Wiktionary talk:Votes/pl-2017-11/Restricting Thesaurus to English

Rationale

Some benefits of restricting the Thesaurus to English:

• Easier to manage and overview.
• No need to add elements catering for multiple languages, such as:
  • Level 2 heading stating "English" or "Spanish"
  • Language indication in thesaurus page name (largely no lang indication yet; examples could be Thesaurus:Hindi/water, Thesaurus:pt:autêntico, Thesaurus:Polish:złoczyńca)
  • Language indication in thesaurus category names (no lang code yet)
  • Language code to {{ws}} such as {{ws|en|dog}} instead of {{ws|dog}} (no lang code yet)

Why keep the Thesaurus multilingual in the English Wiktionary:

• Mainspace is multilingual and so should the Thesaurus. If English can benefit from features of the Thesaurus, so should other languages even if it makes things a bit more complicated.

--Dan Polansky (talk) 10:27, 15 November 2017 (UTC)

Thanks for drafting this vote.

The main thing I wish to avoid is reduplication. That's why en.wikt doesn't host translation tables for every word in every language, but only for English words (cf. a recent example of misunderstanding). We have Wiktionaries in other languages; they're meant to be used.

I acknowledge, however, that I'm not being entirely consistent: I favour having full-blown foreign entries (with inclusion of synonyms, antonyms, etc.), even though this seems, too, like a waste of time, as that job has theoretically already been done elsewhere. I think this is somewhat of a grey area.

Furthermore, I'm holding to a rather idyllic view; I'm afraid that in practise, certain Wiktionaries won't ever attract enough contributors to hope to become a reliable resource; and that they won't even have as good a coverage of their own language as en.wikt does. Besides, certain languages don't even have their own Wiktionary. Ancient Greek would be an example; as such, I'd certainly favour hosting here Ancient Greek Thesaurus entries.

But for major languages like French, German, Spanish, Russian, whose own Wiktionary versions are thriving... no. --Barytonesis (talk) 11:02, 15 November 2017 (UTC)

As for "reduplication", you mean that, e.g., French thesaurus entries can be in fr.wikt thesaurus, and do not need to be duplicated in en.wikt thesaurus, right? --Dan Polansky (talk) 11:14, 15 November 2017 (UTC)

Yes. --Barytonesis (talk) 11:17, 15 November 2017 (UTC)

Then it is to be noted that the fr.wikt thesaurus seems to have a somewhat different structural design than the en.wikt thesaurus; en.wikt relies a lot of hyponymy/hypernymy and meronymy/holonymy to organize the material. Thesaurus:cat and fr:Thésaurus:chat/français look quite different. --Dan Polansky (talk) 11:26, 15 November 2017 (UTC)

Yes, that's a problem, and I'm afraid it's going to be impossible to achieve cross-wiki consistency. --Barytonesis (talk) 14:08, 15 November 2017 (UTC)

On an another note, if you include English and Ancient Greek but exclude Spanish and Polish, most of the rationale at the top of the thread collapses since you will need the infrastructure to cater for multiple languages anyway. --Dan Polansky (talk) 12:15, 15 November 2017 (UTC)

Yes. I personally think our decision shouldn't be guided too much by that: once the infrastructure is in place, I don't see that it should require much overhead. It's easy for me to say that though; I don't know how to develop such an infrastructure, so I won't be the one to do it. --Barytonesis (talk) 14:08, 15 November 2017 (UTC)

Current non-English entries

Languages that have at least one Thesaurus entry include Czech: WS:příbuzný, Finnish: WS:juoppo, French: WS:chat, Hindi: WS:मनुष्य, Polish: WS:gruchot, Portuguese: WS:duradouro, and Telugu: WS:కుక్క. From what I remember, Polish and Portuguese are languages with larger Wikisaurus entry counts. --Dan Polansky (talk) 10:37, 15 November 2017 (UTC)

More:

• Chinese: WS:幫助, WS:傻子, WS:非常
• Hindi:WS:यज्ञ, WS:Hindi/water, WS:Hindi/human being, WS:Hindi/God
• Polish: WS:badziewie, WS:gruchot, WS:mierny, WS:nieporządek, WS:zamieszanie, WS:znajomy, WS:złoczyńca, etc.
• Portuguese: WS:amedrontador, WS:avarento, WS:insulto, WS:habitação, WS:lisboeta, WS:menino, WS:treco, etc.
• Telugu WS:ఆయుధము, WS:ఆసియా, WS:కీటకము, WS:కుక్క, WS:కోతి, WS:క్షీరదము, WS:చిలుక, WS:చేప, WS:జంతువు, WS:దినము, WS:ద్రవము, WS:నీరు, WS:పండు, WS:పక్షి, WS:పాము, etc.

Translation subpages, another practice: WS:stupid/translations, WS:beer/translations, WS:penis/translations, WS:vagina/translations

--Dan Polansky (talk) 11:13, 17 November 2017 (UTC)

Expanded. --Dan Polansky (talk) 10:23, 19 November 2017 (UTC)

Oppose

I plan to oppose this vote. My reason is simply that words in languages other than English may have a large number of synonyms (and other -nyms) too.

I'd prefer not to overload the mainspace with synonyms and other -nyms in all languages. We can use the Thesaurus for all languages.

Conversely, if we somehow proved that all languages are fine handling many synonyms and other -nyms in the mainspace, then that system would probably work for English too and the Thesaurus namespace could be deleted altogether as unneeded. (which I don't think is the case)

We have a namespace for Rhymes in all languages and it seems fine. --Daniel Carrero (talk) 00:04, 16 November 2017 (UTC)

I appreciate the same arguments as brought by Daniel Carrero. Additionally I want to point out that it rationally bundles resources to have synonyms on the English Wiktionary rather than on every other Wiktionary. Polyglots work on multiple languages, and it is questionable to have to work in different communities with different rules and layouts each just because of synonyms – if there are such communities accessible and usable at all, which is regularly not the case for dead languages or small languages where one would be alone, or contentious languages (Serbo-Croatian having multiple Wiktionaries). Also the use of another language – English – supports description of the semantic differences. In contrast to that the arguments for restricting the en.Wiktionary Thesaurus to English are very weak. It is the editors’ business what is “easier to manage and overview”, and as lies bare, it is harder to manage and overview languages accross multiple wiktionaries. Palaestrator verborum (loquier) 11:53, 17 November 2017 (UTC)

Hover over texts in non-English entries

Hover over texts seem to be a neat language learning tool. Thus, in Thesaurus:పండు (fruit), when you move the mouse over అనాసపండు, there appears "pineapple". Thus, you can guess the meaning of అనాసపండు and then move the mouse over the item for verification. I placed even more such hover over items in Thesaurus:příbuzný.

This benefit can only be there if we have Telugu thesaurus entries in English Wiktionary; a Telugu thesaurus in Telugu Wiktionary would not provide this feature, that is, if the hover over thing would be there, it would show a Telugu gloss instead.

Admittedly, this may work well only on desktops and not on tablets and mobile phones. Nonetheless, if the non-English items in Thesaurus gain interesting coverage, someone may come up with mobile interface for this as well; the data is there is the markup to show in various ways.

--Dan Polansky (talk) 08:29, 17 November 2017 (UTC)

Insights from French Wiktionary

Hi,

French Wiktionary have a different vision of thesauri, by doing a larger collection of vocabulary around a topic, and in several languages. We made up a policy about thesauri recently. Up to now, French Wiktionary offer 453 thesauri in 54 languages, including 300 thesauri in French. So, a third of the stock is in other languages, and there is a structure that permits to include other languages easily. Thesauri in other languages than French are bilingual. It may not be a perfect solution, but I present it here as an other way to deal with this matter Noé 09:03, 25 November 2017 (UTC)