Wiktionary talk:Votes/2015-09/Creating a namespace for reconstructed terms

From Wiktionary, the free dictionary
Jump to navigation Jump to search

What should we name the namespace?[edit]

Please look at our existing namespaces for established naming conventions. --WikiTiki89 20:43, 29 September 2015 (UTC)[reply]

Was this discussion linked to from anywhere with indication that it's the discussion for choosing a namespace name? I haven't seen it. If not (and prhaps even if so), then, if and after the vote is closed successfully (which seems extremely likely at this point), at least a pointer hereto should be posted in the BP to invite discussion.​—msh210 (talk) 20:29, 4 November 2015 (UTC)[reply]

No it wasn't, and I was going to post about it at the BP after the vote closed anyway. Although I guess it would have made sense to do so sooner. --WikiTiki89 21:10, 4 November 2015 (UTC)[reply]

Unattested:[edit]

  1. Oppose Sounds like too broad a word for me. By definition, it would encompass protologisms and stuff, even [though we know] that we wouldn't really allow those in the new namespace. --Daniel Carrero (talk) 16:50, 28 October 2015 (UTC)[reply]
    By that logic, we should allow uploading Microsoft Word templates in the "Template:" namespace and pictures of appendicitis in the "Appendix:" namespace. --WikiTiki89 14:55, 29 October 2015 (UTC)[reply]
    @Wikitiki89: Good point. Let me elaborate: an "Unattested:" namespace is inviting of unattested protologisms and any words that fail RFV or for some reason don't have 3 citations. If we allowed a single namespace for protologisms (gasp!), "Unattested:" would fit.
    The other namespaces you mentioned are not so much inviting of the things you mentioned: Microsoft Word templates are not dictionary material as far as I'm concerned, and our namespaces are named after purposes and types of content, rather than based on the semantics of the content: we don't have "Appendix:" for appendicitis, as much as we don't have "Dog:" for dogs and countless other possibities. --Daniel Carrero (talk) 06:32, 30 October 2015 (UTC)[reply]
    You probably get my point, but I'll just say anyway, the fact that we name a namespace something does not mean that anything that fits under that something will automatically be allowed in that namespace. We would still have more detailed rules describing what belongs in the namespace and these rules will certainly not include protologisms. --WikiTiki89 15:29, 30 October 2015 (UTC)[reply]
  2. Oppose Per Daniel Carrero: that would encompass English unattested protologisms. --Dan Polansky (talk) 18:39, 28 October 2015 (UTC)[reply]
  3. Oppose per above: too broad a name for its intended use.​—msh210 (talk) 20:30, 4 November 2015 (UTC)[reply]

Reconstructed:[edit]

  1. Support, we don't have "Templates:" or "Categories:" after all. —CodeCat 23:51, 29 September 2015 (UTC)[reply]
  2. Weak Support this or but prefer "Reconstruction". Oppose "Reconstructions" because, as CodeCat notes, everything else is named in the singular except "Rhymes" and "Citations" where the pages are designed to store multiple rhymes/citations. A reconstruction page, by contrast, would typically store only one reconstruction. It'd be useful if we (well, the devs) could also create RC: as a shortcut. - -sche (discuss) 23:57, 29 September 2015 (UTC)[reply]
    Currently all of our namespace names are either nouns describing the content or modifiers modifying the word "namespace". "Reconstructed:" is neither of those. "Template:" and "Category:" are singular because each page can only represent exactly one template/category. A page in "Citations:" can have just one citation or more. A page in this new namespace can have just one term in it, or sometimes more. --WikiTiki89 15:04, 30 September 2015 (UTC)[reply]
    On further consideration, you and Dan are right that a noun would be preferable, as our other namespace-words (Category, Appendix, etc) are nouns. - -sche (discuss) 17:11, 28 October 2015 (UTC)[reply]
    Support Abstain, I prefer "Reconstruction". --Daniel Carrero (talk) 00:05, 30 September 2015 (UTC)[reply]
    Abstain I prefer "Reconstruction" since that is a noun, like Category or Template. But this seems acceptable. --Dan Polansky (talk) 15:59, 28 October 2015 (UTC)[reply]
  3. Support over "Reconstruction(s)", but this still feels a bit off. Collecting reconstructed terms is not quite the point of proto-languages. --Tropylium (talk) 16:29, 28 October 2015 (UTC)[reply]

Reconstructions:[edit]

  1. Support --WikiTiki89 20:43, 29 September 2015 (UTC)[reply]
  2. Support --Diego Grez-Cañete (talk) 18:58, 26 October 2015 (UTC)[reply]
  1. Oppose Plural does not match Template: and Category: namespace conventions. --Dan Polansky (talk) 15:59, 28 October 2015 (UTC)[reply]
    Later: Citations: is in plural but a page in the namespace contains multiple citatious. Rhymes: is in plural, and I am not sure whether a single page contains a single or multiple rhymes; it contains multiple items from which multiple rhymes can be construcred. I see no other namespaces named in plural. --Dan Polansky (talk) 18:44, 28 October 2015 (UTC)[reply]
  2. Oppose (see my rationale further up the page). - -sche (discuss) 17:11, 28 October 2015 (UTC)[reply]
  3. Oppose per Dan P.​—msh210 (talk) 20:31, 4 November 2015 (UTC)[reply]

Reconstruction:[edit]

  1. Support --Dan Polansky (talk) 15:59, 28 October 2015 (UTC)[reply]
  2. Support --Daniel Carrero (talk) 16:46, 28 October 2015 (UTC)[reply]
  3. Support. On further consideration, I'd even prefer this to "Reconstructed:" (see my rationale further up the page). - -sche (discuss) 17:11, 28 October 2015 (UTC)[reply]
  4. Support. I have been convinced that the singular would be better. --WikiTiki89 17:40, 28 October 2015 (UTC)[reply]
  5. Support I think I've been convinced as well. —CodeCat 17:42, 28 October 2015 (UTC)[reply]
  6. SupportAɴɢʀ (talk) 17:56, 28 October 2015 (UTC)[reply]
  7. SupportJohnC5 19:09, 28 October 2015 (UTC)[reply]
  8. Support​—msh210 (talk) 20:34, 4 November 2015 (UTC)[reply]
  9. SupportEnosh (talk) 21:32, 4 November 2015 (UTC)[reply]

*:[edit]

(That is, one page would be [[*:bʰer-]] (or, depending on discussion elsewhere, possibly [[*:Language/bʰer-]]).)

Support but I prefer "Reconstruction:". --Daniel Carrero (talk) 20:38, 4 November 2015 (UTC)[reply]

Does the software even support that as a namespace name? —CodeCat 20:45, 4 November 2015 (UTC)[reply]
Oh I misread it. I thought the idea was "*bʰer-", placing the entries in the main namespace, starting with an asterisk. --Daniel Carrero (talk) 20:47, 4 November 2015 (UTC)[reply]

Discussion[edit]

Purpose/implementation[edit]

If, as discussed, this is intended to replace all current protolanguage appendices (although currently the proposal does not say as much), I'd like to reiterate that for a namespace for "terms", deciding on "lemma forms" becomes a problem. Proto-languages more often than not do not have a set orthography, and a single entry like *ḱerh₂- could easily spawn a dozen or two "alternate reconstruction of / alternate transcription of" entries, eating up effort that could go into actual work for no clear benefit. The pages themselves are also commonly not "terms" as much as clusters of related terms, with multiple derivatives of a single root treated in the same entry.

However, if we kept the current Namespace:Lang/*whatever notation, we could continue handling alternate reconstructions as redirects or the like, and to continue grouping etymologically related data on the same page. --Tropylium (talk) 20:04, 1 October 2015 (UTC)[reply]

I explicitly mentioned that page titles are unaffected by this vote. All proto-language appendices will have only the namespace changed. What you are bringing up are real issues, but should be decided separately. --WikiTiki89 14:39, 2 October 2015 (UTC)[reply]
Moving all the pages is such a big task, we should decide on this beforehand. Otherwise we may have to move twice. —CodeCat 14:40, 2 October 2015 (UTC)[reply]
It's not that bad of a task to move them all twice, especially since after the first move we won't have to sort through non-reconstruction pages in the appendix. --WikiTiki89 15:20, 2 October 2015 (UTC)[reply]
It would also be good to move all the pages over to the new namespace as they are, preserving their histories, before doing any moves that will result in mergers. --WikiTiki89 18:55, 6 October 2015 (UTC)[reply]
Ah, sorry. Sloppy reading on my part. More discussion to be had later, then. --Tropylium (talk) 13:52, 3 October 2015 (UTC)[reply]

Language name in the title[edit]

If this vote passes, the pages for reconstructed terms are still going to use the language name in the page title?

For example: Appendix:Proto-Germanic/wadǭ would become either:

  1. Reconstructed:Proto-Germanic/wadǭ (language name in the title)
  2. Reconstructed:wadǭ (no language name in the title)

I suppose option 2 makes more sense, after all the language should be in the L2 section title. We don't have pages like English/dog. --Daniel Carrero (talk) 03:13, 15 October 2015 (UTC)[reply]

See the thread before this one. Currently, the vote is only to change "Appendix:" to "Reconstructed:", leaving everything else in place, i.e. option 1. Option 2 would require a separate discussion and consensus-finding. —Aɴɢʀ (talk) 06:28, 15 October 2015 (UTC)[reply]
As mentioned above, proto-language terms have no canonical orthography. The phonological interpretations often vary as well. I don't think there's any reasonable metric for when we could claim that two particular reconstructed terms have "the same form" and thus put them on the same page. --Tropylium (talk) 15:19, 28 October 2015 (UTC)[reply]
I agree. The characters used in the reconstruction are meaningless without context and the language name preceding it provides that context. It would not make sense at all to conflate two reconstructions of words in different proto-languages just because the characters we used happen to coincide. --WikiTiki89 15:55, 28 October 2015 (UTC)[reply]
…Or, actually, with a little bit more nuance: reconstructions in unrelated proto-languages cannot be equated especially consistently. When it comes to closely successive protolangs instead, like Proto-Polynesian > Proto-Nuclear Polynesian > Proto-Ellicean > Proto-Eastern Polynesian, or Proto-Germanic > Proto-Northwest Germanic > Proto-West Germanic, I continue to think it should be productive to generally merge such languages altogether and treat differences mainly in prose (to avoid the "*aka > *aka > *aka > *aka" problem).
This suggests that a third option might be to use the format Reconstructed:Germanic/wadǭ, to highlight that not everything reconstructed from e.g. Germanic data will come out as Proto-Germanic proper.
We could also explicitly title any exception cases: e.g. *pannôReconstructed:Proto-Northwest Germanic/pannô (while continuing to encode this as gem-pro). --Tropylium (talk) 16:53, 28 October 2015 (UTC)[reply]
I (and more problematically, our templates) would find it unintuitive for some of the reconstructions to have different-than-canonical prefixes in their pagenames; we might be better off having e.g. Reconstructed:Proto-Germanic/pannô labelled {{lb|gem-pro|Proto-Northwest Germanic}}. For *aka we could have a page for the furthest-back protolanguage and then list the subsequent stages on its descendants page only, i.e. merge the descendant pages' contents to Appendix:Proto-Polynesian/aka. Perhaps even redirect them so that links still work; otherwise, de-link the links (leave them as unlinked text) or remove them altogether. - -sche (discuss) 17:23, 28 October 2015 (UTC)[reply]
It should be Reconstruction:wadǭ, IMHO. Or even better, let's create a vote to allow *wadǭ in the mainspace, and after that vote fails, let's go ahead with the separate namespace thing. I think the reconstruction namespace, if created, should be treated as much as mainspace, and thus with no prefixing by language. --Dan Polansky (talk) 08:57, 7 November 2015 (UTC)[reply]
My preference is for Reconstruction:wadǭ as well. —CodeCat 19:33, 11 November 2015 (UTC)[reply]
Then go and create a vote. I intend to oppose, however. --WikiTiki89 20:12, 11 November 2015 (UTC)[reply]