Wiktionary talk:Votes/pl-2022-07/Adding descendant hubs

I think the number of descendants required for a DHUB entry should be increased. There are a lot of terms that are from unidiomatic phrases or expressions in the source language and I wouldn’t call a DHUB entry with only one descendant particularly “useful”. I’d personally prefer the threshold to be at least 2 or 3 separate / independent descendants. Maybe more than one options can be given for this? —Svārtava (talk) • 13:18, 11 July 2022 (UTC)Reply

@Svartava: Could you provide some of these phrases that have exactly one inclusion-worthy descendant so I get a better idea? — Fytcha〈 T | L | C 〉 15:27, 11 July 2022 (UTC)Reply

@Fytcha: Right now, I can't think of them but any possible DHUB entry with 1 descendant is pretty useless. If there are no such possibilities then it only strengthens the point to propose a threshold for minimum 2 descendants. —Svārtava (talk) • 16:36, 11 July 2022 (UTC)Reply

@Svartava: I can make it so that the vote has two options. — Fytcha〈 T | L | C 〉 18:05, 11 July 2022 (UTC)Reply

@Fytcha: Thanks, OK by me. —Svārtava (talk) • 00:14, 12 July 2022 (UTC)Reply

@Fytcha: Category:Japanese pseudo-loans from English is full of these (although many of them probably aren't attestable in English). 98.170.164.88 00:03, 2 November 2022 (UTC)Reply

Interesting point. Thinking about it again, perhaps an easier example would be German Bodybag where, while a donor term exists, the meaning doesn't match. Do you think there are any unintended consequences of this vote when it comes to these terms? BTW, AFAIK {{desc}} doesn't have a parameter that allows for pseudo-loans. — Fytcha〈 T | L | C 〉 00:15, 2 November 2022 (UTC)Reply

That actually is a good example for the reason you mentioned. Maybe the policy should be clear that the term has to be attestable in the source language, with the right sense. Or at least a closely linked sense, since often meanings do change in the course of borrowing or inheritance.

I think this proposal is a great idea overall, btw. 98.170.164.88 00:23, 2 November 2022 (UTC)Reply

I'm not quite sure I understand which legislative hole we're trying to fill here. English body bag exists anyway so what would the protection granted by WT:DHUB do to it? Do you mean the text is not clear about whether a non-sense sense may be added to body bag?

I think this proposal is a great idea overall, btw. Thank you! Yet another reason for you to sign up ;) — Fytcha〈 T | L | C 〉 01:48, 2 November 2022 (UTC)Reply

Oh, right. If the entry already exists then DHUB wouldn't need to be invoked to justify its existence; not sure what I was thinking exactly...

There still could be cases where a literal phrase exists and is attestable in one language (SOP, so no entry), but then it is pseudo-loaned into another language with an entirely different specific meaning not known to the original. I'm not sure how big of a practical problem that is, though.

For a specific example, Japanese ロングライド (rongu raido, “randonneuring”) < English long + ride. If there was another language that had the same pseudo-loan, would that justify the creation of English long ride, even if the original English phrase didn't have that meaning? Maybe we would even want to have the entry in that case. I'm not sure. 98.170.164.88 03:39, 2 November 2022 (UTC)Reply

That's an interesting point because I can see this as just being the working-as-intended behavior: if the same term has been pseudo-loaned into two languages, having an (attested, SOP) DHUB allows users to discover one pseudo-loan from the other. If this is the route we want to go down on, I'd rewrite the two proposals to "one non-pseudo-loan descendant or two descendants" and "two descendants" respectively (because the discoverability argument is only sound if there are two pseudo-loans whereas a single non-pseudo-loan descendant might already sufficient justification for the attested SOP donor (I can go into more detail as to why I think this is if anybody is interested)). If this is not the route we want to go down on, the two proposals would have to be such that only non-pseudo-loans count towards their respective limits. — Fytcha〈 T | L | C 〉 04:23, 2 November 2022 (UTC)Reply

I've changed proposal 1 now and I've stated my rationale on the vote page. — Fytcha〈 T | L | C 〉 14:37, 5 November 2022 (UTC)Reply

Do descendant hubs need to meet attestation requirements? WT:THUB only exempts entries from the idiomaticity requirement, not the attestation requirement.

I think descendant hub terms need to be attested too, otherwise they would be in the Reconstruction namespace. But the proposed policy text exempts descendant hubs from the entire CFI. @Fytcha could the vote text be changed to be more specific on this front? Perhaps "Multi-word terms that would otherwise be considered non-idiomatic may still be included ..." This, that and the other (talk) 23:57, 1 November 2022 (UTC)Reply

@This, that and the other: Thanks, good point, I'll fix it immediately. — Fytcha〈 T | L | C 〉 23:59, 1 November 2022 (UTC)Reply

Another thing worth considering is neo-Latin phrases that have been widely used in multiple European (and often non-European) languages, but which were probably made by some guy in the 1800s. If the phrase turns out to have had uses in Classical or Medieval Latin, does that count as attestation, or does there have to be evidence that the term was borrowed instead of invented independently?

I suspect such cases could be found by trawling Category:English terms derived from Latin, although admittedly I don't have a good example off the top of my head. 98.170.164.88 00:34, 2 November 2022 (UTC)Reply

foedus pacificum, which I recently added quotations to, would be an example of a phrase used in various European languages made up in its technical sense in the 1790s but with unidiomatic prior attestation. Not sure that makes sense as a "descendant hub" though vs. the hoc anno example linked here. —Al-Muqanna المقنع (talk) 23:35, 5 November 2022 (UTC)Reply