Wiktionary:Beer parlour/2019/February

Proposal: Separate namespace for entries in Category:Chinese terms written in foreign scripts
The main issue: Various discussions from the past: Chinese loanwords that were written in foreign script were originally used only for technical terms such as, , but the advent of globalization has introduced terms such as , , man into the Chinese language. Many of these are in colloquial use, but appears to be unregulated.
 * Talk:Thames河
 * Talk:KTV
 * Talk:Cookie
 * Requests for verification/Non-English (to be merged with Talk:APP) and User talk:Wyang
 * Beer parlour/2018/December

As a dictionary that aims to describe all words of all languages, it would be useful to include such entries, particularly entries such as fighting which has a different meaning from what one would usually expect.

However, of late, this has turned into a rather contentious issue. The main arguments were (1) Chinese terms should be written in Chinese script, not foreign script (2) Chinese terms written entirely in foreign scripts are code-switched. KevinUp (talk) 02:48, 1 February 2019 (UTC)

Prelude: It seems that KTV, which had passed RFV in 2014, was recently removed from Wiktionary for being "not Chinese". I wish to point out that KTV was among the 39 pioneer entries listed in (Xiandai Hanyu Cidian, 3rd edition, 1996) under its appendix for lemmas that begin with the Latin script.


 * The following was listed after the definition for KTV: I'm not sure whether "" is an abbreviation of Chinese  or Japanese, but it should be consistent with the "K" used in . Note that we already have an entry for K.

On the other hand, the appendix for lemmas that begin with the Latin script in (Xiandai Hanyu Cidian) has expanded from 39 entries (3rd edition, 1996) to 239 entries in the 6th edition (2012). Of the 239 entries, 226 entries were capitalized, while only 7 entries -, , , , , , ) were not fully capitalized (the remaining 6 entries contained Greek α,β,γ). For comparison, the original 39 entries found in the 1996 edition are listed below: KevinUp (talk) 02:48, 1 February 2019 (UTC)




 * Of these, I found that the following entries were not found in the 6th edition (2012)



The volatile nature of such entries (note the removal of Internet in the 6th edition) prompted me to come up with the following proposal:

Proposed solution: A separate appendix for Chinese loanwords that are written, either partially or fully in foreign script will be created. These "entries" will have full etymology, pronunciation etc, similar to what we have for English snowclones such as "X is the new Y", "have X, will travel" which are listed in a separate appendix. The template will then be used to redirect entries such as 卡拉OK to a separate namespace such as Appendix:Chinese terms written in foreign scripts/卡拉OK or Appendix:Foreign words used in Chinese/卡拉OK, which is up to the community to decide. KevinUp (talk) 02:48, 1 February 2019 (UTC)

Comments
I don't like the idea of moving these to an Appendix; the Appendix has poor findability. I stand by what I wrote in the 2018 BP thread. —Suzukaze-c◇◇ 03:13, 1 February 2019 (UTC)
 * (I am partial towards User:Fay Freak's idea of adding "code-switching quotes". —Suzukaze-c◇◇ 03:27, 1 February 2019 (UTC))
 * I agree with Suzukaze-c for the most part. Also, putting all these words of varying degrees of acceptance into Chinese (e.g. 卡拉OK vs. part-time) in the appendix seems to sweep everything under the rug and would not be dealing with the core of the issue. — justin(r)leung { (t...) 03:30, 1 February 2019 (UTC)
 * It seems that "Appendix:Snowclones/X is the new Y" has the proper categorization (Category:English lemmas, Category:English phrases). The only difference is the title of the page looks different. Yes, it's a bit hard to search for "X is the new Y", but for Chinese entries we'll use so it is still searchable.
 * The reason of moving such entries into an appendix is for obvious reasons: Chinese entries are generally not written using foreign scripts, unless it is a transliteration. This does not solve the issue of whether or not an entry is part of code-switching. For me, code-switching hold true for overseas communities, but most of the people in mainland China do not speak much or any English at all. KevinUp (talk) 04:05, 1 February 2019 (UTC)
 * Why attach so much significance to whether an entry is in the appendix namespace? Are ordinary users supposed to somehow know what that means? As you say the categories are the same. DTLHS (talk) 04:07, 1 February 2019 (UTC)
 * One reason for this is because we want people to view Wiktionary as a serious project. It does feel awkward to have iPhone among Danish, French, Portugese, Spanish, etc. Another reason, some of these lemmas come and go, e.g. Internet which was found in the 3rd edition (1996) of but removed in the 6th edition (2012). If we have a separate namespace we can better monitor such entries. I'd like to mention that KTV was recently removed without any formal discussion (despite passing RFV in 2014), and the etymology of KTV contains errors (MTV = Movie TV?) KevinUp (talk) 04:27, 1 February 2019 (UTC)
 * I say add usage notes to all entries of this type explaining the situation or just link to an explanatory page, which explains how the situation is controversial. 'iPhone' is used by Chinese people all the time. I don't care if it is considered Chinese or not, but there's no reason for Wiktionary to ignore the fact that Chinese people use that word in amongst Chinese speech in just the same kind of way that 'taco' is used freely in English. --Geographyinitiative (talk) 04:54, 1 February 2019 (UTC)
 * Turns out some scholars disagree with the inclusion of such entries in Xiandai Hanyu Cidian. See also news report here. "《人民日报》高级记者傅振国说：“《现汉》第6版在‘正文’中收录了英语缩略词等词汇之后，等于将汉语汉字的标准规范擅自改变为英语等外语可以进入汉语，英文可以代替汉字. ”" KevinUp (talk) 12:38, 1 February 2019 (UTC)
 * It would be serious enough if there were “code-switching quotes”, foreign language quotes in English sections displayed as “code-switching ▼” instead of “quotations ▼”; in the citation namespace perhaps we change the  template so the wording is not “English citations” but “Citations for English”, and perhaps with an extra parameter for subsections like “Citations for English [x] in [language y]”. The findability is good. One cannot expect anyway that a term in a text one searches is found on Wiktionary in the language of the text. If I have a word in Slovenian a section in Serbo-Croatian will usually help, and Persians make use of the Arabic sections for Persian texts. If something is in Latin script in a Chinese text, one expects people to search it as English more than as Chinese. So you editors just need to get over the novelty of this view.
 * It also works trans-script btw – instead of ridiculous Sanskrit-as-English entries we can put quotes for the terms used in esoteric English texts on the citation pages of the Sanskrit entries. This is like Serbo-Croatian entries in Cyrillic can contain quotes in Latin script since one intends to have mirrored Latin and Cyrillic entries, and like one does not always quote every alternative form on its own page when it would be more useful to centralize to showcase the meaning for example and when variants depend on readings of manuscripts ( actually contains a quote for, for the readings are all uncertain …), and Azerbaijani is now using  like on , and if you quote from audio-records you can’t quote spellings anyway – but I might be too liberal here and you don’t make this second step though assenting to this code-switching quoting. Point is editors need to open their minds for the first step. Sweeping normal quotes into the appendix is caitiff. Fay Freak (talk) 13:35, 1 February 2019 (UTC)
 * I think this would be a better way of dealing with the current situation (to have foreign language quotes in English sections displayed as “code-switching ▼” and “Citations for English [x] in [language y]”). This is usually encountered for proper nouns (personal names, placenames, etc). The Vietnamese Wikipedia often uses the original Latin spelling without converting it to the Vietnamese alphabet. KevinUp (talk) 21:05, 1 February 2019 (UTC)

What about creating a dummy "language"/language code/Language header for cross-linguistic terms- that is, terms that are used in a given language, but don't really belong to that language. We already have "und", which displays as "🇨🇬". We even have entries: see Category:Undetermined language. We would have to set some ground rules so that we wouldn't be basically duplicating our coverage of a term for every language that might use it in running text, and we would still have to weed out translingual terms and genuine borrowings. Figuring out what to do about script support might be tricky, though. Chuck Entz (talk) 05:07, 1 February 2019 (UTC)
 * The scope of this is a bit too wide. We're currently looking at Chinese loanwords that retained part of its foreign script and whether or not such entries can be considered as Chinese lemmas. KevinUp (talk) 12:38, 1 February 2019 (UTC)
 * If there are two languages to ascribe a word to the language of it is no more “undetermined” than the etymology of a word is “unknown” if we have two etymologies we are not sure to choose between. It’s not undetermined, it is underdetermined, a thing that usually isn’t a problem in language.  is an example of what a word of undetermined language is. If a word is sorted as of undetermined language or an etymology as unknown, there is hope that at some point the language is determined respectively the etymology is resolved (we even categorize pages with und language links). For the state of things was known somewhen to someone. In the code-switching examples it is no issue to leave it unresolved. They have arisen in a state of ambiguity. Fay Freak (talk) 13:47, 1 February 2019 (UTC)

Another solution
As suggested by User:Fay Freak, I think the inclusion of foreign language quotes in English sections which are displayed as “code-switching ▼” instead of “quotations ▼” as well as “Citations for English [x] in [language y]” in the Citations namespace would be a better solution. , any further comments? KevinUp (talk) 03:42, 2 February 2019 (UTC)
 * I totally oppose including a quotation in which someone uses one English word in what is otherwise a Chinese sentence under an ==English== section, no matter the formatting. If the sense used in the quotation exists in English, use English quotations to cite it; if the sense doesn't exist in English, then it shouldn't be in an ==English== section and it also strongly suggests the string does deserve a ==Chinese== (or whatever) section. (I would make exceptions for extinct languages attested in Greek manuscripts and things of that sort. But using code-switching to attest, or provide as quotations to illustrate the use of, a WDL? No.) - -sche (discuss) 23:24, 2 February 2019 (UTC)
 * From a wholesome view, this is language and thus to be covered. It is an extra one can have. I do not deem it likely that Wiktionary can overflow with code-switching quotes in any fashion that could significantly give offence. Note that one has the offence already. One has the quotes already but sorted in a crude fashion. This is only about showing the quotes as what they are: Multilingual text. Fay Freak (talk) 22:56, 3 February 2019 (UTC)
 * Yeah, some users might oppose to having foreign language quotations, because it feels weird to see Greek or Chinese text popping up within an English entry. Perhaps a new template similar to in the English section for code-switched/foreign language quotes would be more appropriate. KevinUp (talk) 04:35, 4 February 2019 (UTC)
 * Why add code-switching quotations at all? Our mission is to define all words, not to record all quotations. We can cite and define English words using English quotations. AFAICT the only reason to bring up a Chinese- or German- or whatever- language quotation in which one "English" word has been embeded, is if the "English" word (or sense) can't be attested via English quotations . . . in which case, shoehorning it into an ==English== section in any fashion is wrong, on a basic WT:CFI level, and furthermore strongly suggests the word is in fact a word in the language of the surrounding quotation. - -sche (discuss) 05:26, 4 February 2019 (UTC)
 * Yes, it's not our mission to record such quotations. If we were to create a section for "code-switched quotations", this will have to be restricted to lemmas written in a nonnative script. The main reason for having this section is to prevent entries such as iPhone or iPhone from appearing. KevinUp (talk)
 * Why/how would preventing/banning iPhone require adding code-switching quotations? Anyone who runs across a code-switching quotation and wants to know what any of the words in it means can look up each of them, and find Chinese entries (with Chinese quotations) defining the Chinese words and iPhone (with English quotations) defining the English word, enlightening them on the meaning of all of the words in the quote. I don't see why the code-switching quote itself would need to be recorded. - -sche (discuss) 06:29, 4 February 2019 (UTC)
 * Good point. The thing is, there's currently a loophole in our system. It is not impossible to find quotations for APP, の or iPhone (no quotations yet), but are these words considered actual lemmas in their respective languages? What can we do to prevent users from creating such entries? Guidelines are needed to identify whether quotations provided to lemmas written in a nonnative or nonstandard script qualify as code-switching. KevinUp (talk) 07:41, 4 February 2019 (UTC)


 * Consider also Talk:hiam. I also used and  as the basis for creating, but really I could not find usage of the same phrase in Chinese characters, excluding  calques. —Suzukaze-c◇◇ 06:11, 4 February 2019 (UTC)
 * Interesting. Now we're looking at Min Nan terms code-switched into English. I disagree with the creation of hiam. If it's Singlish or Singaporean English, it should at least be found here: http://eresources.nlb.gov.sg/newspapers/ (Singapore newspaper archive). I think it's Min Nan code-switched into English, because non Min Nan speakers might not be able to catch its meaning (Singapore is a fairly diverse society).
 * As for, I think we can create POJ entries such as ài-súi-mài-miā rather than poorly transcribed "ai swee mai mia" or "ai sui mai mia". Hokkien is often transcribed without tone marks by the locals, but when read its still pronounced exactly like Hokkien. KevinUp (talk) 07:41, 4 February 2019 (UTC)

Citations namespace
This seems perfect for the citations namespace instead of an appendix. It's linked automatically from the entry, and there's no need to specifically label it as a particular language. DTLHS (talk) 23:31, 2 February 2019 (UTC)
 * The important lexicographic information concerning the Chinese entry, like pronunciation or the measure word it takes, cannot be placed on the citations page, so there would be a loss of information. Also, citations pages are standardly labelled by language using . —Μετάknowledge discuss/deeds 19:04, 3 February 2019 (UTC)
 * Yeah but foreign terms which haven’t passed are arbitrarily adapted to the sound system of a language. This has also been shown for “APP” in Chinese, and is well-known of code-switching in general: As is that when multilinguals switch languages it is not unusual to take over the pronunciation of one language when one is in the other, and even conscious speech assumed there lack standards. Even the pronunciation of “passed” words is rather arbitrary, dependent on educational background and also intentionally ridiculized.
 * I have suggested the citation namespace, but with proper earmarking of multilingual quotes. Fay Freak (talk) 22:44, 3 February 2019 (UTC)
 * the use of the citations namespace for entries such as APP which has the exact same definition as its corresponding English.entry . As for its pronunciation, there isn't a proper guideline for that. Xiandai Hanyu Cidian states the following in its appendix for lemmas that begin with the Latin script: I hope we can make a decision on this soon, i.e. (1) entries in Category:Chinese terms written in foreign scripts such as size, part-time, iPhone, which has the same meaning as in English, are to be moved to the citations namespace. (2) Only entries such as man, fighting, NG, which has a meaning different from what one would expect from its usual English definition, are to be included in the Chinese section. KevinUp (talk) 04:25, 4 February 2019 (UTC)
 * The loss of pronunciation information in entries such as size (Cantonese: saai1 si2) is regrettable, but that information belongs to /, not size. Until a lemma has been properly lemmatized into Han script (e.g. → ), its pronunciation is often unclear and varies depending on each individual.
 * Then again, I have no idea why APP is pronounced like an initialism in mainland China, but I think this information can be included in the usage notes of APP instead. KevinUp (talk) 04:25, 4 February 2019 (UTC)

A specific low memory template for compounds of Japanese kanji, Korean hanja, Vietnamese Han characters
The page for is currently out of Lua memory. Even after memory consuming templates such as were removed, the same problem persisted. This may have something to do with Module:columns. I think, we may need to rely on an older version, such as Module:columns/old. I found that uses less memory compared to.

Also, a few months back, a user was confused by the many derived terms in the Japanese section (Tea_room/2018/August), so whatever template used for compounds or derived terms needs to have a customizable title ( doesn't have a title anymore). KevinUp (talk) 02:48, 1 February 2019 (UTC)

I just made a der3 with a custom title right here. Or was there a discussion about deprecating it ASAP? mellohi! (僕の乖離) 19:00, 1 February 2019 (UTC)


 * Yes, the customized title of has been deprecated. Prior discussion can be found at Beer parlour/2018/November. Take a look at wine. The unboxed title looks out of place. KevinUp (talk) 21:06, 1 February 2019 (UTC)
 * Ah, you weren't referring to the unboxed titles. Speaks of me being out of the loop. mellohi! (僕の乖離) 21:09, 1 February 2019 (UTC)
 * No worries. Basically, the Lua memory used for (Chinese compounds) and  (JKV compounds) needs to be reduced. KevinUp (talk) 03:42, 2 February 2019 (UTC)

Update: seems to have enough memory now. , do you know which template/module was using up the memory? I just realized does not use Lua memory. KevinUp (talk) 04:49, 4 February 2019 (UTC)


 * Update 2: By using instead of  and subsituting  with  and, Lua memory in  is now reduced to 44.59 MB. KevinUp (talk) 09:24, 11 March 2019 (UTC)

Kanji compounds for Japanese given names
Previous discussion: Wiktionary:Tea room/2018/August, User talk:Shāntián Tàiláng

I'm interested to know what the community thinks about creation of kanji compounds such as that are only used in given names. There are up to up 148 possible kanji combinations listed at. Are we going to create entries for all of these? Readings for Japanese given names (known as ) is often arbitrary and there are no strict rules on which kanji to use.

When I look at pages such as Category:Japanese terms spelled with 実 read as み, most of it appears to be given names. To isolate actual kanji compounds, one would have to search for — incategory:"Japanese terms spelled with 実 read as み" -incategory:"Japanese proper nouns" intitle:実 to obtain the 14 entries of 実 with reading み that are not proper nouns.

We could perhaps include only the top 5000 kanji compounds used for given names. I think listing the possible kanji forms for Japanese given names at hiragana pages such as is good enough. To find out how to pronounce a person's name written in kanji, we could just use the search box, or check the nanori readings listed at individual kanji pages.

On an unrelated note, most South Koreans still use hanja for their given/personal names, but we don't have any entries for hanja given names. There are only 3 entries in Category:Korean given names, compared to the 6229 entries we have in Category:Japanese given names, while Category:Chinese given names was recently deleted, because most Chinese given names are sum of parts, and any combination is possible as long as it's not a lewd word.

So the question is, should we continue to create such entries, or should we limit this to something like 5000 most popular kanji compounds (I have no idea where to find this). KevinUp (talk) 03:42, 2 February 2019 (UTC)
 * For me, all Japanese given names should be lemmatized at the hiragana form, with the kanji spellings being soft redirects to the hiragana lemmas, where an exhaustive list of kanji spellings can be added. Subsequently, Category:Japanese given names and all its subcategories should be purged of any kanji spellings of given names, leaving only the hiragana lemmas left. mellohi! (僕の乖離) 04:45, 2 February 2019 (UTC)

Pinging over here for their thoughts. mellohi! (僕の乖離) 04:45, 2 February 2019 (UTC)


 * At least, It might be necessary to isolate articles of people names that can be made unlimitedly. It’s necessary to delete if there are which does not have actual usage.--荒巻モロゾフ (talk) 06:30, 2 February 2019 (UTC)


 * My only opinion is that we should not rely on the EDICT names dictionary, and should do at least the minimum effort to make sure that a name or its kanji spelling is actually used. —Suzukaze-c◇◇ 23:44, 4 February 2019 (UTC)
 * I’d like rename all Japanese first name entries to hiragana. One can very freely choose kanji for a given pronunciation. — T AKASUGI Shinji (talk) 00:44, 5 February 2019 (UTC)
 * In response to the original question, speaking as a beginner reader of Japanese, I would like as many kanji personal "first" names and family names as possible to be look-uppable. Sometimes for beginners it may not even be clear that a kanji compound is a personal name. Generally, if I see incongruous characters, e.g. for topographical features or "beautiful flower"-type meanings, then I tend to guess that a personal name is meant, but sometimes it is not obvious for beginners. Mihia (talk) 01:18, 15 February 2019 (UTC)


 * In Japanese texts, especially those that are beginner friendly, given names are often written with the suffix . In addition, Japanese personal names (full names) often consist of four to five character kanji compounds (usually two characters for surname, followed by two or three characters for given name), so it is not that hard to identify a Japanese personal name while reading more advanced texts. I don't think it is practical to have as many kanji personal "first" names because many different kanji variations are possible for the same Japanese given name written in hiragana. Family names, on the other hand, are fixed when it comes to kanji choice, due to strict rules in the system, so kanji compounds for family names can be included as they fulfil our attestation requirements. KevinUp (talk) 03:05, 17 February 2019 (UTC)
 * Erm, well, thanks, I know さん, and that names are character compounds! The thing is that these characters usually have literal meanings too. When these are obviously incongruous to the subject matter it is not too bad, but this isn't always the case. Also there are no capital letters to help, of course. Mihia (talk) 00:04, 20 February 2019 (UTC)

Taxonomic names in individual languages
In Dutch, but I'm sure also in other languages, there are terms for taxonomic clades as well as members of them, that are different from the scientific/Latin-based translingual names. For example, the normal term for Felidae is and for Mustelidae it's. These are plural forms of nouns, and the singulars and  refer to individuals of these groups.

There doesn't seem to be any kind of category tree for such names currently. We have a big set of categories for the translingual taxonomic names, but they don't seem to have equivalents in other languages, only translingual. Given that both the group and its members are part of a single lemma in Dutch, how should these be categorised? Only refers to a group, but it's not a lemma, so it shouldn't have any categories. Should the lemma have something like  as a second definition? —Rua (mew) 13:27, 2 February 2019 (UTC)


 * In English, too, a is a member of the, and the whole group could be referred to by the plural of that word. But if you say (and I tend to agree) the plural isn't a lemma—if its use to refer to the Mustelidae isn't so lexical it needs to be given as a definition on the page  / —why isn't it sufficient to define the singular as "a member of the family "? Is it not comparable to how "humans" in the plural can mean "humanity" / "humankind", but we probably don't need to add a sense to "human" (or "humans") that says "(in the plural) Humanity / humankind", or a sense at "elf" for "(in the plural) elfkind", "dog" "(in the plural) dogkind", etc? As far as categorization, what would you suggest would be needed beyond putting  in Category:nl:Mustelids the way  is in Category:en:Mustelids? - -sche (discuss) 17:04, 2 February 2019 (UTC)
 * I don't really think should be in Category:en:Mustelids, based on WT:Beer parlour/2018/December. But that aside, we have a lot of categories specific to taxonomic names, but only in Translingual, not in any specific language. My question was more related to whether we should replicate this structure in all languages that have terms referring to species/taxonomic groupings (like English and Dutch, as you showed). That is, should  be in a to-be-created Category:en:Taxonomic names or Category:en:Taxonomic names (family), the way that  already is? —Rua (mew) 18:23, 2 February 2019 (UTC)
 * Aha, I see you what you mean. Hmm...if marterachtige(n) / mustelid(s) is categorized as Category:foo:Taxonomic names (family), would and  be categorized as Category:foo:Taxonomic names (species)? Would birds / vogels be categorized as a taxonomic name for a class? And then, would cohosh also be in Category:en:Taxonomic names (species) although it refers to two species? I guess I'm not opposed to that, though the birds/vogels (clearly just a common name/word) and cohosh (ambiguous / two species) examples seem like evidence these aren't truly taxonomic (unambiguous) names. (It seems related to the question of whether mul taxonomic names can have translations, to which the de jure answer may be no but the de facto answer—looking at Navajo, for example—is yes. On that note, I suppose marterachtigen and mustelids should be added to Mustelidae...) - -sche (discuss) 23:17, 2 February 2019 (UTC)
 * Perhaps languages written in non-Latin scripts can give answers. How are taxonomic names rendered in Russian or Chinese? I cannot read Chinese, but 鼬科 is the interwiki for Mustelidae and it has a name in Chinese characters, with the given after it in Latin letters. Куньи is likewise in Russian, and gives the scientific name but labels it "Latin". Would "scientific name" and "taxonomic name" be the same thing? What is the term for native-language equivalents of taxonomic names, like  and ? Should we give them their own categories or just place them in the regular lifeform set categories? —Rua (mew) 21:07, 3 February 2019 (UTC)
 * “Would ‘scientific name’ and ‘taxonomic name’ be the same thing?” In the context of discussing a taxon: yes. See and .  --Lambiam 02:11, 4 February 2019 (UTC)

Format of custom header text in new

 * Can you please change the formatting of the custom header text in the new ? It has the same bold text as Derived terms and this does not make it obvious to readers that the multiple tables under Derived terms still belong to this section. They seem like separate sections. I tried to get used to it but every time I see it, I find it confusing. See . I'd prefer text in italics and parentheses, with closing colon, e.g. (Compound words): Thanks. Panda10 (talk) 20:07, 3 February 2019 (UTC)
 * A change like that needs consensus, though admittedly I didn't get input on what the header text in and similar templates should look like when I chose the style. But you can change it to the style you propose just by adding the following to your common.css:
 * — Eru·tuon 20:40, 3 February 2019 (UTC)
 * I really appreciate the script but I'm not sure if modifying my common.css is the correct solution. I think it is better to see the entries as a Wiktionary reader would see it. Panda10 (talk) 21:46, 3 February 2019 (UTC)
 * Well, okay. I agree that the current style is confusing. I don't like the combination of parentheses and colon myself, but if others like it, I can implement it. In the meantime I should probably make the header not use inline CSS though. — Eru·tuon 22:01, 3 February 2019 (UTC)
 * Why not just hard-format it in a more satisfactory way. Let a thousand flowers bloom and then pick from among them. DCDuring (talk) 00:22, 4 February 2019 (UTC)
 * Well anyway, DTLHS added the CSS . — Eru·tuon 23:38, 9 February 2019 (UTC)


 * Because of the extremely poor presentation of the modified templates (der3, der4 etc.) I have stopped using them completely. I now use der-top3 for new conversions; it's not perfect but the presentation is far better. Progress, huh? DonnanZ (talk) 17:28, 10 February 2019 (UTC)
 * I agree. I'm also using for Han character entries. At least it doesn't use any Lua memory. KevinUp (talk) 08:42, 11 February 2019 (UTC)
 * I'm kind of disappointed, but can't blame you. The new layout doesn't look very good, and it's annoying to have the toggle button at the bottom, because you can't collapse the list when you're reading through a page. I am open to ideas for improvement. I am not great at graphic design or whatever this is. It would be nice to at least bring it to the level where nobody hates it so much that they can't bear to use it. — Eru·tuon 22:56, 11 February 2019 (UTC)

Usage of kanji in Ryukyuan languages besides Okinawan
Unfortunately, is once again out of Lua memory, even though it was working yesterday. I would like to know whether the following languages: (1) Miyako, (2) Northern Amami-Oshima, (3) Oki-No-Erabu, (4) Southern Amami-Oshima, (5) Yonaguni, (6) Yoron, are actually written using kanji (historical or modern times) by native speakers.

The sections at appears to have been added by the following two users: Special:Diff/25636005/25750073. Should these languages be lemmatized using kana instead of kanji? KevinUp (talk) 11:46, 5 February 2019 (UTC)


 * (are they written by native speakers at all, in the first place? 🤔 —Suzukaze-c◇◇ 07:03, 6 February 2019 (UTC))
 * I don't think so. The entry for 海豚 even has (7) Kikai and (8) Kunigami, added by User:Nibiko in this 2016 edit. Some of these languages appear to have test wikis at Wikimedia Incubator, but I'm not sure about the script used. KevinUp (talk) 10:24, 6 February 2019 (UTC)


 * They use kanji in the purposes to write the lyrics of their traditional songs (examples: ). Note that those spellings are not necessarily phonologically strict, and not linked to the spellings for convenience which prepared by researchers. Modern Ryukyuan languages don't have any official orthographies defined.--荒巻モロゾフ (talk) 14:54, 11 February 2019 (UTC)


 * Since these appear to be used in modern times, I think we can use or  at the definition lines of kanji entries for Ryukyuan languages that are not Okinawan. Meanwhile, kanji forms copied from online dictionaries that lack attestation will be removed. KevinUp (talk) 07:24, 23 February 2019 (UTC)

2018 ISO code changes
The [//iso639-3.sil.org/sites/iso639-3/files/reports/639-3_ChangeRequests_2018_Summary.pdf changes the ISO made to codes in 2018] were posted. They: They also added codes: xsj Subi (a lect previously merged with Shubi; we merged Shubi into Rwanda-Rundi, but Subi is said to not be closely related and only often associated by confusion) (done), lvi Lavi (which we current encode as mkh-law)  (done) , lsv Sivia Sign Language, cey Ekai Chin (WP prefers just "Ekai"); the Australian languages wkr Keerray-Woorroong, tjj Tjungundji, and tjp Tjupany, about which see WP; pnd Mpinda, lsn Tibetan Sign Language, and tvx'' Taivoan (Taivuan). If anyone has a reason we should not follow suit on these code deprecations and creations, please speak up. (They also made a number of name changes we could look into.) - -sche (discuss) 06:43, 6 February 2019 (UTC)
 * [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-015.pdf split] and retired ais Nataoran Amis, merging the Amis part into ami "Amis" and creating a new code szy for [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-015_szy.pdf Sakizaya] ([//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/CR_Comments_2018-015.pdf commentary]).
 * ✅. - -sche (discuss) 16:13, 16 March 2019 (UTC)
 * merged asd Asas into snz Sinsauru (Sensauru), and renamed it Kou (alternative spelling: Kow), on [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-012.pdf solid] grounds.
 * ✅. - -sche (discuss) 18:25, 9 February 2019 (UTC)
 * [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-014.pdf split] dud Hun-Saare into [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-014_uth.pdf uth] ut-Hun and [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-014_uss.pdf uss] us-Saare.
 * ✅ - -sche (discuss) 23:46, 26 March 2019 (UTC)
 * retired lba Lui as spurious, [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-016.pdf citing] Wikipedia, which cites that ISO document. But ISO cites other sources too, so it's not just citogenesis.
 * ✅ - -sche (discuss) 07:46, 23 February 2019 (UTC)
 * merged llo Klor / Khlor into ngt Kriang, [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-008.pdf saying]: "In January, 2018 I happened to be sitting next to a man from Sekong province. I asked him about Klor and to my shock and his, he reported that he himself was Klor. He confirmed that it is pronounced [klɔːr] with no aspiration and that the langauge is spoken only in Ko' [kɔʔ] village. He reported that Klor is completely mutually intellibile [sic] with Kriang and that he considers the Klor to be Kriang. We counted to ten together and it was indeed the same as Kriang. This leads me to propose that Khlor [llo] be retired and Klor (note spelling difference) be added as a dialect of Kriang [ngt]."
 * [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-010.pdf merged] myd Maramba into aog Angoram.
 * ✅ - -sche (discuss) 07:44, 23 February 2019 (UTC)
 * retired myi, which we already retired.
 * [//iso639-3.sil.org/sites/iso639-3/files/change_requests/2018/2018-090.pdf merged] nns Ningye into nbr, which they renamed Numana (from Numana-Nunku-Gbantu-Numbu).
 * ✅ - -sche (discuss) 00:08, 24 February 2019 (UTC)
 * Thanks, . -- 07:03, 6 February 2019 (UTC)


 * Neither here nor there, but their addition of xsj Subi is a reversal of their previous merger of it into suj Shubi in 2014 (which we followed at that time). - -sche (discuss) 19:18, 9 May 2019 (UTC)

Tocharian B
The entries in Category:Tocharian B lemmas are all written in the Latin script. Is this correct? SemperBlotto (talk) 07:31, 8 February 2019 (UTC)


 * They were written in the Tocharian alphabet (also see https://www.unicode.org/L2/L2015/15236-tocharian.pdf]]) and in the Manichaean alphabet. —Stephen (Talk) 08:45, 8 February 2019 (UTC)


 * We cannot write them differently until Unicode encodes the Tocharian script, we have a similar situation with Sogdian and Old Uyghur. Crom daba (talk) 20:07, 8 February 2019 (UTC)
 * Sogdian was already added to Unicode 11.0, as was Manichaean (back in 7.0), so technically you could be creating Tocharian B entries in when attested. But yes, alas, Tocharian has yet to be encoded. --  21:08, 9 February 2019 (UTC)

Use of the term "West Frisian"
On Wiktionary, the Frisian language as spoken in the Netherlands is always referred to "West Frisian". This is its usual name outside the Netherlands, contrasting with East Frisian and North Frisian. However, in Dutch "West-Fries" usually refers to a dialect spoken in the province of North Holland. This variety is Dutch, not Frisian, but is called "West-Fries" because it is spoken in the historical region of West-Friesland. So far, so good. We all know what is meant by it, and we usually don't add word from Dutch dialects. However, there is an actual variety of Frisian that is extinct today but was spoken in (pockets of) West-Friesland until about 1700. Not much has survived of this language, but it would love to add those words that have. But to do so, we would have to settle on names. I can't call these entries "West Frisian", since that name is already in use for the living language that us Dutchpeople call "Westerlauwers Frisian". My proposal would be to adopt Dutch terminology: rename all existing West Frisian lemmas to "Westerlauwers Frisian" and reserve the name "West Frisian" for this language. I admit it would be cumbersome, but at least it would be unambiguous. What do you think? Steinbach (talk) 12:46, 11 February 2019 (UTC)
 * Another option would be use a geographical description like "Noord-Holland" or the historical term "". ←₰-→ Lingo Bingo Dingo (talk)  15:52, 11 February 2019 (UTC)
 * Do linguists treat this West-Frisia Frisian as a separate language from Westerlauwers Frisian? — Ungoliant (falai) 17:12, 11 February 2019 (UTC)
 * That's a difficult question. As you know, linguists tend to stay away from the arbitrary distinction between "language" and "dialect". The two varieties were clearly distinct, however. A 17th century Frisian poem could with certainty be identified as being from North Holland, not Friesland, by its text alone. At least one defining feature that sets Westerlauwers Frisian apart from East Frisian, the words sa and ta rather than so and to, did not occur in West Frisia Frisian. Some innovations relative to Old Frisian are shared, some aren't. In combination with the geographical and political separation, a solid case can be made to treat the two varieties as separate languages. Steinbach (talk) 20:24, 11 February 2019 (UTC)
 * Could you give a pointer to literature about this Frisian lect? The proposal is to move away from the usual terminology in English, so it would be useful to see how others deal with it. ←₰-→ Lingo Bingo Dingo (talk)  08:02, 12 February 2019 (UTC)
 * Give me some time. I'm not in that stage myself, I've been inspired to this proposal by an article in mainstream press. For the time being, here's a link to the sole surviving longer text in this dialect. It can give you an impression of how it differs from Westerlauwers Frisian. Steinbach (talk) 08:19, 12 February 2019 (UTC)
 * What I understand from the article is that the language of this sole surviving text of 331 words (160 different words) was known to be a quaint variant of Westlauwers Frisian, and has about a year ago been identified as being specifically a North-Holland variant (not quite surprising, seeing as it is one of the song texts in a collection titled d'Amsteldamsche Minne-zuchjens). Interesting, but hardly a reason to upset the Frisian language classification. And redefining “West Frisian” to mean neither nor the  as the term is commonly understood by linguists, but to reserve it for this variant, will be utterly confusing. Just like  has a label , we can use some label like   for words found only in this variant.  --Lambiam 13:25, 12 February 2019 (UTC)
 * This article (in an issue of De Vrije Fries from 1906) discusses possible printing errors in the text – apparently not considering the possibility that the language may be a variant of West Frisian. (BTW, pejeer may be an attempt to render .) --Lambiam 13:42, 12 February 2019 (UTC)
 * I agree, setting apart a new language code goes too far for this. Any words can be included as obsolete West Frisian. ←₰-→ Lingo Bingo Dingo (talk)  08:09, 13 February 2019 (UTC)
 * It is definitely more than obsolete West Frisian. It differed greatly from seventeenth-century Westerlauwers Frisian, too. The work of Gysbert Japiks already looks rather similar to modern day Frisian, something that can't be said of this text. Steinbach (talk) 09:00, 13 February 2019 (UTC)
 * I wouldn't claim it was merely obsolete Westlauwers, just that it should be included under West Frisian and labelled as obsolete, in addition to a geographical tag. So something like fy should in my view do the trick. I am also curious about the extent that the similarity of Japicx's Middle Frisian to modern-day West Frisian is due to his orthography influencing later orthography. ←₰-→ Lingo Bingo Dingo (talk)  13:44, 13 February 2019 (UTC)
 * Technical discussions aside, that's a hilarious poem. Soo molle bolle Femke! — Mnemosientje (t · c) 14:59, 12 February 2019 (UTC)
 * I suppose that is one way of dating it to the seventeenth century, beside the title page and the spelling. Maybe nice for use on Valentine's Day? ←₰-→ Lingo Bingo Dingo (talk)  08:09, 13 February 2019 (UTC)

English-based creoles of Suriname
The Surinamese creole languages Sranan Tongo, Aukan and Saramaccan currently do not have any ancestors recognised by Wiktionary's classification. For Sranan and Aukan it is uncontroversial that these are English-based creoles (some consider Saramaccan a Portuguese-based creole instead); they in many ways resemble Guayanese Creole (which also has no ancestor languages in the categorisation) and Jamaican Creole (which is recognised as a descendant of English). Several scholars posit also posit a common creole ancestor to those variety. Implementing that latter view might go too far now, but it seems a good idea to at least enable Sranan Tongo and Aukan to have terms as inherited from English. ←₰-→ Lingo Bingo Dingo (talk)  15:47, 11 February 2019 (UTC)
 * (Note the prior discussion at Talk:dofu.)
 * Adding English as an ancestor of Sranan at least seems sensible per our earlier discussion, the other languages I can't well judge. But if Aukan is so obviously English-based, then I personally don't see the harm. — Mnemosientje (t · c) 15:55, 11 February 2019 (UTC)


 * My feeling is that it is misleading to say that a word like was inherited from English. It suggests that some branch of the English language tree evolved in some way so as to morph into Sranan. But these words were incorporated into the creole language as it was crystallizing out of a pidgin that was not a language in the usual sense of that term but an unstable mishmosh varying from plantation to plantation. For English to be an ancestor, there should be intermediate versions of the language that are closer to English than modern Sranan is, while also closer to modern Sranan than English is.  --Lambiam 21:46, 11 February 2019 (UTC)
 * I too disagree that we should be marking lexifiers as ancestors of creoles; is the best template to use. See this discussion, among others. —Μετάknowledge discuss/deeds 21:59, 11 February 2019 (UTC)

Another matter is whether the Surinamese creoles should be linked in a similar way. Aukan is generally considered to descend from (very) Early Sranan or Proto-Sranan, and the same is often considered for Saramaccan. ←₰-→ Lingo Bingo Dingo (talk)  15:50, 11 February 2019 (UTC)


 * The book Pidgins and Creoles: An Introduction contains a chapter on Sranan in which the authors write: “As far as the shared histories of [the Atlantic group of English-based creole languages] are concerned, we may point to such aspects as the common supplier of the vast majority of the imported slaves — the Dutch, and the history of colonization, whereby a new colony was founded by groups from one or more existing colonies. Surinam, for instance, was first settled from Barbados, St. Kitts, Nevis and Montserrat. In this way [Sranan] is linked to the other Caribbean English-based creoles. [...] Within this group Sranan belongs to a clearly defined Surinam subgroup. This subgroup can be demonstrated in historical linguistic terms (with languages Sranan, Ndjuka-Aluku-Paramaccan-Kwinti, Saramaccan-Matawai). Outside this subgroup Sranan has a particular relationship with Krio, and other similar languages on the West African coast, as well as with the Maroon Spirit Language of Jamaica (Bilby 1983).” (Ndjuka is another name for Aukan.) Unfortunately, Google does not allow me to view most of the section entitled “History and current status”. --Lambiam 12:43, 12 February 2019 (UTC)
 * You can find it on Library Genesis. Not sure if we're allowed to link it here. Some other comments of interest:

"So we cannot say that Sranan (the major English-lexifier creole of Surinam; see chapter 18) derives in any gradual fashion from Early Modern English – its most obvious immediate historical precursor. [...] we are dealing with two completely different forms of speech. There is no conceivable way that Early Modern English could have developed into the very different Sranan in the available 70 or so years. [...] So creole languages are different from ordinary languages in that we can say that they came into existence at some point in time. [...] we have to reckon with a break in the natural development of the language [...] The parents of the first speakers of Sranan were not English speakers at all, but speakers of various African languages, and what is more important, they did not grow up in an environment where English was the norm."
 * From the section you couldn't see on the Google preview, here are some comments of interest:

The origins of Sranan (see also chapters 2 and 10) must be sought in the seventeenth century. Surinam started its post-Amerindian history as an English colony in 1651. The period of English occupation only lasted officially until 1667. English influence can be considered to have become negligible by 1680. So the period in which the direct linguistic influence of English can be assumed to have been operative was less than thirty years. [...] How precisely English functioned in the development of Sranan is highly controversial. In for instance the bioprogram hypothesis of Bickerton (see chapter 11), English lexical items and language universals combined to produce Sranan. In the substrate approach the African language(s) of the early slaves had a decisive influence (chapter 9).
 * From this, I have to agree with Metaknowledge that perhaps a simple may be best, at least in the case of Sranan. — Mnemosientje (t · c) 14:50, 12 February 2019 (UTC)
 * Thanks. I didn’t know about LG. Not having access to a research library, it looks like a useful addition to my research tools. --Lambiam 00:00, 13 February 2019 (UTC)
 * Yes, in that case der also looks like the best choice for Aukan terms deriving from seventeenth century English. ←₰-→ Lingo Bingo Dingo (talk)  07:57, 13 February 2019 (UTC)
 * Short note to participants in this discussion ( - I have created WT:About Sranan Tongo as a draft (including a note about the use of etymology templates), perhaps some of you might be interested to improve it. I think it might be useful to add a note about Sranan orthography. I have noticed variant spellings being a thing but don't know enough to add a useful description of the issue on the think tank page. — Mnemosientje (t · c) 10:15, 24 February 2019 (UTC)
 * Do you all agree with the use of bor or borrowing for terms that were taken from Dutch without an intermediary? ←₰-→ Lingo Bingo Dingo (talk)  14:24, 5 March 2019 (UTC)
 * Fine with me. But what do we do with terms taken from Portuguese like and ? I think these, as well as terms originating from African languages, like  and, should use der just like terms in early Sranan coming from the English lexicon.  --Lambiam 20:30, 5 March 2019 (UTC)
 * I'd be all right with that change. In many cases the exact trajectory for loans from Portuguese seems hard to establish, apparently some scholars suppose that there was a Portuguese-based pidgin or creole involved at some point. As for African languages, I'm not entirely sure. In some (but perhaps not all) cases the terms must have been directly borrowed into the creoles via native speakers, right? ←₰-→ Lingo Bingo Dingo (talk)  13:21, 6 March 2019 (UTC)
 * @LBD – There is a striking commonality of words found in Atlantic creoles that originate from Portuguese. It is generally acknowledged that a possible and even plausible explanation is that they had been swirling around in the mishmosh of Atlantic pidgins developed along the Atlantic coast of Africa already generations before the slave trade, and that they travelled with the slave ships to the Americas. For all we know the same may hold for words in Atlantic creoles originating from African languages, which also display such commonality, like ~  and  ~ . It is furthermore generally assumed that a new creole language is forged by the first generation born into a society that has no shared language and communicates through a pidgin. It is therefore (in my opinion) entirely possible that the words stemming from African languages were copied from the pidgin of the older generations, but not necessarily via a native speaker of the original language. If we were to confer language status to the Atlantic pidgin mishmosh (something we definitely shouldn’t do) we could use    and    in describing a two-stage scenario: term in [creole language] inherited from Atlantic mishmosh, borrowed into Atlantic mishmosh from [African language].  --Lambiam 19:21, 8 March 2019 (UTC)

Blocker role
What are peoples thoughts on creating a blocker role so that non-sysops can issue short-term blocks to be reviewed later by an admin? -- 21:59, 12 February 2019 (UTC)
 * What would be the point of "reviewing" them if they were short? Most blocks are short anyway. What action would a "reviewer" take? What happens if they aren't reviewed? Why would an admin want to review blocks that they could have done themselves? . DTLHS (talk) 22:30, 12 February 2019 (UTC)
 * The point would be to have more users to catch vandals in the act. If we don't need more people to do that, why are we even having admin votes for that role? You could think of them more like blocker-bots, and the second they're doing a poor job of it, you decommission them and take away the role. -- 02:16, 13 February 2019 (UTC)
 * I have no problem with giving more people vandalism fighting abilities. My main issue is with the "reviewing" that probably wouldn't happen. DTLHS (talk) 02:44, 13 February 2019 (UTC)
 * That's fair. At the very least, the blocker role users could issue their block and request an perma block when needed. -- 02:55, 13 February 2019 (UTC)
 * Also oppose, if there was actually a time that no admin was active there are emergency options (stewards). In my mind this is the functions of admins which is most "powerful", so anyone I would want having this ability I would be happy to have as an admin. - TheDaveRoss  00:12, 13 February 2019 (UTC)
 * You yourself were questing the quality of admins be have these days. In this way, we can have people stopping vandals in their tracks, while still holding admins to a higher standard. -- 02:16, 13 February 2019 (UTC)
 * I am saying I would hold the blockers to the same standard that I hold admins. - TheDaveRoss  02:53, 13 February 2019 (UTC)
 * Which, IMO, ensures sub-quality admins and not enough vandal blockers. -- 03:02, 13 February 2019 (UTC)
 * Do you feel that vandals frequently go unblocked for long periods of time? In my experience blocks happen within minutes if not seconds of vandalism taking place. If there are times that vandals are able to persist for longer periods I would be interested to hear about that, since it would be happening while I am unaware. My opinion remains unchanged about the need for a distinct role, I would not vote to approve anyone as a blocker if I would not also vote to approve them as an admin. My bar to become an admin is fairly low, I would guess I have voted yes in well above 90% of admin votes in which I have voted at all. - TheDaveRoss  15:19, 13 February 2019 (UTC)
 * I don't really feel anyway one way about it, but you see the conflict in the statements "judgement [...] has been a problem with existing admins of late" and "my bar to become an admin is fairly low", right, haha? -- 04:48, 14 February 2019 (UTC)
 * Just because the bar is low doesn't mean everyone clears it. I think it is fairly easy to be civil, e.g., which is one of my criteria for voting yes on an admin vote, and yet there are some current admins who place very little value (seemingly) on civility. Most admins (and other editors, and proposed admins) easily demonstrate the level of civility that I hope for in an admin. I think I have a similar viewpoint about judgment and the other criteria which I value, most people easily surpass my expectations, some few fall short. I don't see a conflict with having a low bar and not always determining that everyone clears it. - TheDaveRoss  13:38, 14 February 2019 (UTC)


 * My feeling is that the need for this arises from not having admins in a certain time zone. "No admins are awake, but 4chan is attacking us, and creating a zillion stupid pages! Luckily, a blocker is here!" This raises the issues that (i) it really just means you don't have enough admins, or not a wide enough geographical spread of admins, and (ii) even if you had a special "blocker" role it would be susceptible to the exact same issue that maybe all blockers are asleep too. Equinox ◑ 02:51, 13 February 2019 (UTC)


 * Anyway oppose because it's easy to be whitelisted (by creating 100 entries in some under-loved language) but a lot harder to get admin status, and the ability to stop people from editing is a very significant and powerful one. (Mostly unrelated thought: what if admin responsibilities included dealing with x-percent of untouched anon edits in your language? Sometimes I find stuff I did two months ago not logged in that still hasn't been reviewed.) Equinox ◑ 05:53, 13 February 2019 (UTC)

"Eskimos have 50 words for snow"
https://popula.com/2019/02/11/white-words/ —Justin ( koavf ) ❤T☮C☺M☯ 07:39, 13 February 2019 (UTC)
 * . The article originally appeared in The New Scientist of 18 December 2012 under the title “Are there really 50 Eskimo words for snow?”. Instead of 50 you also find other numbers like 40, 52 and even 100, so “Eskimos have X words for snow” is a snowclone. --Lambiam 09:25, 13 February 2019 (UTC)
 * Eskimos have 50 snowclones. —Justin ( koavf ) ❤T☮C☺M☯ 01:47, 14 February 2019 (UTC)

Ideophones as ur-language
https://aeon.co/essays/in-the-beginning-was-the-word-and-the-word-was-embodied —Justin ( koavf ) ❤T☮C☺M☯ 07:57, 13 February 2019 (UTC)

On crafting scientific language in Zulu
https://www.theopennotebook.com/2019/02/12/decolonizing-science-writing-in-south-africa/ —Justin ( koavf ) ❤T☮C☺M☯ 01:48, 14 February 2019 (UTC)
 * Text's here; both sources aren't durably archived, thus no sources for WT (would be nicer if the article appeared in print). --Brown*Toad (talk) 07:32, 17 February 2019 (UTC)

Layout of "of" qualifiers
I see "of" qualifiers written in two different ways, as these definitions, respectively from wet and fast, illustrate:


 * 1) Of weather or a time period: rainy.
 * 2)  More sensitive to light than average.

Is the second style preferred? Should the first style generally be converted to the second style when encountered? Mihia (talk) 20:36, 14 February 2019 (UTC)
 * I think the second is much more common. It also makes some sense to mark such qualification for searches. with of within the label serves as a fairly natural marker. DCDuring (talk) 21:54, 14 February 2019 (UTC)

Proto-Bantu Verbs
Currently, all Proto-Bantu verb entries have the default suffix -a. However, I think it would be better if this suffix were removed from the lemma forms of PB verbs, as it's not part of the verb root, and not all Bantu languages make use of this suffix. Smashhoof2 (talk) 06:24, 15 February 2019 (UTC)
 * We've taken the route of trying to reconstruct what PB actually looked like (so putting the final vowel on verbs, putting noun class prefixes on nouns), which is contrary to the BLR style, which just shows lexical roots. I don't know what's better, but reconstructing words rather than roots is more in keeping with Wiktionary being a dictionary and attempting to treat languages similarly when possible. —Μετάknowledge discuss/deeds 19:23, 16 February 2019 (UTC)
 * That's fair. Smashhoof (talk) 21:32, 16 February 2019 (UTC)

Stub entries and minimum required content
My talk page contains post to the effect that there exists some additional requirements for minimum content of entries that I am unware of. Such requirements can be created if desired, so let's have an amicable conversation about it.

My understanding of minimum content of an entry is as follows. The entry needs:
 * 1) Language header
 * 2) Part of speech header
 * 3) Somewhat controversially, a definition, translation or, for non-lemma entries, the required content for a definition line. I say controversially since some people thought that it would be a good idea to create many definitionless entries, but there was no consensus either way, from what I remember. Furthermore, a dump analysis can show that nearly all English Wiktionary lemma entries have a definition line with a definition or a translation.

The above seems consistent with WT:EL except that the example speaks of references, which are demonstrably lacking in an overwhelming majority of en wikt entries.

I am not aware of any further requirements on minimum entry content. In particular, as far as I know, there is no requirement on provision of pronunciation and inflection. During my time of contribution of Czech entries to the English Wiktionary, I mostly avoided entering pronunciation and inflection, focusing rather on semantics.

What do you think? Should there be increased requirements on minimum content beyond the three items above? Should such requirements be specified on a per-language basis? If so, should the decision be delegated to a small group of editors of a particular language, say 3 editors if there are no more? Thus, should the English Wiktionary be split into small oligarchies rather than there being One English Wiktionary?

--Dan Polansky (talk) 19:20, 16 February 2019 (UTC)
 * We don't need a legalistic framework or "small oligarchies". Dan, nobody I can think of wants to institute strict rules about what entries need to have at minimum. We were just asking you to put in a slight amount of effort, like putting in the gender of a noun when the very dictionary you're referencing gives the gender, or even just using a template like rather than . That's it. —Μετάknowledge discuss/deeds 19:21, 16 February 2019 (UTC)
 * On my talk page, it says a Russian entries need to 1) include the accent, [...], 3) include the declension or conjugation, and 4) include the pronunciation. I ask the editors if they would be so kind and indicate whether they want to establish minimum content above my three listed items. --Dan Polansky (talk) 19:23, 16 February 2019 (UTC)


 * One of WF's tricks is to create rfdef entries with nothing but a lazy quotation from the sports news, sometimes SoP. But the worst ones I can remember were the dozens of [name_of_country] Sign Language entries with no definition and often not meeting CFI. I don't think it's been a big enough problem to need policy yet (in English anyway). Equinox ◑ 19:31, 16 February 2019 (UTC)
 * By the way, I was one of those vehemently opposing volume creation of definitionless entries. Semantics is the life and soul of a dictionary, by my lights. --Dan Polansky (talk) 19:39, 16 February 2019 (UTC)


 * I think this whole thing has gotten seriously out of hand. Meta maybe came across as a bit officious, but it was a reasonable request- as a request. Dan interpreted it as more of an order, and got defensive- after which it escalated. There are legitimate issues about burdening editors in specific languages with fixing up terms that they wouldn't have created themselves and hijacking their priorities- but that's a matter of courtesy, and far too complex to reduce to rules. We've all created entries that needed work by others, and the dictionary would be a fraction of what it is now without that. We need empathy and consideration, not arguments and battles- it's too easy to drive away good editors over such things. Chuck Entz (talk) 22:13, 16 February 2019 (UTC)
 * You see, in User_talk:Dan Polansky/2018, I received the following order from Metaknowledge: "do not create entries in languages you do not know and have not studied". I think interpreting communications from the same contributor on the same subject as orders in disguise is pretty reasonable. But this thread is about policy, not about me in particular, and is merely triggered by certain posts on my talk page. The key question is, shall small subcommunities be able to increase the requirements for minimum entries per language, and therefore, should the English Wiktionary be understood as a collection of oligarchies, small ruling groups? --Dan Polansky (talk) 08:04, 17 February 2019 (UTC)
 * The key is cooperation. You shouldn't say that you refuse to do it and say the entries are fine as they are if they are not for editors of that language. You can simply add and  The request for higher standard of entries based on existing entries is legitimate, even if it gets harder to keep the same minimum level of quality of entries is already high. You can make simpler stubs for languages with low contents but you can still mark them with  so that other editors can at least find entries that require attention. As for Russian entries, it takes more effort, knowledge and time but it's not that Russian inflections and genders are unavailable. It doesn't belong to poorly documented languages. But, since it can also be error-prone, editors with less knowledge of a language shouldn't be completely discouraged from editing but are asked to mark them incomplete. Everybody does it. I did too for languages I wasn't confident in and when I knew what was required, for example, languages with complex scripts. It's strange that you vehemently opposed definitionless entries with  for otherwise great entries for high frequency words. It often takes much less effort to add a definition than reformat headers and add inflections. --Anatoli T. (обсудить/вклад) 09:07, 17 February 2019 (UTC)
 * The subject of this thread is minimum content, not marking. To address your subject (out of scope of this thread) of marking entries with attention: no such marking is required since if there is consensus that entries with ru need to be in a convenient category, head can be instructed to place such entries into a maintenance category automatically. Czech entries without inflection are not marked with attention and as far as I know, such marking is not a common practice for most languages, and I can make a dump analysis to check the actual facts; "everybody does it" is easily verified to be false. Here again, the general question that I saw no clear answer to so far is, should small groups make up their own rules for other editors to follow? --Dan Polansky (talk) 09:19, 17 February 2019 (UTC)

You said “I want my undivided attention to be channeled toward making sure that the semantic information I am entering is correct” but I deny that using the appropriate templates excludes it. The templates all have the same names and you can even care for it it after you added the glosses. And of course for Russian the stress is one of the main reasons why one consults the dictionary, so unless one has no information about the pattern because it is a kind of archaic word nowhere included, one can already give the complete information in the headword and in the table, which latter is important because Benwing’s bot creates the non-lemma forms and users like to look into the tables. About adding pronunciations: For English it is not easily predictable, so editors don’t add it because they don’t know it (English is the only example for “irregular” in ). In most other languages the pronunciation sections have indeed the character of clutter we only add because we have unlimited room, but the stress mark in the head-word or declension-table is what you need to know the pronunciation already, unless such a case like which you could guess wrongly either if you know the stress is on the beginning but have not heard the word. For languages like Arabic and Aramaic where multiple pronunciations can be on the same page I am for avoiding adding IPA pronunciations because it only makes the layout complicated without adding additional information (because as I said the full vocalization or transcription gives all information already) and indeed I create the pages faster and with better overview if I do not add the pronunciation, so I think they divert me and the reader. For Russian, perhaps the bot can add IPA pronunciations since the cases should be all included already. Just ask yourself what the reader would like to know from Wiktionary: It is the stress pattern and the gender for the languages that have such, and the meaning, and even if you have the pattern you have the gender already most likely in Slavic languages and it is only one letter, all if only you know it, so the demands are really low. There being links to other dictionaries is a bad argument to omit stresses and patterns, since copying over the stresses and patterns is what you should do, and for the languages in question many web searches can confirm. “Accuracy combined with verification” does not stop you to tell people what you already know. Also add surface derivations, if you have reasonable ideas of them, else others have to add it. BTW is an incomplete wrapper of, some times I used it I had to use , because it does not support m / f. Fay Freak (talk) 13:48, 17 February 2019 (UTC)
 * Dan, my main concern is that you work *with* the main contributors in a given language. Overall, I completely agree with what said. This is not a matter of enforcing rules but of (a) keeping up the overall quality of Wiktionary by attempting to follow the example of existing entries, and of (b) maintaining harmonious relationships with others. In this case, if you had tried to figure out the prevailing structure and templates of a Russian entry, and found it too complex, and instead inserted  or  or a similar request template, I'd have no problem with this. But you seem to have made no such attempt, and in general appear to show little interest in working with others or maintaining consistency. If everyone did this, the whole project would descend into chaos. Benwing2 (talk) 19:59, 17 February 2019 (UTC)
 * Is it your position that a Russian noun entry must contain pronunciation and inflection as a minimum, or is it not your position? I am puzzled. --Dan Polansky (talk) 20:26, 17 February 2019 (UTC)
 * As for what I am doing, which is out of scope of this thread per its title, I am interested in using the generic tools for setting up an entry, which is head, since I am basically little like a slow-moving Tbot working with a plethora of languages, using general human intelligence to verify semantics in applicable sources. Since I work in so many languages, I am not interested in learning any template peculiarities that various language groups may have set up. I need the minimum entries as places to attach verification artifacts and further reading goodies, which happen to be the same thing. As much must be pretty clear to anyone who saw my recent batch of contributions. I am not acting out of malice or disregard for wishes of particular groups, but my enterprise can only work economically if I can work with generics, or non-demanding templates such as be-noun, which I am now starting to use. I am absolutely not interested in pronunciation or inflection. I am no worse than Tbot, and in fact, I am better in multiple ways: Tbot checked in other Wiktionaries whereas I am checking in external sources even for entries that not a single other Wiktionary has, and I do human checking of semantics, not just checking for existence. I will run out of gumption pretty soon, I guess, and return to creating Czech entries; my best hope is that other editors will pick up the work, including new editors. --Dan Polansky (talk) 20:48, 17 February 2019 (UTC)
 * : I would discourage from using and language-specific templates because this can produce incorrect results - a wrongly detected gender, animacy and a stress/inflection pattern (many things are automated) without your knowledge. It also requires a correct word stress. All we ask for is adding maintenance templates, so that appropriate editors could bring the entry to the required standard. When I said "everybody does it", I meant everybody who is asked to do it. E.g. people know that Chinese entries require traditional and simplified forms. What if you don't know? You need to ask people who know. --Anatoli T. (обсудить/вклад) 22:38, 17 February 2019 (UTC)
 * I would add that Tbot added a maintenance category to every entry it created, so that others would know to go back and check on it- in that respect, these current entries are inferior to Tbot's. Chuck Entz (talk) 01:35, 18 February 2019 (UTC)
 * Thank you for adding the template to трактористка. That allowed me to find it and fix it up. Benwing2 (talk) 16:30, 18 February 2019 (UTC)
 * I can keep adding attention to Russian entries, but a proper (economical, systematic) way to address this need would be to teach head to put ru to a maintenance category. Otherwise, you will need to keep asking newcomers to do something that machines can do. As for Tbot, it was only a bot doing no real verification so it had to mark the entries with a reader-visible template, indicating a significant risk of inaccuracy; I am doing human verification and take credit for accuracy.
 * On another note, I came up with a motto: Make yolk and hub and skip all fluff. I maintain that pronunciation and inflection are fluff, things insubstantial, an order of magnitude less important than definitions and translations. Let us do the following thought experiment: you want to learn Serbo-Croatian and you are offered three dictionaries; which one do you buy?
 * Dictionary 1 has 100 000 lemma entries with pronunciation and inflection and no semantic information.
 * Dictionary 2 has 5 000 lemma entries with definitions or translations, and pronunciation and inflection.
 * Dictionary 3 has 20 000 lemma entries with definitions or translations but no pronunciation and inflection.
 * Which one do you buy to learn Serbo-Croatian? The thought experiment explains why I oppose creation of definitionless lemma entries in volumes; I admit that creating definitionless lemma entries in small volumes (up to 1000?) can be useful in that people can fill in the definitions/translations in reasonably short time. --Dan Polansky (talk) 07:35, 23 February 2019 (UTC)

Module:la-headword
There is an (AFAICT) undiscussed removal of valuable information going on leading to incomplete and incombrehensible head lines. --Hamator (talk) 11:47, 17 February 2019 (UTC)

Classical Malay?
I changed the meaning of the worklang= param in etc. Formerly it took either a single language code or an arbitrary string like "French and Latin" or "Classical Malay". I changed it so it takes one or more comma-separated language codes, but doesn't allow arbitrary text. I fixed up all the resulting errors except for two, which are in -kah and -kan, which have quotations in Classical Malay, for which we don't have any language code. Could someone add this? I'm not sure if it should be an etymology-only language or a proper language in its own respect. (And what about Old Malay?) Benwing2 (talk) 19:15, 17 February 2019 (UTC)


 * Thank you for bringing this up. A few months ago I bought up a similar suggestion at Beer parlour/2018/September. Currently, Classical Malay (14th to 18th century) and Old Malay (7th to 14th century) do not have proper language codes defined for them. However, because there is a lack of effort to digitize texts from Classical Malay (written in Jawi script) and Old Malay (written in Pallava script or Rencong script) in its original script form, I think we can wait for ISO 639 to define a proper language code for these languages.
 * Currently, only two Classical Malay works are available on Wikisource: Hikayat Hang Tuah and Hikayat Bayan Budiman. Modern transcriptions of Classical Malay works are often written in the Latin script, so it is slightly problematic to figure out its original orthography in the absence of an original manuscript.
 * By the way, Classical Malay and Old Malay is the missing link between the Proto-Malayic language and the modern Malay language. KevinUp (talk) 22:36, 18 February 2019 (UTC)
 * I have removed the  parameter in  and  because the texts have been translated and modified to suit readers proficient in modern Malay, rather than transcribed word-for-word based on the original manuscript. KevinUp (talk) 22:36, 18 February 2019 (UTC)

Should I suppress the "(please add an English translation of this quote)" message for Scots?
Many of the Scots quotations given are so close to English that they are readily understandable without any "translation". Example:
 * "Och, it's the lassies will be the pleased ones, coiling the blankets round them; it's Auld Kate that kens," and then she gave a screitchy hooch and began to sing in her cracked thin voice-- 'The man's no' born and he never will be, The man's no born that will daunton me.'

Not surprisingly no translation is given, but if this is tagged with sco, you'll see "(please add an English translation of this quote)". Given the predominance of this situation, should I special-case Scots to remove this message? Benwing2 (talk) 02:51, 19 February 2019 (UTC)
 * No. If Scots isn't to be translated it shouldn't be a separate language. DTLHS (talk) 02:52, 19 February 2019 (UTC)
 * Agreed. I'm not 100% on what this means. Per About Scots, we consider it a separate language instead of a dialect of English. If we consider it English, it's a different story. —Justin ( koavf ) ❤T☮C☺M☯ 03:59, 19 February 2019 (UTC)

Constr d uction namespace
Has it been suggested that constructed languages, like Esperanto, be moved to a "Construction" namespace, ex. Construction:Esperanto/eburo? -- 11:15, 19 February 2019 (UTC)
 * Seems reasonable. We already have a "Reconstruction" namespace. SemperBlotto (talk) 11:24, 19 February 2019 (UTC)
 * Is the title of this section meant to be something else? - TheDaveRoss  13:27, 19 February 2019 (UTC)
 * This is a great idea. Fay Freak (talk) 14:07, 19 February 2019 (UTC)
 * Ooh... it's a game! Reconstruction>Construction>Contraction>Distraction>Destruction... it almost worked...
 * Seriously, though, I'm not impressed by the name: Reconstruction: houses reconstructions, but Construction: would house terms in construct ion ed languages. Chuck Entz (talk) 14:42, 19 February 2019 (UTC)
 * And I would have gotten away with it too, if it wasn't for you meddling kids! -- 19:30, 19 February 2019 (UTC)
 * I think what Chuck is saying is he'd prefer "Constructed:" instead of "Construction:". Personally I'm on the fence as to whether this is needed at all, under any name. Benwing2 (talk) 15:39, 19 February 2019 (UTC)
 * I'm not married to the name; I just thought it in keeping with the "Reconstruction:" namespace. I'm fine with "Constructed:" or simply "Construct:", but it could be "Conlang:" or "Artificial:" for all I care. I just think if we keep reconstructions off the main namespace, why should conlangs be shoehorned into natural (if you will, non-constructed if not) languages? I think it's confusing to the reader, as they might mistake Esperanto, for example, for some inherited descendent of Latin, as we have no indicator that it's a constructed languages, like we do reconstructions. I also find that they clutter up entries and every few months there seems to be some vote on allowing another (forgive the hyperbole, but you get my point). -- 19:30, 19 February 2019 (UTC)
 * I see your point. I don't find it especially confusing but I imagine it might be different for users who haven't heard of Esperanto, Interlingua, Lojban, etc. (OTOH a page like a already has a huge number of random languages on it, and the average user isn't likely to have heard of Kalasha, Mandinka, Lower Sorbian, or Mezquital Otomi, to name a few on that page, and won't get any more confused by the additional presence of Esperanto, Interlingua, Ido, Novial, etc. on the same page.) Benwing2 (talk) 19:56, 19 February 2019 (UTC)
 * Exactly. What's stopping layman users from thinking Esperanto and Kalasha are categorical equivalents? -- 20:09, 19 February 2019 (UTC)
 * So instead they're supposed to think Na'vi and Esperanto are equivalents? There are clearly more nuanced distinctions to be drawn than "constructed" vs "not constructed". DTLHS (talk) 21:54, 19 February 2019 (UTC)
 * Yes, I would say more so than Esperanto and Kalasha. -- 22:53, 19 February 2019 (UTC)


 * Oppose. Esperanto has become too big to be cordoned off, and unlike nearly every other constructed language, people are going to be looking for it where they look for other languages. —Μετάknowledge discuss/deeds 20:24, 19 February 2019 (UTC)
 * , and what would stop them to finding them at another namespace other than main? -- 21:18, 19 February 2019 (UTC)
 * Let's see... they won't come up in search, they won't be in translation tables... you'd have to be looking for them to find them, which is good for Novial, but bad for Esperanto. —Μετάknowledge discuss/deeds 21:24, 19 February 2019 (UTC)
 * And by "won't come up in search" you mean in the search dropdown because reconstructions certainly show up in search results. There maybe be a technical solution for that, ditto for translation tables, though I'm less familiar with the problem there. -- 21:32, 19 February 2019 (UTC)
 * Oppose. I am opposed to deciding we don't want to include languages, and then including them anyway in a roundabout way. Conlangs that aren't in mainspace shouldn't be included anywhere at all, not in any namespace. —Rua (mew) 20:30, 19 February 2019 (UTC)
 * @CodeCat, I don't think anyone is suggesting moving poorly attested conlangs like Lojban and Novial to this namespace, as those are being relegated to Appendix:. I'm cliefly referring to Esperanto and Interlingua. -- 21:16, 19 February 2019 (UTC)
 * : FYI, the following vote seems relevant to your position: Votes/2019-01/Moving Novial entries to the Appendix. --Dan Polansky (talk) 14:13, 23 February 2019 (UTC)
 * Oppose. If a constructed language is so unused it should be banished to an appendix or deleted, do that. But e.g. Esperanto is more widely used than at least several hundred of the natural languages we include, and even has some native speakers; I don't see a reason to segregate it into a separate namespace away from e.g. Mbariman-Gudhinma or Berbice Creole Dutch just because we can identify who coined most of Esperanto's words and not the other two languages'. - -sche (discuss) 23:29, 19 February 2019 (UTC)
 * Oppose. I don't understand the motivation for the concept of "categorical equivalents" (I guess German and Dutch are "equivalent"? What does that even mean?) or the need to keep them separate. I don't understand why someone mistaking Esperanto for a Romance language is a danger that we should take seriously. It seems no more plausible or harmful than someone seeing and concluding, on that evidence alone, that English is a Romance language. It seems extremely condescending to think that someone looking up words in a language would not have such basic knowledge about the language, and I don't see that we're at all responsible for someone using the dictionary that stupidly.__Gamren (talk) 12:49, 23 February 2019 (UTC)
 * Oppose. Esperanto is attested in use; reconstructions such as PIE are not. Mainspace is the natural place for users to look up terms, whatever the language. As for the poor souls who cannot look up entry Esperanto or check Wikipedia's Esperanto, they should try harder and look it up. I think PIE could have been in the mainspace, with entries starting with asterisk (*), but that ship has sailed. --Dan Polansky (talk) 14:23, 23 February 2019 (UTC)
 * Oppose. Languages that have actually been used in human communication should be in mainspace. I'd rather move languages like Mbabaram (dog), which has a tiny recorded vocabulary and no recorded literature, away from mainspace, than move languages with literatures away from mainspace.--Prosfilaes (talk) 08:12, 24 February 2019 (UTC)

Constellation name definitions
Considering that the IAU has recognized 88 constellations, should the constellation names be defined as translingual, and should the English definitions be moved there? -Mike (talk) 23:39, 19 February 2019 (UTC)

Scots again, and Middle English
A number of entries have quotations from Template:RQ:Dictionary of the Scottish Language used to illustrate English terms. An example is forspeak, for which definition #1 says "(transitive, dialectal, Northern England and Scotland) To injure or cause bad luck through immoderate praise or flattery; to affect with the curse of an evil tongue, which brings ill luck upon all objects of its praise." Should we allow this? If so, what language should I use to tag the Scots portions of the quoted text? en (English) or sco (Scots)? If not, what should happen to these quotes? (Move to a Scots L2 section? But what about the "Northern England" label?) Note that on the same page is also a Scots entry for forspeak, defined as "To bewitch or cast a spell over, especially using flattery or undue praise; to seduce." Examples like this make me think that the entire decision to include Scots as a separate language may have been wrong, because (a) most terms (like this one) that exist in Scots and don't exist in Standard English also exist in Northern England dialects; (b) in general there's no way to make a clear distinction between Scots and nearby dialects of English. Note that if Scots had a standard literary form the situation would be different, because then we could define the nucleus of Scots as consisting of that literary form.

A related issue: A number of English entries have illustrative quotations from Middle English. An example is ashame, which has a quote from Wycliffe's Bible dated to 1390, complete with translation: "Ashame thou, Sidon, seith the se, the strengthe of the se, seiende, I trauailide not with child, and bar not, and nurshede not out ȝung childer, ne to ful waxing broȝte forth maidenes." (translated as "Be ashamed, Sidon, says the sea, the strength of the sea, saying, “I did not travail with child [give birth], and did not nurse boys, nor to full waxing bring forth maidens.") Should we allow this? If not, what should happen to these quotes? (Move to a Middle English L2 section?)

Benwing2 (talk) 23:42, 19 February 2019 (UTC)


 * Regarding Middle English quotations: my inclination is to move them to Middle English sections, but some editors have argued they are tolerable in English sections to demonstrate age/continuity of use. (They don't count towards attesting the term, obviously, but neither do e.g. quotations from websites, which are nonetheless infrequently included alongside ATTEST-satisfying citations if they are particularly good illustrations of a term.) - -sche (discuss) 23:54, 19 February 2019 (UTC)


 * Regarding Scots quotations: right now, they should be moved to Scots entries. Since both Scots and English are WDLs, merging them might need a vote, or at least strong consensus support. But it would certainly simplify distinguishing Scots from Scottish English at RFV if we, erm, didn't distinguish them. And we already include several rather divergent dialects (e.g. Geordie) under English, so I don't expect the issues of e.g. different inflected forms and the like to be much harder to handle than for those dialects. And other (monolingual) English dictionaries tend to include Scots as English. - -sche (discuss) 01:43, 20 February 2019 (UTC)


 * I considered it odd that Scots is separated from English on Wiktionary, but I've held back from raising the topic myself (it's potentially an emotive subject!). I'm a Geordie and have lived in Scotland, and I find that most of the Scots and Geordie terms are the same (supporting Benwing2's "(a)"). There's work to do in adding Northern English terms to Wiktionary; it's something I've been avoiding so far. Having had a look through, I'd come to the conclusion that the bulk of the job would be to take Scots entries and duplicate them as English (Geordie) entries. As a simple example, User:Stelio/Tyneside Songs has a bunch of orange links (if you've turned on that gadget) most of which are Scots terms, and the songs themselves could easily be misidentified as Scots to someone unfamiliar with their context. -Stelio (talk) 11:00, 20 February 2019 (UTC)
 * As someone who has spent a significant time around drunken Geordies, my vote is for the Geordie lect being considered its own language. XD -- 17:46, 20 February 2019 (UTC)
 * Mebbees like, but how man, dinna fash yersel'. That's wark, reet? ;-) -Stelio (talk) 14:25, 21 February 2019 (UTC)
 * Why aye man, canny wark! -- 14:53, 21 February 2019 (UTC)


 * OK, I drafted Votes/pl-2019-02/Treat Scots as English. Please improve or postpone as needed. - -sche (discuss) 23:42, 23 February 2019 (UTC)
 * Thanks. Maybe you should note in the vote that combining Scots with English on Wiktionary makes no statement as to whether Scots should be considered a separate language, but is being suggested due to the difficulty of drawing a clear line between Scots and English (at least, that is my view ...). Benwing2 (talk) 01:59, 24 February 2019 (UTC)
 * OK, I took out the part about "treating Scots and English as a single language". But the very act of treating Scots as English, no matter how it's worded, is inherently treating it as not being a separate language... - -sche (discuss) 02:50, 24 February 2019 (UTC)


 * English and Middle English are different languages, hence Middle English quotations belong into Middle English entries only. To show age/continuity there's the section "Etymology". --Brown*Toad (talk) 18:24, 10 March 2019 (UTC)

Talk to us about talking
The Wikimedia Foundation is planning a global consultation about communication. The goal is to bring Wikimedians and wiki-minded people together to improve tools for communication.

We want all contributors to be able to talk to each other on the wikis, whatever their experience, their skills or their devices.

We are looking for input from as many different parts of the Wikimedia community as possible. It will come from multiple projects, in multiple languages, and with multiple perspectives.

We are currently planning the consultation. We need your help.

We need volunteers to help talk to their communities or user groups.

You can help by hosting a discussion at your wiki. Here's what to do:
 * 1) First, sign up your group here.
 * 2) Next, create a page (or a section on a Village pump, or an e-mail thread – whatever is natural for your group) to collect information from other people in your group.  This is not a vote or decision-making discussion: we are just collecting feedback.
 * 3) Then ask people what they think about communication processes.  We want to hear stories and other information about how people communicate with each other on and off wiki.  Please consider asking these five questions:
 * 4) When you want to discuss a topic with your community, what tools work for you, and what problems block you?
 * 5) What about talk pages works for newcomers, and what blocks them?
 * 6) What do others struggle with in your community about talk pages?
 * 7) What do you wish you could do on talk pages, but can't due to the technical limitations?
 * 8) What are the important aspects of a "wiki discussion"?
 * 9) Finally, please go to Talk pages consultation 2019 on Mediawiki.org and report what you learned from your group.  Please include links if the discussion is available to the public.

You can also help build the list of the many different ways people talk to each other.

Not all groups active on wikis or around wikis use the same way to discuss things: it can happen on wiki, on social networks, through external tools... Tell us how your group communicates.

You can read more about the overall process on mediawiki.org. If you have questions or ideas, you can leave feedback about the consultation process in the language you prefer.

Thank you! We're looking forward to talking with you. Trizek (WMF) 15:01, 21 February 2019 (UTC)

Words of language X used in language Y

 * some previous discussions: December 2018, September 2017

It is not difficult to find uses of the English word in Dutch texts. In many cases this is obviously an instance of code switching (the word is italicized or put between quote signs, or occurs in a longer English phrase) but in some there is no obvious giveaway. Yet I expect most Dutch speakers will agree that this is not a Dutch word. On the other side I expect most Dutch speakers will readily agree that belongs to the Dutch vocabulary, including those speakers who pronounce the word as. A giveaway may be that the Dutch diminutive trammetje is actively used, while *happinessje does not occur. What set off this musing was a request for verification for (allegedly). I think this is an English term used in Dutch texts for lack of a native term. Elsewhere there was a reference to ; I am inclined to think this is a Middle English word used in Old French texts for lack of a native term. Is there a way to make this more firm? What criteria can we use to help us decide when uses of a word from language X in language Y stop being instances of expedient code switching and become evidence of incorporation in the lexicon of Y? --Lambiam 06:20, 23 February 2019 (UTC)


 * This has come up before (I added two of the most recent discussions, one involving "marketing" as a Greek word, up top). I agree it's a thorny issue. Italics, quotation marks and different script are all good clues to code-switching, although it's worth noting that something can only be code-switching if the word and sense exists in the supposed donor language: French is frequently italicized and thought of as a loanword or code-switching, but cannot be code-switching/English and must be a French word because English  never means "a celebrity". Since we've had 3+ threads about this recently, I'll start a draft / think-tank page for this, a la English adjectives. - -sche (discuss) 07:01, 23 February 2019 (UTC)
 * The draft is now live at Code-switching. - -sche (discuss) 07:33, 23 February 2019 (UTC)


 * (Beer parlour/2019/February isn't entirely irrelevant. —Suzukaze-c◇◇ 07:25, 23 February 2019 (UTC))


 * I would like to share this journal article I found which was published in the International Journal of English Linguistics in 2011: Code-Mixing of English in the Entertainment News of Chinese Newspapers in Malaysia. KevinUp (talk) 19:46, 23 February 2019 (UTC)
 * I sort of agree with you but am going to play devil's advocate for a moment, because I'm not sure we should remove such entries in all cases. To take the statement 'I think this is an English term used in Dutch texts for lack of a native term.' -- well, is this not how many loanwords are borrowed in the first place? Especially for specialized terms with no native Dutch word, a borrowing-entry (where it meets CFI) might be useful: after all, if one hypothetically were to look for the Dutch translation over at anti-roll bar, they would then find it and understand that there is no current native term, but that this originally English term is in fact in use in Dutch by Dutch speakers. — Mnemosientje (t · c) 20:04, 23 February 2019 (UTC)
 * We can still explain that in the translation table without having to create an entry: "no equivalent in Dutch; the English term is used instead" or something of the sort. The relevant quotes can be put in Citations:anti-roll bar. Per utramque cavernam 20:14, 23 February 2019 (UTC)
 * I'd call it a proto-loanword. The distinction between a proto-loanword and a real loanword is not clear though, one flows into the other. is very clearly French to a Dutch speaker, yet it's used widely in Dutch despite the existence of . Perhaps a solution would be to use the "hot word" approach here: if the word continues in use for X years, then assume it has entered the lexicon of the borrowing language. —Rua (mew) 17:10, 24 February 2019 (UTC)
 * Yes, I agree with the first part of your message. I'm less sure about the idea of using the "hot word" approach; would being used for X years be a sufficient proof that a word has entered the lexicon? Per utramque cavernam 17:20, 24 February 2019 (UTC)
 * It's a subjective question, so what matters is whether it's proof enough for Wiktionary. —Rua (mew) 17:27, 24 February 2019 (UTC)
 * The point of a hot-word approach would be there would be a review some number of years after the date of first attestation. I still wonder what makes something qualify for the proto-borrowing status. Three cites with no typographic marking of the term, but a substantial (75?, more?) majority of appearances with typographic marking? Would a proto-borrowing become an ordinary borrowing the first year there were three of more attestations without typographic marking and fewer attestations with such marking? Is there a simpler approach that has some face validity. DCDuring (talk) 20:12, 24 February 2019 (UTC)

Categories which don't exist, but have been added by users anyway
There are a large number of non-standard categories which have been chosen by users which appear in Category:Categories with invalid label. The question is, of course, whether these should be created, or whether any of them can be reassigned to existing categories. Some are unusual to say the least, for example Category:ceb:Dried fish. DonnanZ (talk) 19:32, 23 February 2019 (UTC)


 * I just made Category:en:Bricks, which should exist. --Wonderfool early February 2019 (talk) 19:35, 23 February 2019 (UTC)
 * There has been a user who has contributed a very large number of entries for Cebuano names of organisms. I'm not surprised that there are more ceb categories than one might naively expect. If dried fish is important in the diet of the Philippines, why not have the category? DCDuring (talk) 23:08, 23 February 2019 (UTC)
 * There are 11 entries in Category:en:Seasonings. There are 10 in Category:ceb:Dried fish. DCDuring (talk) 23:12, 23 February 2019 (UTC)
 * I could be wrong, but half of those look like SOP compounds, with bulad and buwad referring generically to dried fish, and the second words referring to the type of fish that's dried. Chuck Entz (talk) 00:19, 24 February 2019 (UTC)
 * I want to note that "nonstandard" categories only get sorted into Category:Categories with invalid label if they try to use . This means that 1) categories which are valid for one language but maybe not useful to others can exist, they just need to be put into parent categories manually and without using, and 2) nonstandard categories that don't use  are out there. (We could potentially find them by searching a database dump (or possibly using the built-in site search) for pages in the category namespace that don't contain .) "Dried fish" might be useful to a number of Asian / Pacific languages, I don't know; if it is, it could be added to the official category tree so that  would "play nice" with it. - -sche (discuss) 23:24, 23 February 2019 (UTC)
 * Also, it is hardly a surprise that a new contributor would not master our baroque category implementation.
 * It is largely due to the efforts of Chuck Entz that so few (1,167) items remain in Special:WantedCategories. If only there were similarly diligent efforts to clear Special:WantedTemplates and Special:WantedPages, either by removing the links or by adding the "wanted" templates and pages. DCDuring (talk) 23:32, 23 February 2019 (UTC)
 * It's quite been a while since I patrolled Special:WantedCategories. Now User:DTLHS does most of the heavy lifting with his bot and autocat is very easy to use, so I haven't felt the need. I spend most of my time approaching categories from other angles. Chuck Entz (talk) 00:09, 24 February 2019 (UTC)
 * 23. Category:en:Towns in Alberta, Canada should be removed from that (see below), but is incorporated in an entry somehow. DonnanZ (talk) 09:30, 24 February 2019 (UTC)
 * That was happening somewhere in the innards of Module:place. fixes it. Have to do the same edit for all other Canadian provinces and territories though; not very efficient. — Eru·tuon 09:57, 24 February 2019 (UTC)
 * Ah, thanks. DonnanZ (talk) 10:32, 24 February 2019 (UTC)
 * It was agreed upon a while ago that category names for subdivisions would always include the country name. So the name with "Alberta, Canada" is actually the right one going forward. —Rua (mew) 17:02, 24 February 2019 (UTC)
 * Well, we know what happened to US states, and who was responsible. There is only one Alberta, New South Wales, Florida etc., so that's not the way to go, IMO. DonnanZ (talk) 17:13, 24 February 2019 (UTC)


 * I did create a couple of standard categories: Category:Towns in Alberta and Category:Villages in Finland. These are still on the invalid label list, but will hopefully disappear soon. DonnanZ (talk) 00:41, 24 February 2019 (UTC)


 * Dried fish is important for Russians too, in case you try to conceive usage fields. Every grocery store targetted at the Russia-related population in Germany has a section filled with various kinds of dried fish, which however are conceived as supplementary foodstuff and not as staples. They are employed like roasted sunflower seeds, somewhat cliché. Fay Freak (talk) 00:45, 24 February 2019 (UTC)
 * To note, if a topic cannot contain enough entries in a sufficient number of languages to justify adding category data but the grouping in one language is recommendable, the solution is to create a thesaurus entry. Fay Freak (talk) 00:54, 24 February 2019 (UTC)


 * That's another thing that I have noticed before; looking at Category:sms:Lakes in Finland, Category:Lakes doesn't have any subcategories for "Lakes in ...". I don't see why it shouldn't; although I don't envisage it should contain every lake in Finland or Canada for example. Currently the category lists terms relating to lakes, e.g. in Category:en:Lakes, as well as actual lakes all over the world. DonnanZ (talk) 16:20, 24 February 2019 (UTC)
 * It is a set category though, so it should not contain terms related to lakes. —Rua (mew) 17:04, 24 February 2019 (UTC)
 * I don't see why not, and this is where "Lakes in ..." subcategories would come in handy. If there is no objection to that, I think I can set them up. DonnanZ (talk) 17:25, 24 February 2019 (UTC)
 * I do not support converting Category:Lakes into a topical category. —Rua (mew) 17:30, 24 February 2019 (UTC)
 * That's odd, because you created this. DonnanZ (talk) 17:36, 24 February 2019 (UTC)
 * I don't see anything odd about it, it's like Category:Countries and Category:Cities. —Rua (mew) 17:49, 24 February 2019 (UTC)
 * The odd thing about it is that it appears in Category:Categories with invalid label because no module has been set up, so do we fix that or not? DonnanZ (talk) 18:00, 24 February 2019 (UTC)
 * If you don't want to play ball a move to Category:sms:Lakes would be an alternative. A module can easily be created for that already exists. DonnanZ (talk) 21:32, 24 February 2019 (UTC)

Venn diagrams for semantic nuances
I think it would improve the entries to add visual examples such as Venn diagrams, for example for the semantic relation in some fields between the terms assimilation, inclusion, exclusion, segregation and integration. --Backinstadiums (talk) 15:48, 27 February 2019 (UTC)
 * I think these visualizations can easily be misunderstood. “Exclusion” might illustrate expulsion. “Assimilation” might illustrate expulsion followed by cleansing. “Integration” could serve as an illustration of the Nazi concept of . --Lambiam 23:21, 27 February 2019 (UTC)

Anglo-Norman
We must add Anglo Norman tag xno to the tag repository.Aearthrise (talk) 22:19, 27 February 2019 (UTC)
 * We consider Anglo-Norman a dialect of Old French. Please enter terms under the L2 ==Old French== and label them fro to categorize them in CAT:Anglo-Norman Old French. —Mahāgaja · talk 22:53, 27 February 2019 (UTC)
 * I come across this in etymology, where it is necessary to use en (or similar). DonnanZ (talk) 00:33, 28 February 2019 (UTC)
 * Actually you can use en; the template will display "Anglo-Norman" but link to "Old French". —Mahāgaja · talk 11:15, 28 February 2019 (UTC)
 * I wasn't aware of that, I will try that next time. Thanks. DonnanZ (talk) 11:23, 28 February 2019 (UTC)
 * It works for all the etymology-only languages. —Mahāgaja · talk 11:28, 28 February 2019 (UTC)
 * OK, I knew you could do that with Late Latin, Medieval Latin etc. DonnanZ (talk) 11:33, 28 February 2019 (UTC)