etymology sections should be listed in order of historical development; however, listing non-lemma sections after lemma sections is a reasonable exception to that principle, which I support. 0DF (talk) 10:41, 20 November 2023 (UTC)
 * @0DF heh, was actually the entry which generated the offline discussion that I eventually brought over to Beer Parlour. To me, it's the opposite of what I recommend in paragraph 2 of the proposal, since the lemma in this case is archaic and dialectal, whereas the non-lemma form is current and common, and therefore a likelier target for a user doing a Wiktionary lookup.
 * The lemma for the non-lemma form goes back to the proto-language, so it's not obvious which etymology came first historically. However, it's still the case that even as an educated native speaker, I'd never heard of  as a type of knife, and I immediately thought of the fish. More generally, I think having to do work in order to get to the most common thing is not a user-friendly experience for an online dictionary. Research in other areas of online life tends to show that people want to be in and out quickly when using something that's supposed to be a tool or convenience. I think an etymology order based around commonness is a better match for that. I'd still be interested in finding out why you think a more chronological order might be better. Chernorizets (talk) 11:03, 20 November 2023 (UTC)
 * I'll respond to this on your talk page at User talk:0DF (eventually!). BTW, you may want to note Equinoctis correction of the proposal text at Wiktionary talk:Votes/2023-11/Ordering of etymologies within an entry. 0DF (talk) 23:58, 19 December 2023 (UTC)
 * Of course, the Bulgarian skeleton/perch form preserves user-unfriendliness by requiring a second click to actually get a hint of the meaning.
 * Ordering etymology sections by etymological development only makes sense when the etymology sections are being abused, e.g. to separate cognate words with the same spelling but different pronunciations, as with read. --RichardW57m (talk) 14:51, 20 November 2023 (UTC)
 * 1)  weakly. Overall I don't object to the proposal, but I feel it should be made clearer that the order of precedence should be as follows: 1. Etymologies for lemmas should precede etymologies for non-lemma forms. 2. Among etymologies for lemmas, they should be arranged by commonness of the terms in the language. (Not entirely sure how this is to be determined, but I suppose terms in common use should precede terms which are only dialectal or obsolete.) 3. However, this should be subject to the principle that etymologies should be arranged by historical development, as  has pointed out above. What I mean is that if, for example, if a noun is derived from a verb, it seems logical to give the etymology for the verb first—even if the verb is now obsolete—followed by the etymology for the noun (which should then refer to the earlier etymology and need not repeat common portions of it). — Sgconlaw (talk) 22:21, 20 November 2023 (UTC)
 * I would advocate confining everything related to historical development to the etymology section itself. Order the headers by commonness and mention in the etymology, if it isn't already obvious, which sense derived from which. —Caoimhin ceallach (talk) 22:55, 20 November 2023 (UTC)
 * as written. I would support a rewrite that said etymologies that develop from older ones should come after the original, but otherwise they should be sorted by frequency. פֿינצטערניש (Fintsternish), she/her (talk) 19:05, 12 January 2024 (UTC)
 * @פֿינצטערניש This vote has already ended. Theknightwho (talk) 19:35, 12 January 2024 (UTC)

Abstain

 * 1)  I support Sgconlaw's way of writing this. I am not really sure if I should be against this or not. However, I just found one more example that really shook me- Yunan. Check out Citations:Yunan where I have a list of every major modern news agency plus a few books from the 19th and 20th centuries and etc mistaking "Yunnan" province as "Yunan" province. I don't know if (probably not?) this proposed policy might be construed to "recommend" that the misspelling of "Yunnan" province as "Yunan" province appear before the legitimate use of "Yunan" (which is a real name of a real county in southern China). It's almost a certainty that this word "Yunan" appears as a misspelling of the name of "Yunnan" province much more "commonly" than as the legitimate name of "Yunan" county. I don't know how I feel about any of this. Do with this what you will- reorganize that Yunan page if you feel so inclined. I have no firm opinions. I think Sgconlaw's way of writing this is much better than the way this is currently written. --Geographyinitiative (talk) 12:38, 29 November 2023 (UTC) (Modified)

Decision
Passed 21-2-1. @Benwing2 or another admin to please make the change to WT:EL, including Equinox's suggestion to use emdash instead of hyphen if/as appropriate (see the Talk page). Chernorizets (talk) 01:31, 20 December 2023 (UTC)


 * @Chuck Entz for visibility. I'm not sure what the process is for changing WT:EL. Chernorizets (talk) 01:40, 21 December 2023 (UTC)