Wiktionary:Beer parlour/2016/April

discussion rooms: Tea roomEtym. scr.Info deskBeer parlourGrease pit ← March 2016 · April 2016 · May 2016 → · (current)

Category:Classical LatinEdit

Aren't Latin entries presumed to be classical unless otherwise noted? Should we have this category? DTLHS (talk) 03:18, 2 April 2016 (UTC)

Indeed, that's the standard we've used. The use of this as a context label is fine when indicating a spelling, but can we prevent it from categorising? —Μετάknowledgediscuss/deeds 03:27, 2 April 2016 (UTC)
For just Latin or everywhere? (AFAIK no other languages use the Classical label) KarikaSlayer (talk) 18:30, 5 April 2016 (UTC)
@KarikaSlayer Classical vs. Vedic Sanskrit comes to mind. —JohnC5 18:38, 5 April 2016 (UTC)
@JohnC5 I was thinking more about the code under "classical" in Module:labels/data/regional. To my knowledge Latin is the only language that uses that specific label (Classical Hebrew has its own tag, and it doesn't look like there's a Category:Classical Sanskrit). KarikaSlayer (talk) 18:48, 5 April 2016 (UTC)
Note that I just changed "Classical Hebrew" to be called "Biblical Hebrew", which is the much more common name. --WikiTiki89 18:53, 5 April 2016 (UTC)
Classical Arabic, Classical Chinese, etc. See w:Classical language for a long list. Although for Arabic we say "Koranic" or "Modern Standard". Not sure about the others. Benwing2 (talk) 23:32, 5 April 2016 (UTC)
There is a lot of Classical Arabic that is not Koranic. In fact, I would say Koranic Arabic is pre-Classical. --WikiTiki89 23:49, 5 April 2016 (UTC)

Catalan/Old Provencal and Walloon/Old FrenchEdit

I'm not sure if this has been discussed already, but there are a few issues with the way we handle Old Provencal in relation to Catalan. For one, it isn't universally accepted that the former was truly or technically an ancestor/parent language of the latter, even though I personally basically think so. Or at least I can agree an early form of what we may call Old Provencal or it's immediate ancestor was parent to both Catalan and Occitan. The problem is that many of the Old Provencal entries we have happen to be from later in time, in later stages of the language after certain sound shifts distancing and differentiating it (and the later Occitan) from what would become Catalan happened (or Old Catalan). So it may not be accurate to list a certain Catalan term as a descendant on an Old Provencal page of a term that was attested after the split Old Catalan made from it, as Old Provencal by then underwent several sound changes that already made it look more like Occitan but less like Catalan.

Like for one example, Catalan dolç vs. Old Provencal dous and Occitan doç, or another example: Catalan ocell vs. the 'au' diphthong arising in Old Provencal auzel and Occitan aucèl (a later unique development not to be confused with a direct inheritance from the original Latin 'au' in some cases). Or occir and aucir. In some cases, Catalan may have also received influence from neighboring Castillian Spanish, further differentiating it. I do agree that inherited Catalan terms can be listed as deriving from (some form of) Old Provencal, but not necessarily all the forms we have listed (the same even goes for some Occitan words, unless we specify which variant they came from, and not just the main lemma term; for example, how do we relate Cat. caure and Oc. caire, also càder and càser to Old Prov. chazer? Or the descendants on eu?). For the majority of terms, it isn't a problem, as they are very similar or identical, but there are some notable exceptions.

I've been making some of these descendant entries on Old Provencal or etymology entries on Catalan terms, based on what others have already done, but now I'm not so sure about the way we deal with it.

It does seem at least partly a matter of semantics, and what we call or how we define Old Provencal and Old Catalan and Old Occitan.

Also, is it really accurate to call Old French a parent language of Walloon? We're defining Old French that broadly, so as to basically include any Oïl language? Word dewd544 (talk) 20:12, 2 April 2016 (UTC)

  • First point, what else could be the parent language of Walloon? I can't think of anything. I think the answer to your question is yes Old French by our definition includes all the Oïl dialects including what are now the British Isles (including Ireland), France, specifically the northern half of France and what is now Belgium. I don't see how we say is came directly from Vulgar Latin as that leaves a gap of several hundred years. And there's no language or hint of a language called 'Old Walloon'. Renard Migrant (talk) 11:17, 20 April 2016 (UTC)
  • Not sure who the other user is creating Old Provençal entries. There's an argument for renaming the whole language Old Occitan because we merged Provençal and Occitan some years ago when ISO 639 retired prv. As for chazer it's the spelling used by Bernard de Ventadour as it's available on Wikisouce in s:fr:Catégorie:Œuvres de troubadors. FEW lists it as cazer, caire, caer (top of the second column, code is apr. for German Altprovenzalisch). Essentially the existence of chazer does not imply that cazer does not exist. I wouldn't worry too much about having the exact form (or mostly likely of the forms) that the word descended from. Note that Old French soloil originally listed French soleil as a descendant, but this got moved to Old French soleil. These things aren't set in stone.
  • Date-wise I have noted in Wiktionary:About Old Provençal that while we have a code for Old Catalan, roa-oca we don't have a cutoff either geographical or in terms of dates for Old Catalan. Of course if Catalan caure comes from Old Catalan, that also does not imply that it doesn't come from Old Provençal before that. Renard Migrant (talk) 11:24, 20 April 2016 (UTC)
A related issue, see the etymology of vriþa. The Old Norse form is Old Icelandic, which has a change of word-initial vr- > r-. But Swedish doesn't have this change, so it makes it look like the v disappeared and then reappeared. —CodeCat 17:53, 23 April 2016 (UTC)
Good points. That clarifies some things for me. I'm also for making Catalan entries potentially both descended from Old Catalan, if there is an attested term, and further on "Old Provencal". I suppose it's mostly the name of the language that causes the issues for some, as it makes it sound Occitan/Provence/southern France-centered, and may mislead people. Just wish there was a better general term for "Occitano-Romance", but it works.
I have no problem with Walloon being derived from Old French, but I was a bit confused by the fact that Gallo was not also considered a descendant language of Old French. It seems someone corrected that recently, however. Thanks. Word dewd544 (talk) 22:33, 30 May 2016 (UTC)

PIE verb lemmasEdit

Please see Wiktionary talk:About Proto-Indo-European#Lemmatising PIE verbs. —CodeCat 15:06, 3 April 2016 (UTC)

Singapore English entriesEdit

It looks like someone (maybe a teacher) in Singapore has noticed Wiktionary and has persuaded a group of people to start adding Singlish terms. Most of them are OK, but etymology sections can be over-complicated and formatting can be a little strange. I think it's better to clean the bad ones up rather than deleting them. On your guard! SemperBlotto (talk) 15:50, 3 April 2016 (UTC)

Maybe we should build a list so that they can be re-examined after the wave dies down. —suzukaze (tc) 06:29, 4 April 2016 (UTC)
Yeah, it seems they've been tasked with creating entries in a specific format (with "usage notes" that are not lexical and thus not appropriate here) and keep reverting if we change them. Not a great idea on a public, shared wiki project really. Equinox 16:12, 4 April 2016 (UTC)
The reversion isn't good, but the entries are interesting. DCDuring TALK 17:49, 4 April 2016 (UTC)
Can someone provide some links for those of us who have no idea where to look? --WikiTiki89 17:52, 4 April 2016 (UTC)
See contributions from Coffeeandbiscuits (talkcontribs), Potatogarcia (talkcontribs), Whatyoumean2016 (talkcontribs), Razif07 (talkcontribs), Kellytongjy1990 (talkcontribs), Syy03 (talkcontribs), Chingwennnn (talkcontribs), Afbak (talkcontribs), Nahte79 (talkcontribs), Maeaeae (talkcontribs), Heroaldchern (talkcontribs), Wani.lee (talkcontribs), E-van-316 (talkcontribs) (and others?) (not in any particular order). SemperBlotto (talk) 20:22, 4 April 2016 (UTC)
And Category:Singapore English. DCDuring TALK 21:03, 4 April 2016 (UTC)
For Singapore English cat members with usage notes: [1] DCDuring TALK 21:07, 4 April 2016 (UTC)
Well, that exercise seems to have ended, and teacher is now marking their work. I don't suppose we will ever hear of the results. SemperBlotto (talk) 18:08, 6 April 2016 (UTC)

Filter watchlists and recent changes by languageEdit

I added an option to WT:PREFS: "Filter watchlist and recent changes to only show changes for certain languages." (Third from the bottom of the "Experiments" section.) Suggestions, bug reports, feature requests, etc would be most welcome. --Yair rand (talk) 18:49, 3 April 2016 (UTC)

It seems it only filters out mainspace entries? --Giorgi Eufshi (talk) 06:27, 4 April 2016 (UTC)
Yes. Should it filter out other namespaces? --Yair rand (talk) 14:32, 4 April 2016 (UTC)
I would think that when you use it, it should show only changes in the mainspace and reconstruction namespace. Another flaw is that this is all done client-side, which means that the number of results you see is very small and inconsistent. If only this could be done server-side... But still, a useful gadget. Thanks, Yair! --WikiTiki89 14:58, 4 April 2016 (UTC)
Since we separate reconstructions by language anyway, it can be assumed that watching such a page means you're interested in that language. Or am I understanding this wrong? —CodeCat 15:03, 4 April 2016 (UTC)
But you might not be watching every page in the language you're interested in and want to see all recent changes related to that language. I guess that makes more sense in recent changes than it does in the watchlist. --WikiTiki89 15:10, 4 April 2016 (UTC)
That's true. —CodeCat 15:22, 4 April 2016 (UTC)

Russian Church SlavonicEdit

I asked over at the Grease Pit about creating an etymology-only language for Russian Church Slavonic. I wrote:

I'm pretty sure it should be treated as a dialect of Old Church Slavonic (code cu). Not sure what code to use, maybe cu-ru? (Although most such codes seem to have 3 letters after the hyphen.) While we're at it, we might want similar entries for other Church Slavonic dialects; Wikipedia mentions three: Old Moscow, Croatian and Czech, but all of them in rather limited usage. Benwing2 (talk) 03:53, 3 April 2016 (UTC)

Wikitiki responded:

This is more of a Beer Parlour issue, since the real question is whether we want to have these. User:Ivan Štambuk probably has some opinions about it. --WikiTiki89 04:02, 3 April 2016 (UTC)

Any comments? I'm not sure why it wouldn't be a good idea to have these. They are clearly different dialects from Old Church Slavonic, with words spelled differently, etc. Benwing2 (talk) 04:07, 4 April 2016 (UTC)

(You mentioned there a comparison with Medieval Latin, but we do not have etymology-only languages for "French Medieval Latin" or "English Medieval Latin".) But anyway, if you give me an example of where you intended to use this code, then I can more easily consider whether or not it's a good idea. --WikiTiki89 14:46, 4 April 2016 (UTC)
@Benwing2: In case you missed my previous post, I would like an example of where you intended to use an etymology-only code for Russian Church Slavonic. --WikiTiki89 14:26, 5 April 2016 (UTC)
@Wikitiki89 Thanks, I did miss your post. The comparison with Medieval Latin was meant to be Classical vs. Medieval, not French Medieval vs. English Medieval; but if the issue of Russian vs. Croatian vs. Czech bothers you, then I'd be OK with just Russian. An example of where it would be used is осени́ть (osenítʹ). Vasmer specifically says it's borrowed from Russian Church Slavonic rather than from Old Church Slavonic. Benwing2 (talk) 21:42, 5 April 2016 (UTC)
@Benwing2: I can't find an entry for that in Vasmer at all. Can you link me to where you found it? --WikiTiki89 23:56, 5 April 2016 (UTC)
Hmmm. I actually got it from ru:осенить, which says it comes from Vasmer and another source, but you're right that it's not in the online version of Vasmer. Perhaps it's the other source, or the online version of Vasmer doesn't include everything that was published? Benwing2 (talk) 00:09, 6 April 2016 (UTC)
What you may have thought was another source is actually just a link to a "list of literature". Anyway, assuming the information is correct, if a Russian word is derived from Russian Church Slavonic, that can mean one of two things: either the word was inherited from Old Church Slavonic, but not actually attested in the OCS period, or the word was borrowed into Russian Church Slavonic from Russian itself (with a possible spelling change). In the former case, we can just call it OCS; in the latter case, we can just say that its spelling was influenced by OCS. I don't think late varieties of Church Slavonic were ever used as a written lingua franca, nor do I think much innovation occurred within them, which is why they are much less useful as etymology languages than Medieval Latin. But I would like User:Ivan Štambuk to confirm this, since he knows a lot more than I do about it. --WikiTiki89 00:36, 6 April 2016 (UTC)

Old HindiEdit

I've found a grammatical analysis of Old Hindi, which shows considerable differences from modern Hindi (more cases, less schwa dropping, etc.). Could we make hi-old into a full language, not just etymology-only? The code inc-ohi also works. —Aryamanarora (मुझसे बात करो) 20:26, 4 April 2016 (UTC)

Is it different from Sauraseni Prakrit psu? —Aɴɢʀ (talk) 19:22, 5 April 2016 (UTC)
Sauraseni Prakrit is an earlier stage, I think maybe 1000 years before Old Hindi (although the time period of Prakrit is quite long). Between the two was Sauraseni Apabhramsa. If I'm not mistaken, there was a time around maybe 1100 AD when some people still wrote in Apabhramsa and others in Old Hindi, with significant differences in the case system. Apabhramsa still has the old inherited case system to a large extent while Old Hindi is closer to the modern agglutinative system. For this reason, Old Hindi is put in the "modern" stage while Apabhramsa is in the "middle" stage. (This would suggest that the present system of having Old Hindi be an etymological variant of modern Hindi is consistent with the linguistic consensus.) However, take everything I just said with a grain of salt as I'm going by memory and might be wrong in some particulars. Benwing2 (talk) 21:52, 5 April 2016 (UTC)
You got the gist - Old Hindi is a much more simplified Sauraseni Prakrit. It is different enough to be a different language, with (I think) only five cases (modern Hindi has three). Literature is rare, but the book I've found shows some very old poetry in the language, from the region of Rajasthan. Overall, at least a few words can be added in Old Hindi with adequate citations. —Aryamanarora (मुझसे बात करो) 23:11, 5 April 2016 (UTC)
(Hang on, did you just say Hindi is agglutinative? While it does have plenty of postpositions, it is not agglutinative, at least not like Sanskrit) —Aryamanarora (मुझसे बात करो) 23:13, 5 April 2016 (UTC)
I guess I'm thinking more of other Modern Indo-Aryan languages which have more clearly agglutinative-like case systems, i.e. the plural forms are essentially the same as the singular ones. These are in origin postpositions, and in Hindi you can still analyze them this way and say there are only 3 cases. IMO it's pretty clear that these case systems are heavily influenced by the Dravidian ones. I wouldn't say Sanskrit was exactly agglutinative, though; rather, it let you form long compounds and had extensive derivational morphology of the inflected type. Benwing2 (talk) 23:25, 5 April 2016 (UTC)
BTW having more cases doesn't necessarily mean it has to be a separate language -- Early Middle English, for example, had 4 cases and 3 genders, and Late Middle English had no cases and no genders, but they're clearly dialectal variants of the same language. Benwing2 (talk) 23:27, 5 April 2016 (UTC)
@Aryamanarora You say Rajasthan right? Are you sure those aren't in fact Old Marwadi (if such a term exists), or Old Western Rajasthani (Old Gujarati)? DerekWinters (talk) 05:02, 6 April 2016 (UTC)
@DerekWinters No, it's definitely old Hindi. The verb "to be" has the 3p.s.ind. form hai, 1p.s.ind. hū̃, unlike Gujarati che. —Aryamanarora (मुझसे बात करो) 15:13, 10 April 2016 (UTC)
We shouldn't make up our own codes; if a new code for old Hindi is needed, it should be requested from SIL/ISO 639-3 (if a full language code is believed justified) or from IETF-languages (if a subcode will do.)--Prosfilaes (talk) 00:30, 15 April 2016 (UTC)
Good luck with that. —CodeCat 00:40, 15 April 2016 (UTC)
I know for a fact that you've never tried to get a code from IETF-languages, since I've been on that list for a long time. It's not that hard, provided someone is willing to come up with cites to published descriptions and justify why this is a distinct lect.--Prosfilaes (talk) 07:19, 16 April 2016 (UTC)
That would take at least a few months – I want to add entries now. —Aryamanarora (मुझसे बात करो) 11:41, 15 April 2016 (UTC)
So you want to do a lot of work, but don't care if it's useful or will be preserved? It could be done under a private use tag and switched over once we have a real one. If you're going to do something, do it right, instead doing it right now.--Prosfilaes (talk) 07:19, 16 April 2016 (UTC)
Why would the work be less useful or not be preserved if he uses the inc-ohi code now instead of waiting for a new code? If and when the new code is created, the forms can be moved. We had a lot of entries in Norman varieties using our own local codes before the code nrf was created, and once it was, we moved them to the new code. Nothing was lost. —Aɴɢʀ (talk) 07:44, 16 April 2016 (UTC)
Or we could use a correct code, like inc-x-ohi, that never will get confused with anything else. New codes don't pop into existence; they get created when someone proposes them. I can deal with the bureaucracy of IETF-languages, but I know nothing of Old Hindi, and couldn't possibly come up with a list of published descriptions, and am not the best person to make the argument for it. If we just go on and create Old Hindi entries, the odds this will ever get any sort of standard code will be minimal.--Prosfilaes (talk) 23:35, 19 April 2016 (UTC)

Entry layout for reconstructed termsEdit

Please see some thoughts at Wiktionary talk:Reconstructed terms#Layout proposal. I am considering creating a new draft version of the policy to include considerations such as these at some point. --Tropylium (talk) 09:00, 6 April 2016 (UTC) 

Old East Slavic adjectivesEdit

Right now we have some of these lemmatized at their long forms (e.g. бѣсовьскꙑи (běsovĭskyi) rather than бѣсовьскъ (běsovĭskŭ)), and some of them lemmatized at their short forms (e.g. четвьртъ (četvĭrtŭ) rather than четвьртꙑи (četvĭrtyi)). Is this intentional? I’d think they should be consistently at one or the other, but am I missing something? Vorziblix (talk) 16:04, 7 April 2016 (UTC)

I think they should all be lemmatized at their short forms, unless the short forms are unattested. --WikiTiki89 17:09, 7 April 2016 (UTC)
Why does attestation matter? We lemmatise words in other languages consistently even if the lemma form happens to be not attested. And the lemma form of Slavic adjectives is entirely predictable from any other form. —CodeCat 17:33, 7 April 2016 (UTC)
Because the short forms may have not existed for some adjectives. Compare how the short forms of adjectives with *-ьskъ do not exist in any modern Slavic language. I wouldn't want to lemmatize a short form when it did not exist. --WikiTiki89 17:48, 7 April 2016 (UTC)
We can easily determine from OES grammars which classes of adjectives had only the longer forms. Also consider that OCS did still have short inflections for almost all adjectives. —CodeCat 18:54, 7 April 2016 (UTC)
If that can be determined, then sure. OES ≠ OCS. --WikiTiki89 19:01, 7 April 2016 (UTC)
I would assume so, unless OES attestation is so sparse that nobody has been able to determine it yet. —CodeCat 19:11, 7 April 2016 (UTC)
Short forms of *-ьskъ adjectives do seem to be rare, but even a cursory search turns up at least one instance: »Азъ… хотѣвъ вкоуситі оучительскаго съказаниꙗ, готова прѣложити въ словѣньскь ꙗзꙑкъ.« In any case, I’ll move добрꙑи, тѧжькꙑи, and бородатꙑи, which are each attested in their short forms on the very pages they link to. Vorziblix (talk) 22:03, 7 April 2016 (UTC)


When and why did this stop capitalizing the first letter of the word? Was this discussed? I feel like the link should reflect the spelling in the Wikipedia entry, and thus automatically capitalize the first letter. --WikiTiki89 18:53, 7 April 2016 (UTC)

It was done because not all Wikipedias capitalise all entries. I think it was the Lojban Wikipedia that was having trouble, all the links to it were broken. —CodeCat 18:54, 7 April 2016 (UTC)
So why can't we do that only for the Lojban Wikipedia and whatever few other ones there may be? --WikiTiki89 19:00, 7 April 2016 (UTC)
Ask the person who made the edit and the people who participated in the discussion. —CodeCat 19:12, 7 April 2016 (UTC)
What a review of the Wiktionary:Grease_pit/2015/June "discussion" shows is that, of the three technical adepts pinged 2015, August 5, based on their having contributed to the modules and templates, none responded. DCDuring TALK 20:59, 7 April 2016 (UTC)

Italian pronunciationEdit

In this page you read:

'See Italian phonology at Wikipedia for a thorough look at the sounds of Italian. In addition, the Wikipedia help page for IPA for Italian has some useful English approximations.'

Neither in Italian phonology nor in Help:IPA for Italian you can find any reference to asterisks (*) used in phonetic transcriptions of Italian language to indicate the so called "syntactic gemination", but in Appendix:Italian pronunciation it is specified that the asterisk is the symbol used for that. The reason you can find it in this dictionary and not in the main encyclopedia is that it was inserted without any consensus by an Italian user and nobody has noticed it during these months. This user, IvanScrooge98, did the same on en.wikipedia, and there he was contested for that: user Macrakis created a discussion in Help talk:IPA for Italian#Syntactic gemination asking to remove the asterisks and users Peter238 and Aeusoes1 joined the discussion and agreed with him. These users are expert in phonetic issues and IPA and decided to delete this arbitrary symbol deliberately inserted. You can read the full talk in the link, and I am reporting the main passages here:

  • We should not show the syntactic gemination (SG) symbol (*) in our transcriptions of Italian, except in articles that are specifically about Italian phonology, for three reasons: 1) cases are predictable except for a small closed set of function words; 2) the actual set of words varies by variety; 3) it is not a universal feature of standard spoken Italian.
  • Syntactic gemination is fully predictable — they occur after stressed final syllables, including stressed monosyllables (phonological SG) and after a small, closed set of unstressed monosyllables and some penultimate-stressed polysyllables. This closed set includes only function words and not nouns, which are the typical words for which we give pronunciations.
  • The actual set of exceptional words varies by variety of Italian, even among those varieties which show SG. But in any case, it does not include nouns, if I'm not mistaken.
  • Many varieties of standard Italian spoken outside central Italy do not show SG at all.
  • Also, the only time the "*" annotation is useful is if the word is being composed with another word, e.g., "il Po superiore", which is presumably [il'possuperi'ore] rather than [il'po.superi'ore]. But if you know enough Italian to compose words like this, you don't need the "*".
  • I agree that there's no need to transcribe SG when the last syllable is stressed, but I'm not sure about the rest.
  • Why couldn't we just show the actual gemination if it occurs in our transcription? It would be like somehow marking French transcriptions with the final consonant that is normally elided except in cases of liaison.
  • In looking back over the discussion, I see that I wasn't specific enough about the case in question. I didn't intend to discuss the transcription of syntactic gemination in the interior of phrases. That is, I see no problem with [kafˈfɛ lˈlatte]. I was concerned with cases like the Po (river) article, where the name of the river is transcribed as [pɔ*]. This is parallel to transcribing Alaska (and every other noun ending in 'a') as [əˈlæskə(r)] because in sentences like "Alaska is big", RP speakers say [əˈlæskər ɪz 'bɪg], which I hope no one is proposing.

Consider, please, above all the last 2 points. It is completely useless, not to say counterproductive, to add an asterisk to show something that, if the asterisk is used, does not happen at all. When it happens, in a sequence of words, a consonant is doubled or it is used the ː symbol as for vowels. Moreover, the International Phonetic Alphabet (which is supposed to be the conventional alphabet used in this dictionary to transcribe the pronunciation of words and names in any language) has never, ever, employed asterisks for syntactic gemination, in any language. They are used sometimes for consonants pronounced fortis in Korean language and maybe in other cases, so using it for Italian words (either stressed on the final vowel, or monosyllabic, or a half dozen of polysyllabic words) would just confuse readers, as it happened to Macrakis on en.wikipedia, and even here to other users. Why should we keep a totally subjective convention introduced without asking anyone and just at one user's will? I frankly find it absurd to have to open a new discussion to ask for permission to remove something introduced without consulting anyone and already crossed out from en.wikipedia, but I am following the rules and probably also the 3 users who have already discussed about it in the talk page I have linked above may join too. I entrust your common sense about your final decision. 21:34, 7 April 2016 (UTC)

i consider this argumentation reasonable Phfnhyn (talk) 23:55, 8 April 2016 (UTC)

The IP editor above has correctly quoted my arguments against including the * symbol for "syntactic gemination". This is certainly an interesting phonological phenomenon, but is not tied to individual words (except for a small number of function words). I see no point in including * in the phonetic transcription of Italian words on Wikipedia or in Wiktionary. Even for the small closed class of function words, it is not universal in Italian, and really belongs in a grammar or phonological description, not in a dictionary. --Macrakis (talk) 18:04, 9 April 2016 (UTC)

I don't know much about it, but I agree using the asterisk to mark places where syntactic gemination may occur is kind of silly. If its triggers are not predictable in all cases, then maybe it would be better to mark certain words as triggering it, the way certain Irish words (e.g. go, i, bhur, etc.) are marked as triggering certain initial mutations. —Aɴɢʀ (talk) 19:56, 9 April 2016 (UTC)
Silly is the word! As quoted above, it's the same as marking French words ending with "d p s t x z", because when followed by a word starting with a vowel such consonants are read, unlinke if the words are pronounced alone or followed by consonant: the logical solution is just writing the consonant when it's actually pronounced in the 1st case. E.G.: "ils" (they) is pronounced [il], while "ils ont" (they have) is pronounced [ilz ɔ̃] and "ils ont un" (they have a) [ilz ɔ̃t œ̃]... 20:39, 9 April 2016 (UTC)
Actually, writing French liaison consonants has a lot more merit, because it is unpredictable and has to be memorised for each word. —CodeCat 20:45, 9 April 2016 (UTC)
Yes, maybe I shouldn't have switched language... Italian words with syntactic gemination are fully predictable, and in the page we're talking about there're links to specific articles about Italian phonology where this phenomenon is explained. It'd make a lot more sense marking French words subject to liaison than Italian words subject to syntactic gemination, and here French word are marked with nothing when they're subject to liaison! 20:59, 9 April 2016 (UTC)

I didn’t notice this practice until quite recently. I’ve never seen any dictionary or encyclopaedia inserting asterisks into their pronunciative instructions, for Italian or otherwise. I’d be quite fine with them being subtracted. --Romanophile (contributions) 08:28, 11 April 2016 (UTC)

@Angr I would like this discussion to have aroused more interest, but after 4 days only a few people have replied, none of them liking the use of the asterisk notation in question, though; I cannot say whether this is just not an important matter for the community or the community prefers truly to take the asterisks away, but in both cases, and since the only ones who wrote down their opinion expressed themselves against this convention, I wonder if this is not enough to start considering their removal. 17:30, 11 April 2016 (UTC)

I'd give it a week, but I agree it looks unlikely that the asterisks will be kept. —Aɴɢʀ (talk) 17:44, 11 April 2016 (UTC)
Disclaimer: I am not for removing information just because it is predictable. Many languages allow pronunciation to be predicted from spelling or inflection from verb class, yet get both a pronunciation section and a declension table. But since this case is apparently an idiosyncratic practice marking a non-general prosodic side effect, I'm for removing it. Korn [kʰũːɘ̃n] (talk) 08:05, 12 April 2016 (UTC)

@Angr I see that another use has commented negatively the asterisk notation, tomorrow it will be a week and probably we shall be able to take a decision, if you agree. 18:48, 13 April 2016 (UTC)

Now that a discussion has been held, I have no objection. I was never in favor of the asterisk, I was only opposed to a large-scale change in our representation of Italian pronunciation without discussion first. —Aɴɢʀ (talk) 09:25, 14 April 2016 (UTC)
O.K. then, thank you! Later I'll fix the asterisks. 09:33, 14 April 2016 (UTC)

New policy: Reconstructions must have descendants or derived termsEdit

Reconstructions are always based on descendant forms which point to their existence. I have already been following this rule as an unofficial policy myself, making sure I can find descendants before creating reconstructed entries. I wonder if we can make this an official policy, to be added to WT:RECONS? —CodeCat 19:12, 8 April 2016 (UTC)

Is there a problem that reconstruction entries are actually being created without descendants, or is this purely a hypothetical issue? --WikiTiki89 19:34, 8 April 2016 (UTC)
There have certainly been a fair few yes. I just deleted them all in the past, but we had no policy stating that it was grounds for deletion. I'd feel better if I could point to a policy when deleting. —CodeCat 19:47, 8 April 2016 (UTC)
I see. The bug is in the {{policy}} template, which seems to imply that a "guideline or common practices page" "must not be modified without a VOTE". We're allowed to enforce guidelines and common practice, so if we were able to modify the page without a BP discussion rather than a vote, that would be the ideal solution. --WikiTiki89 20:09, 8 April 2016 (UTC)
I'm actually asking if people agree though, I don't know if this is a common practice. It's just a practice I've applied, myself. —CodeCat 20:11, 8 April 2016 (UTC)
Ok. I agree. I also think this has been discussed before. --WikiTiki89 20:20, 8 April 2016 (UTC)
The loophole is that you can add {{policy}} without a vote but then any subsequent changes, such as reverting that edit, do need a vote. That's why I think sometimes {{policy}} shouldn't be taken as gospel. Just make the edit. Renard Migrant (talk) 22:47, 8 April 2016 (UTC)
  • I strongly agree. I think it's worth adding to the page. —Μετάknowledgediscuss/deeds 20:16, 8 April 2016 (UTC)
  • If it isn't a problem, it doesn't need a solution. —Aɴɢʀ (talk) 22:45, 8 April 2016 (UTC)
    • Do you think reconstructions without descendants or derived terms are fine, then? —CodeCat 22:48, 8 April 2016 (UTC)
      • No, I think common sense doesn't need to be regulated. —Aɴɢʀ (talk) 09:44, 9 April 2016 (UTC)
  • Not sure what I think. I feel like I need some context. What are some examples of entries that someone created but you deleted? And in what situation and for what purpose would a scholar (assuming the person used scholarly sources and didn't just make things up) reconstruct forms that have no descendants? — Eru·tuon 00:07, 9 April 2016 (UTC)
  • Definitely. —JohnC5 05:24, 9 April 2016 (UTC)
Whether we officialise it with a policy or not I am in support of the proposal. I don't think we have an issue now with it, but creating a policy now will certainly make it easier to deal with should a problem ever arise in future. Leasnam (talk) 21:02, 9 April 2016 (UTC)
  • Hmm. Corner case: suppose we know that in subfamily X of family Y, a base vocabulary item — let's say 'water' — has been replaced by loanwords. However, the loaning has taken place later than proto-X (which might be inferrable e.g. due to external history, due to the loanwords being from numerous distinct sources, or from internal chronology of sound changes). Would it be then legitimate to reconstruct the inherited Proto-X form for 'water', despite its later extinction, e.g. for the purposes of rounding out a Swadesh list for Proto-X? --Tropylium (talk) 16:00, 22 April 2016 (UTC)
    Another corner case that comes to mind: it's occasionally possible to find a reconstruction referenced in a particular source, but without any actual descendants listed (e.g. due to the reconstruction being compared to data from a related branch). I suppose that in such cases, it should still be OK to create the reconstruction on the basis of the source, and worry about looking up the actual descendants later. --Tropylium (talk) 16:06, 22 April 2016 (UTC)
    In such cases, I refrain from creating the entry. —CodeCat 16:12, 22 April 2016 (UTC)
    For your first corner case, no. You don't know that the language inherited the native word. It could have fallen out of use before the borrowing. Perhaps there was another intermediate borrowing, or perhaps people stopped talking about water, who knows. For your second corner case, I would say that you can probably create the entry with the intent to add descendants later, but once it is clear that no such descendants can be found, then you would have to delete it. --WikiTiki89 16:15, 22 April 2016 (UTC)

Black's Law Dictionary 10th Edition (2014)Edit

If anyone has access to this book, could they check that the edits of (talk) are not coyvios please. They look like they might be word-for-word copies SemperBlotto (talk) 11:45, 9 April 2016 (UTC)

@BD2412, do you have a copy? - -sche (discuss) 22:02, 9 April 2016 (UTC)
Not of the 10th edition, I'm afraid. I'm stuck with the one I picked up in law school - the eighth. By the way, if anyone is interested, Wikisource has the second edition in progress. bd2412 T 18:39, 11 April 2016 (UTC)
The user is not doing a great job either, e.g. default en-noun adding an English -s to Latin terms, even ones that end with -s (and the others rarely take an English plural anyhow). Equinox 12:50, 12 April 2016 (UTC)

Russian -стрелить, -скочить, -прячь, etc. -- suffixes, verbs, roots, or what?Edit

I've been creating entries for verbal roots like -стрелить, -скочить, -прячь, for convenience in creating etymologies and listing derived verbs. These are cases where there are prefixed verbs, e.g. застрелить, напрячь, and logically the base verb should be attested but it isn't. What part of speech should they be labeled as? I've been using "suffix" but don't feel totally comfortable with this, maybe it should be "verb" instead? Benwing2 (talk) 23:08, 9 April 2016 (UTC)

I've listed morphemes like -fico, -φρων (-phrōn), -γραφία (-graphía) as combining forms. That might be appropriate. I kind of wish categories and {{affix}} were able to recognize the category, though. It currently treats them as suffixes, so σώφρων (sṓphrōn) is placed in Category:Ancient Greek words suffixed with -φρων. — Eru·tuon 23:36, 9 April 2016 (UTC)

Appearance of templates in tempEdit

Right now, templates referenced using {{temp}}, like {{en-noun}}, have a border around the reference and gray background. This was not so recently, AFAIR. I think it's pretty ugly. Is this a change made via MediaWiki software update or did someone edit some Wiktionary global css? I see that {{temp}} uses the <code> element, so I can achieve the same appearance by using code directly, like this. By contrast, tt element does not do this, like this, and has the reasonably plain appearance that I prefer. --Dan Polansky (talk) 07:39, 10 April 2016 (UTC)

<tt> got 'deprecated', whatever that means. I mean if it's deprecated why does it still work? Anyway have you considered just getting used to it? Humans are pretty adaptable after all. Renard Migrant (talk) 15:59, 11 April 2016 (UTC)
Deprecated means that even though it may still work, it may be removed in the future. Now I have no idea why it's deprecated or who deprecated it. As for {{temp}}, I have come to like to the surrounding box thing. --WikiTiki89 16:47, 11 April 2016 (UTC)
It's deprecated by the HTML standard. —CodeCat 16:57, 11 April 2016 (UTC)
I am indifferent regarding the box, but we can easily create a css class which emulates the tt tag so the deprecation is moot (other than that we should not use it). - TheDaveRoss 17:55, 11 April 2016 (UTC)
  • I like the appearance of {{temp}}. --Dixtosa (talk) 17:10, 11 April 2016 (UTC)
  • Like Wikitiki, I've come to like the surrounding box. - -sche (discuss) 05:45, 22 April 2016 (UTC)

self‐proposal for adminshipEdit

I think that we need more administrators here, so I’d like to suggest adminship for myself. This has been a long time coming, but I’m seeing a lot of crap popping up and I just don’t have the utilities to deal with it by myself. Considering my trustworthiness, I will confess that I have inserted jocular content into the mainspace a few times, but I did have the decency to alert other users about it later (or just fix it myself), so I don’t think that any jokes that I made are still rotting in a lexical entry somewhere. I am a bureaucrat over at Wikcionario, and you simply won’t find any silly jokes there, no matter how arduously you try. The users there seem quite satisfied with my position, and my (few) errors there aren’t particularly outrageous. @Peter Bowman could give his opinion if he’d like. --Romanophile (contributions) 06:16, 11 April 2016 (UTC)

Coming from a returning user who’s infamous for goofing around, that doesn’t surprise me. --Romanophile (contributions) 08:19, 11 April 2016 (UTC)
I wouldn't support you probably because you swear too much, and seem to get stroppy when things don't go your way. --F909fef0j (talk) 08:36, 11 April 2016 (UTC)
In recent times, I don’t think that I really have a tendency towards obstinance; I hate being called stubborn. Also Dick Laurent (talkcontribs) is probably a worse coprolaliac than I am. --Romanophile (contributions) 09:39, 11 April 2016 (UTC)
The real reason you won't be getting support from him is because he is ineligible to vote. And he's already gotten himself blocked, but chances are he'll be back. --WikiTiki89 18:59, 12 April 2016 (UTC)
  1.   Support I think that we need more administrators here, too. JackPotte (talk) 13:52, 11 April 2016 (UTC)
  1.   Support I’m surprised this isn’t already the case. Vorziblix (talk) 18:44, 12 April 2016 (UTC)
  1.   Support I actually disagree; we have too many admins. However, Romanophile would be a good admin, so I can't vote against him. --WikiTiki89 18:59, 12 April 2016 (UTC)
  1.   Support I was also surprised that you weren't already. --Robbie SWE (talk) 19:05, 12 April 2016 (UTC)
  2.   Abstain You'll be relieved to know that I won't be seeking adminship. Donnanz (talk) 19:22, 12 April 2016 (UTC)
  3. (comment) That's correct, Romanophile is a serious and solid colleague on eswiktionary. Regards, Peter Bowman (talk) 19:56, 12 April 2016 (UTC)
  4.   Support. I too thought you were already an Admin. Leasnam (talk) 20:35, 12 April 2016 (UTC)
  5. (comment) I still think you are an admin. --Dixtosa (talk) 21:01, 12 April 2016 (UTC)
All right, the chances look reasonable. Now I just have to figure out how to set up an official election, or wait for somebody else to do it for me. --Romanophile (contributions) 00:52, 13 April 2016 (UTC)
Here: Wiktionary:Votes/sy-2016-04/User:Romanophile for admin. --Romanophile (contributions) 03:36, 13 April 2016 (UTC)


Should we put our plateau entries under the category of "Mountains" or "Plateaus"? E.g. Mexican Plateau, Armenian Highland, Tibetan Plateau, Loess Plateau, Mongolian Plateau, etc. ---> Tooironic (talk) 10:25, 13 April 2016 (UTC)

@Tooironic: Plateaux. — I.S.M.E.T.A. 15:28, 14 April 2016 (UTC)
No, the more common English plural is preferable; there's no need for us to be pretentioux. - -sche (discuss) 05:42, 22 April 2016 (UTC)
@-sche: Plateau is also a verb, so that simple Ngram doesn't prove anything. And plateaux is hardly pretentious. — I.S.M.E.T.A. 15:56, 5 May 2016 (UTC)
Even the OED shows plateaux as a secondary (British) spelling, the primary being plateaus. DCDuring TALK 16:02, 5 May 2016 (UTC)

Proto language translationsEdit

It was my understanding that proto languages were not to be included in translations. See my edit on gold that were reverted by @I'm so meta even this acronym. Is this an official policy or not? DTLHS (talk) 15:06, 13 April 2016 (UTC)

FYI, here are: my request, DTLHS’s reversion, and my reversion. — I.S.M.E.T.A. 15:13, 13 April 2016 (UTC)

Yes. The official policy is that words in proto-languages are not to be included in the main namespace other than in etymologies. --WikiTiki89 15:10, 13 April 2016 (UTC)
@DTLHS, Wikitiki89: It seems legitimate that proto-languages' translations be included in those tables. It's information people might well want, and how else would they find it? In this case, I was trying to find out what the indigenous Celtic word for "gold" was, since all the Celtic languages' terms I could find on here derive from the Latin aurum. — I.S.M.E.T.A. 15:16, 13 April 2016 (UTC)
No, we have consistently kept the standard that only languages which are in mainspace can be added as translations, so no Proto-Celtic or Klingon or what have you. There's a lot of information some people might hypothetically want, but that doesn't mean it's appropriate for us to include. —Μετάknowledgediscuss/deeds 15:22, 13 April 2016 (UTC)
(e/c) We don't consider reconstructed words to be real words. That's one reason. As for your specific example, if no Celtic language has an indigenous word for gold, then there would be no basis on which to reconstruct one. --WikiTiki89 15:25, 13 April 2016 (UTC)
What Μετάknowledge said. - -sche (discuss) 15:56, 13 April 2016 (UTC)

*Humph* Fine. Does anyone know of any word in any of the Celtic languages that means "gold" and which doesn't derive from the Latin aurum? There's a fair bit of gold in Wales, so I'd assumed that there would exist such a word. — I.S.M.E.T.A. 15:30, 14 April 2016 (

I'm unaware of any Celtic word for "gold" besides loanwords from aurum. Maybe in Gaulish or Celtiberian, but I don't think there's anything in Insular Celtic. The gap isn't especially surprising, though; there are all sorts of semantic gaps in proto-languages where we might not expect them. For example, there's no reconstructable PIE word for "rain", despite the fact that PIE speakers must have been aware of and had a word for it. —Aɴɢʀ (talk) 21:30, 19 April 2016 (UTC)
@Angr: How frustrating. Thanks anyway. — I.S.M.E.T.A. 15:50, 5 May 2016 (UTC)
Hmm. There seems to have been a PIE verb, at least. Gamkrelidze and Ivanov's Indo-European and the Indo-Europeans says that although taboo replacement led to a "low number of attested Indo-European languages preserving *seu-/*su- in the sense 'rain' [...] the ancient root is preserved in its original sense" of "to rain" in Greek, Albanian, Old Prussian (suge in the Elbing vocabulary), and Tocharian (su- per Adams' Dictionary of Tocharian B), and Reconstruction:Proto-Germanic/sūpaną mentions it.
What would be the expected Celtic reflex of PIE aus- / h₂é-h₂us-o-?
- -sche (discuss) 05:33, 22 April 2016 (UTC)
@-sche Probably Old Irish áu (attested in the meaning 'ear' from *h₂ṓws) and Middle Welsh eu/Modern Welsh au. —Aɴɢʀ (talk) 18:33, 22 April 2016 (UTC)

This is a bit insensitive to visitors - if we have so many reconstructed terms, they ought to be linked in translation tables. —Aryamanarora (मुझसे बात करो) 20:46, 19 April 2016 (UTC)

@Aryamanarora: Yes, that's my thinking. — I.S.M.E.T.A. 15:50, 5 May 2016 (UTC)
No, they oughtn't. They ought to be linked in etymology sections of real words. —Aɴɢʀ (talk) 21:16, 19 April 2016 (UTC)
Why not both? Semantic shifts occur too, you know. Theoretically, the page "fire" would link to *péh₂wr̥ in its etymology section, but have both *péh₂wr̥ and *h₁n̥gʷnis in its translation table. —Aryamanarora (मुझसे बात करो) 20:17, 5 May 2016 (UTC)

How to render "someone" in lemma phrases in Russian?Edit

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter I wanted to create an entry for the expression ободра́ть кого́-нибудь как ли́пку meaning "to rob someone blind" (literally "to strip someone like a young linden tree"). We have the corresponding English expression under "rob someone blind" with the word "someone" embedded into it. The only issue is that there are three more-or-less synonymous words for "someone" in Russian: кто-нибудь, кто-либо and кто-то. I've been using кто-нибудь in usage examples; Wanjuscha prefers кто-либо. My primary dictionary uses кто-нибудь but Zaliznyak and some others use кто-либо (often abbreviated кто-л or кто-л.). I just put the expression under ободрать как липку (to "rob blind", leaving out the "someone"), but the English analogy suggests that we should include "someone" (there's no rob blind). What should be done? Benwing2 (talk) 18:48, 13 April 2016 (UTC)

One thing to be considered is that in English we add these words because word order is important and there must be an object in between "rob" and "blind", otherwise it doesn't make sense. In Russian, since word order is flexible, we can just create the entry at ободра́ть как ли́пку (obodrátʹ kak lípku). The reason many dictionaries include the word for "someone" is in order to indicate which case it should be in, but we can do that just as easily with tags and/or usage examples in the entry. --WikiTiki89 18:54, 13 April 2016 (UTC)
I think "rob blind" makes perfect sense as a lemma. It can even be used in real sentences that way. —CodeCat 18:57, 13 April 2016 (UTC)
Actually, in this particular case, I would also prefer "rob blind", although this is a dictionary-only form, since in the real world there is always an object in between. But there are many cases where the "someone" is required in the lemma. --WikiTiki89 19:02, 13 April 2016 (UTC)
Not always. "They robbed blind the very people who were trying to help them." —CodeCat 19:04, 13 April 2016 (UTC)
You're right. I definitely support moving rob someone blind to rob blind. --WikiTiki89 19:07, 13 April 2016 (UTC)
To expand on what Wikitiki said, you don't need to represent or add the word "someone" in the Russian entry titles. It's fine for a Russian entry with no word that corresponds to "someone" to link to an English entry containing "someone",; you can pipe the link if you think it would be misleading otherwise: [[rob someone blind|rob blind]]. - -sche (discuss) 01:57, 21 April 2016 (UTC)

ist der Ruf erst ruiniert, lebt es sich ganz ungeniertEdit

What is our policy on (excessive?) referencing like in this one and also in some other edits by the user:Caligari, who is an admin from the German wiktionary. I don't want to denunciate, I'm just asking because the entry doesn't look like our entries usually look. And I'm relatively sure that we don't want our lemmas to look like the German entry for Haus, which lists 57 sources and an additional 15 references. Kolmiel (talk) 23:53, 14 April 2016 (UTC)

@Kolmiel: The entry for ist der Ruf erst ruiniert, lebt es sich ganz ungeniert looks fine to me. — I.S.M.E.T.A. 12:10, 15 April 2016 (UTC)
I also think that the Wikiwörterbuch's entry looks fine, and that the amount of sourcing that appears therein is proportionate to the amount of substantive content in the entry. — I.S.M.E.T.A. 12:13, 15 April 2016 (UTC)
I agree, but I also agree that the reference material for de:Haus is excessive and I don't want our entries to look like that either. —Aɴɢʀ (talk) 12:19, 15 April 2016 (UTC)
@Angr: I'm not a fan of the Wikiwörterbuch entry layout. I've yet to see how all that reference would look if converted to our format. I expect we'd probably stick it in a collapsible table if there was that much of it. — I.S.M.E.T.A. 12:36, 15 April 2016 (UTC)
If we had many entries with that amount of reference material, someone would doubtless propose a new References tab next to the Citations tab. —Aɴɢʀ (talk) 12:42, 15 April 2016 (UTC)
@Angr: Since Wikipedia hasn't done that already (and I've seen some articles with hundreds of citations), I think it's very unlikely that we'd do so. One of the main justifications of the Citations: was that it would give a place to house citations for non–CFI-satisfying terms; I see no analogical use for a References: tab. — I.S.M.E.T.A. 15:40, 15 April 2016 (UTC)
I didn't say it was likely we would do it, merely that it was likely someone would propose it. It would probably also be really difficult to get <ref> tags to generate text in a different namespace on a different tab. —Aɴɢʀ (talk) 15:43, 15 April 2016 (UTC)
@Angr: I'm afraid I wouldn't know about that. — I.S.M.E.T.A. 15:47, 15 April 2016 (UTC)

This entry seems fine to me. My main issue―and I hate to be a schoolmarm about this―is that Caligari's seems to be in violation of WT:USER#User pages. Furthermore, I can barely read the talkpage. This may just be silly of me and irrelevant, but I thought I'd mention. —JohnC5 14:47, 15 April 2016 (UTC)

@JohnC5: The user page isn't too offensive, even if it does violate WT:USER#User pages, but the talk page's background colour is nauseating. — I.S.M.E.T.A. 15:40, 15 April 2016 (UTC)


I have created a vote with the intent that we can finally decide on a new logo for the English Wiktionary. I welcome all discussion on how to make this vote as likely as possible to be successful at representing the consensus of Wiktionary editors, so please feel free to edit the vote if you can improve it. —Μετάknowledgediscuss/deeds 03:25, 16 April 2016 (UTC)

Thank you! I've always been bothered by the current logo. Benwing2 (talk) 05:21, 16 April 2016 (UTC)
I've always been bothered by the RP pronunciation. It doesn't seem right to endorse British, American, or any other English. We could change the pronunciation to enPR or just omitting it altogether in favo(u)r of something more universally accepted and comprehensible like hyphenation (Wik‧tion‧ar‧y); though on second thought, there might be some debate about that too. It's also strange that the headword format in the logo is not one we use (viz. we don't use PoS abbreviations). —JohnC5 05:43, 16 April 2016 (UTC)
We already did this. Why are we doing it again? --Yair rand (talk) 19:31, 21 April 2016 (UTC)
We haven't done "this", whatever you mean by that, for quite a while, and there is great dissatisfaction with the current logo (just look at the comments above yours). —Μετάknowledgediscuss/deeds 14:18, 26 April 2016 (UTC)
  • The vote has begun! Please vote — I would really like to get maximal turnout so we can find a logo that's broadly acceptable to the community. —Μετάknowledgediscuss/deeds 14:18, 26 April 2016 (UTC)
    Why not give the link in the News for editors top part of the site? Or, *gasp*, at the same place as the News for editors link? — Dakdada 10:46, 27 April 2016 (UTC)

Server switch 2016Edit

The Wikimedia Foundation will be testing its newest data center in Dallas. This will make sure Wikipedia and the other Wikimedia wikis can stay online even after a disaster. To make sure everything is working, the Wikimedia Technology department needs to conduct a planned test. This test will show whether they can reliably switch from one data center to the other. It requires many teams to prepare for the test and to be available to fix any unexpected problems.

They will switch all traffic to the new data center on Tuesday, 19 April.
On Thursday, 21 April, they will switch back to the primary data center.

Unfortunately, because of some limitations in MediaWiki, all editing must stop during those two switches. We apologize for this disruption, and we are working to minimize it in the future.

You will be able to read, but not edit, all wikis for a short period of time.

  • You will not be able to edit for approximately 15 to 30 minutes on Tuesday, 19 April and Thursday, 21 April, starting at 14:00 UTC (15:00 BST, 16:00 CEST, 10:00 EDT, 07:00 PDT).

If you try to edit or save during these times, you will see an error message. We hope that no edits will be lost during these minutes, but we can't guarantee it. If you see the error message, then please wait until everything is back to normal. Then you should be able to save your edit. But, we recommend that you make a copy of your changes first, just in case.

Other effects:

  • Background jobs will be slower and some may be dropped.

Red links might not be updated as quickly as normal. If you create an article that is already linked somewhere else, the link will stay red longer than usual. Some long-running scripts will have to be stopped.

  • There will be a code freeze for the week of 18 April.

No non-essential code deployments will take place.

This test was originally planned to take place on March 22. April 19th and 21st are the new dates. You can read the schedule at wikitech.wikimedia.org. They will post any changes on that schedule. There will be more notifications about this. Please share this information with your community. /User:Whatamidoing (WMF) (talk) 21:07, 17 April 2016 (UTC)

Proposal to globally ban WayneRay from WikimediaEdit

Per Wikimedia's Global bans policy, I'm alerting all communities in which WayneRay participated in that there's a proposal to globally ban his account from all of Wikimedia. Members of the Wiktionary community are welcome in participate in the discussion. --Michaeldsuarez (talk) 14:53, 18 April 2016 (UTC)

That user has only two edits to English Wiktionary, both of which have been deleted. —Aɴɢʀ (talk) 14:55, 18 April 2016 (UTC)
And one of those was to a transwiki that was later deleted, while it was still at Wikipedia. Their only real edit here was creation of a user page that was deleted an hour and a half later as "promotional material". Definitely not a participant here. Chuck Entz (talk) 01:27, 19 April 2016 (UTC)

Cascading protection of the main pageEdit

The main page has cascading protection. This prevents non-admins from editing Words of the Day on the day the word is featured, while allowing them to usually edit and set Words of the Day, which is good and is AFAICT the intention behind the cascading protection. However, it has some negative effects, too: it blocks people from editing certain modules, such as Module:labels/data, when labels appear on the main page. (See discussion.) Should we remove cascading protection from the main page? Alternatively, I think labels used to be "converted" by hand to (''text'') before being plugged into the WOTD templates, whereas that doesn't happen anymore... we could make an effort to convert the labels on all the WsOTD that have been set. - -sche (discuss) 00:48, 20 April 2016 (UTC)

I think labels should be converted to {{qualifier}} for WOTDs. --WikiTiki89 16:04, 20 April 2016 (UTC)
Or even a qualifier template used only on the Main Page! Renard Migrant (talk) 21:58, 20 April 2016 (UTC)
Better yet, they should be substed. Chuck Entz (talk) 01:32, 21 April 2016 (UTC)
I'm not sure that would work. Substing {{lb|en|chemistry}}, for example, yields {{#invoke:labels/templates|show}}. —Aɴɢʀ (talk) 13:51, 21 April 2016 (UTC)
The template code can be tweaked to make substing possible, but I don't think that's a good solution, since it would add too much unnecessary mark-up. --WikiTiki89 15:01, 21 April 2016 (UTC)
Modules can be substituted, and there is also the Lua function mw.isSubsting() which a module can use to return different results depending on whether it's substed or not. —CodeCat 15:45, 21 April 2016 (UTC)
Yes they can be, but they won't be unless the template's code is changed to something like {{<includeonly>safesubst:</includeonly>#invoke:labels/templates|show}}. --WikiTiki89 15:57, 21 April 2016 (UTC)

Let's kill nds-de/nds-nl.Edit

I've always been hesitant to bring it up because it's not really a high priority issue and since I don't know how to bot, I would be either heaping the work on somebody else or have to work through 500 pages manually. Also I was partially at fault for the adoption of these tags in my early days on Wiktionary and learning about Low German. But being more acquainted with both, I don't see any justification for this split, especially considering that "Low German" covers the last 400 years and not only post WWII. The only actual distinction that runs sharply on the border is orthographic tradition, though you can find some Dutch traditions in Western Germany as well, e.g. "zy" for [zɛɪ] in Eastern Frisia around 1890, which is when modern Low German had its greatest international public attention.
For reference: The distinction between nds-nl and nds-de arose on nds.Wikipedia because they couldn't settle on a way to spell things, but that is an issue we as a descriptive dictionary don't face. We currently have 99 nds-NL lemmas, 405 nds-DE lemmas and 548 nds-only lemmas. Opinions, commentary? Korn [kʰũːɘ̃n] (talk) 10:23, 21 April 2016 (UTC)

I'm fine with merging the codes as long as we have regional labels for them allowing the categories Category:Dutch Low German and Category:German Low German to be retained as regional varieties of Low German. We could even have separate categories for all of the subdialects of Dutch Low German that have their own ISO codes. —Aɴɢʀ (talk) 13:49, 21 April 2016 (UTC)
I know nothing or almost nothing about the language, however based on what I've said on Wiktionary this seems like a good idea. The Norman Wikipedia I know has similar issues with spelling because Jersey and Guernsey spellings are so different. Renard Migrant (talk) 20:08, 21 April 2016 (UTC)
I've had problems deciding which template to use, and now I just use the nds template. With Norwegian words it's hard to tell whether they're of Low German or Middle Low German origin, so I refer to Den Danske Ordbog as well, where they usually differentiate between nedertysk and middelnedertysk. What language came before Middle Low German, by the way? Donnanz (talk) 20:22, 21 April 2016 (UTC)
Old Saxon (osx). —Aɴɢʀ (talk) 21:36, 21 April 2016 (UTC)
I think we should merge these languages only if Norwegian is also merged, it's more or less the same issue. —CodeCat 21:41, 21 April 2016 (UTC)
Ooh, don't start that one up again. What about merging Scots with English?
@Angr: I thought so, cheers. Donnanz (talk) 21:53, 21 April 2016 (UTC)
I'm definitely opposed to making this merger dependent on some unrelated merger. —Aɴɢʀ (talk) 22:08, 21 April 2016 (UTC)
If there is only around 1000 nds entries in all we have only scratched the surface. I have come across many more without entries when entering etymology, and the same goes for gml. Donnanz (talk) 22:31, 21 April 2016 (UTC)
While the situation within Low German might be similar to that of Norwegian, this is not bound to national borders, which is what we wrongly imply. Westphalian dialects exist in both Germany and the Netherlands, they are very similar to each other and markedly different to the other dialects within the same nation. East Frisian and Twents on the other hand are two dialects which have the political border running right through them without too great an impact. Sometimes not even on orthography, as the zy-example shows. The dialects within Germany might agree roughly on which glyphs not to use. But since they have vastly different phonetic inventories, and are often influenced by different regional traditions, they do not employ identical spellings either, even if they agree on the pronunciation. Further they do not agree on declension patterns, number and form of articles and pronouns, and number of grammatical cases. So while the situation within Low German might be like that of Nynorsk/Bokmål/Rigsdansk, that is simply not the issue nds-de and nds-nl are dealing with. What they are dealing with is spelling. They would separate entries with the same pronunciation, meaning and grammar into separate L2 headers purely based on spelling. And this is what I ask to do away with. Korn [kʰũːɘ̃n] (talk) 10:48, 22 April 2016 (UTC)
I know that we are descriptive, but we must always be forward facing. What does the future hold for these two varieties? Will they trend towards greater divergence in future or no? I know, we can cross that bridge when we get to it, but why wait? Planning today for the future is never foolish. Leasnam (talk) 19:11, 22 April 2016 (UTC)
I don't know about Dutch Low German, but I suspect German Low German will become extinct before it has a chance to either converge with or diverge from Dutch Low German. Trying to find a fluent native speaker of German Low German below the age of 50 is like trying to find a hay-colored needle in a hay stack. —Aɴɢʀ (talk) 19:19, 22 April 2016 (UTC)
Hmm, good point. Then what's left will be just the one variety. But (hopefully) like Bavarian and Swiss German, there is always hope :) Leasnam (talk) 19:25, 22 April 2016 (UTC)
I don't see why being forward facing would include making a random guess at the outcome of the next few decades of development. Whatever their result, they won't undo the last centuries either. And if they change that much, they probably deserve their own L2 as a new phase of the language. My point was that it seems only sensible to me that either we separate the variants based on how different they are, or group them all as one. If we split them up, German Low German has to be parted into the actual dialects - which range from something between 4 and 40, depending on what you split by. If we're not going to split up Low German into its actual dialects, I don't see why we would create the singular case that groups of spellings would get a separate language tag here. (My understanding is that Nynorsk also differs in grammar from Bokmål, but, again, "both" Low Germans already differ in grammar and spelling from themselves plenty.) Korn [kʰũːɘ̃n] (talk) 20:07, 22 April 2016 (UTC)
No one in their right mind would suggest a random guess. But we can gauge based on the past 100 years or so whether they are becoming more alike, remaining static, or diverging. Leasnam (talk) 21:16, 22 April 2016 (UTC)
Maybe there's a parallel with Flemish and Swiss German which aren't regarded as separate languages here, despite the fact they are spoken in countries other than the home country of the original language. Donnanz (talk) 13:45, 25 April 2016 (UTC)
This comparison seems correct to me. I do not see how a divergence of dialects in the future justifies the split as we have it (for less than half of the entries), though. The future is neither the present nor the past, which is what we record. And I don't see why a one difference should get special treatment over all other differences which are not tied to a political border. Be as it may, since there doesn't seem to be a hard opposition, I'll just use the plain NDS-tag in etymology and descendence sections, especially since it's the major one anyway. Korn [kʰũːɘ̃n] (talk) 15:06, 25 April 2016 (UTC)
We are a dictionary, not a crystal ball. Predicting the future is not our business and should not drive any of our decisions. --WikiTiki89 15:10, 25 April 2016 (UTC)
When Wiktionary merged the various dialect codes which the Ethnologue / ISO encoded alongside nds (Sallands, Westphalian, etc) into just two codes mirroring the two Wikipedia codes, I hoped it would make the situation less messy, avoiding some of the debates over spelling and capitalization because some trends are much clearer (even if not entirely uniform) inside nds-de and nds-nl than in nds as a whole, partly as a result of the separate pull that Dutch has exerted on nds-nl and that German has exerted on nds-de.
nds-de and nds-nl are far more dissimilar than e.g. sr, hr and bs (which are functionally identical, all based on one specific subdialect). There are dialects on the edges of each which bleed into the other, as is also true of e.g. Scots vs English (where sentences can sometimes be hard to classify as Scots vs Scottish English). There is also fine internal variation, especially historically: for example, some trends and specific words distinguished similar varieties of Low Prussian from each other, not just from Western Pomeranian vs Mecklenburgish, etc — this is true of most Germanic lects, e.g. Central Franconian, Swedish, even English (contrast da yooge boid ate da olykoek /də judʒ bɜjd eɪt də ˈ(oʊ~oə).lɪ.kʊk/ and the huge bird ate the doughnut /ðə hjudʒ bɝd eɪt ðə ˈdoʊ.nʌt/).
One user previously proposed merging not only nds-de and nds-nl but also pdt into nds as nds.Wiktionary does; I oppose merging pdt into nds. I neither support nor oppose merging nds-de and nds-nl.
- -sche (discuss) 16:31, 25 April 2016 (UTC)
What's your reasoning for Plautdietsch? And what is your definition of it? Many nds-de entries have a tag including Low Prussian, which is what I understand as Plautdietsch. Korn [kʰũːɘ̃n] (talk) 22:40, 26 April 2016 (UTC)
Plautdietsch is a Low-Prussian-derived lect that was historically (and still is) spoken outside Prussia, in America and Ukraine and Russia and elsewhere. It has been kept separate on account of its separate geographic development, just like Luxembourgish and Transylvanian Saxon have been kept separate from each other and from Central Franconian, and Pennsylvania German and Volga German have been kept separate from Rhine Franconian, etc. Low Prussian is the lect that was historically spoken inside Prussia. We have entries in it because it is well- and accessibly-documented.
I have periodically questioned if it was sensible to do things that way — merge all the "inland" varieties of Rhine Franconian under one code, but keep "outland" varieties like Pennsylvania German and Volga German all separate, and likewise Hunsrik, etc, etc. It is convenient in some ways, and the weight of precedent is firmly behind it when it comes to German lects. But I periodically muse that we seem to be doing just the opposite of Ethnologue: they split everything except languages they speak, e.g. they split Fula varieties but not New York vs London English; whereas, we merge Fula but split up (some of) the languages we speak.
- -sche (discuss) 01:53, 27 April 2016 (UTC)
I'll state for the record that I think our language codes for languages which are not officially codified by some nation should be divided solely by difference in grammar and phonology. Minor variations thereof, as well as political and lexical differences, should be the criteria for tags and sub-categories. Exempli gratia I would class Berlinerisch as de. Korn [kʰũːɘ̃n] (talk) 21:15, 27 April 2016 (UTC)
Why even make an exception for languages codified by a nation? --WikiTiki89 21:20, 27 April 2016 (UTC)
Because if we were to start defining and splitting Scandinavian lects based on actual differences, we would become unusable to an average people who wants to look up a word in Swedish and not Svea-Geatlandish and or whatever we'd arrive at. My hunch is that anything with an official set of rules and spellings should have its own header, since it's what people would most likely look up. Don't pin me down on it, though, this statement feels like thin ice. Korn [kʰũːɘ̃n] (talk) 11:07, 28 April 2016 (UTC)
There's no need to use funny names; we'd just call it Southern Swedish or something. Wouldn't people wanting to look up a Norwegian word already be confused that there is no such thing as Norwegian, but only "Norwegian Bokmål" and "Norwegian Nynorsk"? And that Serbian, Croatian, and Bosnian don't exist, but instead Serbo-Croatian does? We shouldn't worry too much about whether people will be confused by which languages we choose to merge and which to split. --WikiTiki89 15:13, 28 April 2016 (UTC)
The southern Swedish dialect is known as Scanian. Donnanz (talk) 15:34, 28 April 2016 (UTC)
We got one of the problems right there: Is Scanian really southern Swedish? I've seen it classified as Eastern Danish. And there are things like Tronderish/Jamtlandic, whose area is in both and Sweden for like 50%. Is that Central Norwegian or Central Swedish? Nor do I think that people would understand "South Swedish" to be a supra-term to what they understand to be just "Swedish", rather than a subcategory of it. You got a point about Bokmål and Nynorsk nomenclature. I guess it comes down to whether one can expect users to intuitively understand that the term covers what is looked for, I don't know if that is or is not the case for Norwegian and any other language in that vein. Korn [kʰũːɘ̃n] (talk) 16:42, 28 April 2016 (UTC)


Discussion moved from User talk:Stephen G. Brown#Palochka.

I see that you were discussing the use of uppercase and lowercase Palochka back in 2012 and 2013. Currently, ru.wiktionary (XML dump) contains some 5000 uppercase and 1000 lowercase Palochka (mixed in with a few hundred uppercase Latin I and Cyrillic dotted I), while en.wiktionary and fr.wiktionary today almost entirely use the lowercase version and tr.wiktionary and ce.wikipedia (Chechen Wikipedia) only use the uppercase version. Would it be possible to find a consensus for all WMF projects on which version of Palochka to use? Should ru.wiktionary, which currently has a mix, change all to uppercase or to lowercase? --LA2 (talk) 00:56, 17 April 2016 (UTC)

I would much prefer that all uses of palochka be of one case. Originally there was only one palochka, Ӏ (u04c0), and I think it should have stayed that way. Today, the original palochka has become the uppercase palochka, so I support the use of the uppercase for all cases. Since uppercase and lowercase palochkas look identical, having it in two cases would lead to a lot of misspellings and trouble.
It would be great if a consensus among all WMF projects could be reached, but I do not know how to go about doing it. —Stephen (Talk) 07:28, 17 April 2016 (UTC)
I agree that lower-case palochka should never be used. --WikiTiki89 14:26, 18 April 2016 (UTC)
I'd prefer lower-case palochka to be used where appropriate. Many wiki-projects go through normalisation. Wikipedias in languages that use palochka will catch up eventually. It's only correct to spell "Qatar" as Къатӏар (Q̇aṭar) (lower case) in Chechen but the spelling "КъатӀар" (upper case palochka) is common. --Anatoli T. (обсудить/вклад) 09:02, 19 April 2016 (UTC)
I noticed that on my iPhone, the lower-case palochka is actually shorter than the upper-case palochka. In other fonts (on my computer), the lower-case one looks like a lower-case "L/l", and the upper-case one looks like an upper-case "I/i"; and so they range from distinguishable by serifs only, to nearly indistinguishable. @Atitarev: Can you link to professionally type-set printed text in a language that uses palochka? --WikiTiki89 19:37, 19 April 2016 (UTC)
No, I don't think I will be able to link to properly type-set texts in North Caucasian or other Russian minority languages, even for common words like лугӏат (luġat). Cf. Chuvash сăмахсар (Roman ă) vs сӑмахсар (sămahsar), Ossetian цæхх (roman æ) vs цӕхх (cæxx). The Internet penetration and digitization of these minority languages is very poor and hard to verify. You'll find that misspellings are often more common than standard or standardized forms. Vahagn might know more about North Caucasian spellings, fonts and the mix-up with palochka and other special symbols, which are missing in other Cyrillic-based languages. We should try and normalize these spellings. I think we can compare the use of Hebrew "׳" vs a simple apostrophe ' or Arabic hamza, which is used inconsistently out there but we use it here. As a dictionary, we can afford using the latest standard forms. --Anatoli T. (обсудить/вклад) 01:42, 20 April 2016 (UTC)
The professionally typeset books of Soviet times do not distinguish the case of palochka. I am in favour of distinguishing lowercase palochka from the uppercase one, because proper nouns can start with a palochka. For example, Chechen Ӏаьрбийн Цхьанатоьхна Эмираташ (ʿärbīn Cḥanatöχna Emirataš, United Arab Emirates). --Vahag (talk) 06:00, 20 April 2016 (UTC)
@Vahagn Petrosyan, Wikitiki89 Our transliteration modules can't handle all possible misspellings either and the translation adding tool User:Conrad.Irwin/editor.js needs some attention for North Caucasian languages (see lines starting after "var diacriticStrippers"). Compare Avar Къатӏар (Q̇̄aṭar) (lower case palochka, correct) and КъатӀар (Q̇̄aṭar) (upper case palochka). The latter is transliterated incorrectly "Q̇̄atӀar". If we use I, l, 1, | instead of proper palochkas, the results are even worse. --Anatoli T. (обсудить/вклад) 08:00, 20 April 2016 (UTC)
@Vahagn Petrosyan: What do current professionally type-set texts do? How would professionally type-set text have rendered United Arab Emirates then and now? --WikiTiki89 15:45, 21 April 2016 (UTC)
According to the few modern sources I have, they too do not distinguish the case of palochka. Compare this extract from {{R:ce:Aliroev}}. --Vahag (talk) 16:06, 21 April 2016 (UTC)
Do your sources say anything about whether the next letter is capitalized (i.e. something like ӀАьрбийн Цхьанатоьхна Эмираташ (ʿÄrbīn Cḥanatöχna Emirataš))? --WikiTiki89 17:26, 21 April 2016 (UTC)
They don't capitalize the next letter. --Vahag (talk) 06:28, 22 April 2016 (UTC)

Yes, we should aim for normalization. But is it settled that the lowercase Palochka is the future? Maybe adding it to Unicode 5.0 was a mistake that we should better ignore? (In the 1970s, the Swedish Academy tried to change the spelling of zebra to sebra and juice to jos, but these reforms didn't catch on, nobody used them, and they are now a laughing matter of a bygone era when overly optimistic planners thought they had an influence that indeed they lacked.) If there is a scientific/societal consensus to use the lowercase Palochka, then we should do so. But is there? It seems to me that the Chechen Wikipedia is already fully normalized and standardized on using the uppercase Palochka. What exactly is wrong with that? --LA2 (talk) 22:09, 20 April 2016 (UTC)

I really favor using only the uppercase as the Chechens do. The palochka is nothing more than a version of the apostrophe, and we would not want to have upper- and lowercase apostrophes, would we? And there is no reason why the unicase palochka couldn’t be used as the initial letter of a proper noun.
It should not be a difficult choice. At the moment, it seems that the speakers and writers of the affected languages themselves prefer to use the unicameral palochka, and we should follow suit. If at a future time the writers of these languages decide that they want two cases for the palochka, it will be a simple matter for us here to make this change on Wiktionary. Just as the readers and writers of a language should have the right to decide the spellings of their words, and how to punctuate their language, those same people should have the say in whether the palochka is unicase or dual case, and we should bow to their usage. It can always be changed later on if the users of the languages want to do it. —Stephen (Talk) 01:02, 21 April 2016 (UTC)
I think this topic should be moved to Beer parlour, it concerns language policies, transliterations of a few languages. Whatever decision is made, could be accompanied by some bot work, changes to translit modules and policy pages. --Anatoli T. (обсудить/вклад) 21:53, 21 April 2016 (UTC)
Topic moved from User talk:Stephen G. Brown#Palochka. —Stephen (Talk) 08:41, 22 April 2016 (UTC)
I think the original reason for adding a lowercase character is that Unicode rules were changed such that case-pairings are fixed and can never be changed (after the glottal stop debacle). As such, a lowercase-only character can get an uppercase counterpart encoded later, but the opposite isn't possible anymore. As such, lowercase counterparts were added for all uppercase-only characters in Unicode (there were only about 5 or 7 or so) so they wouldn't run into problems in the case that a lowercase counterpart would later be needed. -- Liliana 08:55, 22 April 2016 (UTC)
It may the background story but we need to decide, if we're going to use upper case palochka or use upper/lower case palochkas in the concerned languages. I no longer insist on using both, like with other, usual upper/lower case letters. It seems there is more evidence that we need to stick to upper case palochka. --Anatoli T. (обсудить/вклад) 09:23, 22 April 2016 (UTC)
I too don't care that much about which option we choose, as long as we follow it consistently. --Vahag (talk) 09:47, 22 April 2016 (UTC)
In previous discussions, it seemed intuitive to me that we should make the case distinction, for the reasons Anatoli gives above. However, there is a big different between using what amounts to CamelCase, and mixing scripts as in the цæхх vs цӕхх example. If the speakers themselves apparently always or mostly use uppercase, then I guess we should go with that. There is at least one other language with a standard orthography that uses only one member of a pair of cased letters, although it's a constructed language (Klingon uses only I and no i). - -sche (discuss) 15:18, 22 April 2016 (UTC)
The other difference is that with the Latin æ vs. Cyrillic ӕ, we can say that even though we are going against widespread internet practice, we are still staying true to the printed orthography, since these codepoints are not typographically different; however, the lower-case palochka goes against the printed orthography as well, and therefore I oppose using it. --WikiTiki89 15:34, 22 April 2016 (UTC)

Should definitions and glosses reflect the lemma form of a lemma?Edit

For Latin verbs, most of our entries show the definitions in the first person singular. For example abdico. This is presumably because the lemma form is the first person singular present. I'm not sure if this practice makes sense, though, because the lemma represents all forms and not just the first-person singular present. The Latin lemma should be translated with an English lemma, and in English, it is usual to lemmatise verbs in the infinitive, so we should use it in definitions also. Compare verb lemmas for Bulgarian, Irish, Macedonian, Welsh and probably many other languages that have no infinitive, where we already do this. —CodeCat 18:51, 22 April 2016 (UTC)

I agree that lemmas should be translated and glossed with lemmas. For example, dīcō should be defined as "to say". Just like Arabic قَالَ(qāla), which is the 3rd person singular past tense, is defined as "to say", and Macedonian каже (kaže), which is the 3rd person singular present tense, is defined as "to say". --WikiTiki89 19:01, 22 April 2016 (UTC)
I agree in principle, but it will be difficult to get people to comply, perhaps especially for Latin and Ancient Greek. This may be due in part to how these languages are taught. When I was in school and competing in certamen at Junior Classical League conventions, your answer would be considered wrong if you said something like "dīcō – to say". We were constantly reminded to answer along the lines of either "dīcō – I say" or "dīcere – to say". —Aɴɢʀ (talk) 19:16, 22 April 2016 (UTC)
I disagree, as we are defining the word and dico, for example, does not mean "to say". Lemma-to-lemma makes sense on some levels, but I think it's better to provide an accurate translation of that form of the word, especially because beginners in the language might not be aware that the lemma of a Latin verb is not the same form of the verb as the English lemma, and this could be a potential source of confusion. Andrew Sheedy (talk) 11:51, 26 April 2016 (UTC)
Would you then agree with defining Arabic verbs with things like "said", "opened", "defined", etc? And Welsh verbs with "saying", "opening", "defining"? —CodeCat 13:20, 26 April 2016 (UTC)
@CodeCat: That is one reason why Welsh verbs' lemmata should not be the verbnoun. — I.S.M.E.T.A. 16:01, 5 May 2016 (UTC)
We aren't defining the word. We are defining the paradigm. In some languages, the lemma form does not even cleanly correspond to any English tense. Furthermore, many languages lack an infinitive, so what would we put in translation tables? --WikiTiki89 15:10, 26 April 2016 (UTC)
This is another good point. Since we translate from English lemmas to whatever lemma the other language uses, we should translate back into English from lemma to lemma too. It doesn't make much sense for a translation of a Latin verb to appear on the English infinitive page, only for that verb to define itself as a 1st person form. —CodeCat 15:49, 26 April 2016 (UTC)
This is an old subject with previous discussion, Wiktionary:Tea_room/2009/November#nāscor. There is even a vote draft that was intended to address this issue: Wiktionary:Votes/pl-2009-12/Definition layout. --Dan Polansky (talk) 14:14, 30 April 2016 (UTC)

Categorize ditransitive verbs by template labelEdit

Please do. Thanks.--Dixtosa (talk) 12:01, 23 April 2016 (UTC)

Done. - -sche (discuss) 02:27, 27 April 2016 (UTC)

Global ban of Liliana-60Edit

FYI, I sent a message today to the e-mail in the global account log[2] of Liliana-60 to ask exactly why she was globally blocked. I'm waiting for their response. --Daniel Carrero (talk) 19:06, 23 April 2016 (UTC)

Death threats. — Ungoliant (falai) 19:26, 23 April 2016 (UTC)
  • The transparency has been quite underwhelming. I'm curious to see how they respond to your query. —Μετάknowledgediscuss/deeds 19:33, 23 April 2016 (UTC)
    Contrast #Proposal to globally ban WayneRay from Wikimedia above; it links to guidelines that seem to have been ignored in this case. —Μετάknowledgediscuss/deeds 19:36, 23 April 2016 (UTC)
    Im guessing you'll get no reply at all. Renard Migrant (talk) 20:41, 23 April 2016 (UTC)
  • Before you complain about transparency, note that the m:WMF Global Ban Policy clearly states "Also, to protect the privacy of all involved, the Wikimedia Foundation generally will not publicly comment on the reason for any specific banning action." and "Please note that questions about specific WMF global bans will not be addressed, to protect the privacy of all involved." I think this is perfectly reasonable. If the ban was wrongful, Liliana should be the one to sort it out. Also, I'm not sure whether this is significant or not, but three other accounts were globally banned on the same day. --WikiTiki89 15:20, 26 April 2016 (UTC)
    I don't know whether it's significant or not either, but it certainly wouldn't put anybody's privacy at risk to tell us. And yet they won't. (Also, as noted before, they already broke their own rules on global bans in this case.) —Μετάknowledgediscuss/deeds 15:25, 26 April 2016 (UTC)
    It seems there is a distinction between community-initiated global bans and WMF-initiated global bans. The rules for the former do not apply to the latter, so no rule was broken. --WikiTiki89 15:40, 26 April 2016 (UTC)
I am unhappy with the process the WMF is using, it seems to me that anyone can be blocked without any evidence being given to justify that block. The whole point of the WMF is that the community is self-governing. - TheDaveRoss 15:38, 26 April 2016 (UTC)
Just because evidence is not released to the public, doesn't mean none was presented. And you can't judge the affects of releasing that information if you don't know what that information is. I'm sure the WMF doesn't take any of this lightly and certainly would not ban just "anyone". --WikiTiki89 15:49, 26 April 2016 (UTC)
They do not need to reveal the contents of the edit to be at least partially transparent. They could, for instance, post something here (where Liliana was very active) saying that they were taking an action, and justifying that action. Unilateral and silent action is against the spirit of the project, and while I agree that they would probably not wantonly ban anyone, Liliana would likely disagree. She is unable to disagree here. - TheDaveRoss 16:59, 26 April 2016 (UTC)
Liliana is able to disagree in private conversation with the WMF. --WikiTiki89 17:32, 26 April 2016 (UTC)
There is a major flaw with that argument, they hold all of the cards and she holds none. There is a reason why trials in civilized society always include impartial third parties, either a jury or the state or both. When community members make blocks there are other community members who can review the block and judge whether or not it was justified. It happens all of the time on this project. When the WMF blocks someone there is no similar check, so the possibility for abuse is much higher. - TheDaveRoss 17:47, 26 April 2016 (UTC)
The WMF is the impartial third party. Equivalent somewhat to "the state" (but I would caution not to take that analogy too far, there are many differences between an organization and a state). --WikiTiki89 18:02, 26 April 2016 (UTC)
They are not the third party, they are not impartial. The WMF has interests of its own, and has recently shown to go to some lengths to hide those interests from constituents. That whole mess with board members etc. being removed and resigning recently demonstrates that the small group of people who represent the organization officially are not infallible and should work harder to operate in the open. The community writ large is the third party in this situation, the WMF is judge, jury and executioner. - TheDaveRoss 18:28, 26 April 2016 (UTC)
They are a third party. Whether they are impartial depends on what the topic is. As far as Liliana's block is concerned, they certainly are impartial. Liliana didn't do anything personally to them (at least as far as we know). --WikiTiki89 18:32, 26 April 2016 (UTC)
Prove it. You want to argue semantics, I would rather address the issue, which is that the WMF took unilateral action without providing justification or an open mechanism for appeal or community input. I am not the kind of person who distrusts authority on spec, but this process is hypocritical from an organization founded on the concept of openness and community control. I agree that the WMF needs to have the ability to take decisive action to protect the project and the community, but those actions should be very rare and be well justified. - TheDaveRoss 18:46, 26 April 2016 (UTC)
I'm not arguing semantics; if you can't see past the semantics what I'm actually trying to say, then I'll try to explain it better. I can see that your issue is with the lack of openness, which has nothing to do with whether the WMF is an impartial third party (and I have explained why I think so). We are in agreement that there is a lack of openness, but the question remains whether the lack of openness is justified. You say "those actions should be very rare and be well justified". There have been 15 of these bans since 2012. I would say that makes it fairly rare, considering the size of the entire community, but you may disagree with that. Whether they are justified, and whether the lack of openness is justified, is not something you can judge without all of the information, which none of us have. But as long as they remain rare, I am willing to assume they are justified. And I'll reiterate that even if it's not justified, it's not as though Liliana has no recourse. --WikiTiki89 19:12, 26 April 2016 (UTC)
The only way I can judge whether the lack of openness is justified is if the WMF justify their actions, they have not. Fifteen bans is relatively few, the four this week however are another story, but it does seem that the actions are infrequent. Even the infrequent actions, however should be justified. By this I do not mean that they should have justification in the mind of the person taking action, I mean that the person taking action should explain the action to the community. You hit the nail on the head in stating that this "is not something you can judge without all of the information," and I completely agree, which is why more information should be made available. They do not need to reveal the specific offending text, but they should be able to cite the policy under which they acted, and describe the reason why they felt the action was necessary. Concerning Liliana's recourse, the entity to whom she can appeal is the same entity who she believes has acted in bad faith. If I block someone unjustly the person I block can appeal to you, Liliana is not afforded the same recourse. - TheDaveRoss 19:45, 26 April 2016 (UTC)
My point was that if you don't know what they're not saying, you can't judge whether or not they should be saying it. Now did Liliana say that the WMF acted in bad faith, or is it just you who's saying that? And don't forget that the WMF isn't just one person either. --WikiTiki89 19:53, 26 April 2016 (UTC)
Liliana said that the initial set of blocks, and the removal of her permissions elsewhere were unjustified, so I assume she also considers the further action similarly. I have not heard from her since this block came along. You are right that I can't know what they are not saying. I just can't imagine any scenario in which silence is necessary. If Liliana posted nuclear launch codes then they should say that she was blocked for violating whatever policy is in place for such an action. They don't have to give out the codes to explain their actions. Also, though the WMF is more than one person, this action was taken using the anonymous account, so I don't know who actually made the block. - TheDaveRoss 20:09, 26 April 2016 (UTC)
There's a difference between unjustified and done in bad faith. But anyway, I do think it would be reasonable and desirable for WMF to release some information. I just wanted to point out that their current policy itself states that they will not say anything, so we should complain about the policy rather than about the silence itself. I don't know how I even got sucked into this argument. Also, another point is that if WMF states their reason and then is later proven wrong about it, it could be seen as slanderous. --WikiTiki89 20:48, 26 April 2016 (UTC)
In the end it probably comes down to the fact that "the community" can't be sued or tried for a crime whereas WMF can. So WMF needs to have the ability to act quickly and decisively if they feel WMF is at risk. To disclose to the community the details of the allegations and evidence many run the risk of increasing WMF's legal risk.
I am reminded of a Saturday Night Live skit involving a press conference during the Gulf war. DCDuring TALK 16:14, 26 April 2016 (UTC)
As a stakeholder in the WMF, a suit brought against them is a suit against my interests as well. I understand that organizations need to protect themselves (I am on the board of a non-profit myself, so I have some [limited] experience with such considerations). As I said above, it is not that I oppose the action, it is the process by which this action was taken that I am in opposition to. I think that the community should be a safe place for all, and if Liliana's words truly violated that safety then perhaps a ban was called for. Without any form of evidence or justification presented I cannot be comfortable with the action. - TheDaveRoss 16:59, 26 April 2016 (UTC)
@TheDaveRoss Therein is why non-admins don't like admins. To a non-admin, the complaint of blocks being capricious seems hardly surprising. Purplebackpack89 16:21, 26 April 2016 (UTC)
I reject your premise. - TheDaveRoss 16:59, 26 April 2016 (UTC)
  • This block process, especially the global block, has struck me as DerHexer and others talking to the right people privately, rather than anything that approaches a community consensus. Also, aren't you supposed to be disruptive on all of your main projects (of which this is one for Liliana) before being globally banned? Purplebackpack89 17:25, 26 April 2016 (UTC)

Multiple times on Wikis including this, she gave info on where users she had disputes with lived and threatened to visit. On German Wikis, she said she would resort to vigilante justice and that only people who don't shy away from murder get ahead. Multiple people issued blocks over threats and noted the danger of her having privileges that require trust, give access to deleted personally identifying information and allow blocking. As she continued to make threats, people predicted she would be banned. It's sure inexplicable she has been banned. Thrwwy (talk) 20:49, 26 April 2016 (UTC)

This is the e-mail I've sent to them on April 23, they didn't reply:


My name is Daniel Carrero, I am an administrator at the English Wiktionary.
I see that user Liliana-60 has been blocked globally by the WMF, with the summary "WMF Global Ban, questions should be addressed to ca@wikimedia.org".

Up until that block, she was an active member of the English Wiktionary.

If possible, I would like to ask why she has been blocked.

Thank you,
Daniel Carrero

On second thought, I could have said "I know she has made death threats in another project, but really the transparency of this global block has been underwhelming and so an so". Oh, well. --Daniel Carrero (talk) 05:52, 4 May 2016 (UTC)

If WMF said why they banned a user, in some cases that official statement might open them up to legal consequences. ("Oh, they said why User:X was banned, but not why User:Y was banned, so what User:Y did must have been really really bad...") So it's best for them to say that they won't comment on why a user was banned, ever. --Rschen7754 02:54, 6 May 2016 (UTC)
  • A straightforward boilerplate reply, like "We cannot comment on the specifics of any case of banning due to legal considerations," would be much more preferable and professional than simple silence. ‑‑ Eiríkr Útlendi │Tala við mig 03:41, 6 May 2016 (UTC)
    As I've already pointed out, they already say that on the m:WMF Global Ban Policy page. I don't know why we should expect them to respond to individual emails with the same statement. --WikiTiki89 15:01, 6 May 2016 (UTC)
  • Because some kind of response, even boilerplate, is preferable to stonewalling. Flat-out ignoring incoming communications is not professional behavior. (I know a lot of WMF is volunteer; I use this term professional as I lack a better term for expressing all of the bases covered by upstanding, respectful, morally sound, forthright, impartial, etc.) ‑‑ Eiríkr Útlendi │Tala við mig 00:02, 7 May 2016 (UTC)
    Of course receiving a response is preferable to one receiving the response. What I'm saying is that it is unreasonable to expect them to reply when they have a policy of not commenting on such issues. It's not all about what's preferable to us. --WikiTiki89 14:52, 9 May 2016 (UTC)
True I suppose. I also double-checked that I got the e-mail address right and I checked my spam folder just in case. --Daniel Carrero (talk) 03:44, 6 May 2016 (UTC)

Linking to the constituent words in the headword templatesEdit

I find links to

  1. "ex" and "libris" in ex_libris
  2. "Marie Antoinette" in Marie_Antoinette_syndrome

totally useless and potentially misleading.

In the first example libris is an orange link (useless). ex is blue but it has nothing to do with Latin ex (misleading).

In the second case Marie_Antoinette makes one link and if you follow the link you will see a definition that has nothing to do with the syndrome (useless).

So, the questions are

  1. do you agree that both (actually three) of the links should be delinked?
  2. is there any reasonable rule to follow to identify the parts that should be linked?
    and finally super high IQ editors can contemplate on the following
  3. What's the purpose of linking to the constituent words in headword templates at all?

--Dixtosa (talk) 19:54, 25 April 2016 (UTC)

I have always advocated for the auto-linking to be opt-in. Having lost that battle, I tried advocating for a reasonably easy way to opt out other than head=the entire long phrase. I lost that battle too. --WikiTiki89 20:00, 25 April 2016 (UTC)
@Dixtosa How could links from Marie Antoinette syndrome to Marie Antoinette and syndrome possibly be misleading? (I agree that links to Marie and Antoinette are not instructive.) If w:Marie Antoinette is not helpful then perhaps the entry needs an image of a painting of Marie Antoinette when she was relatively young, but had whitened hair. I think it is our job to figure out the best way of answering questions "How did those words come to mean that?"
The default linkage to individual terms is a labor-saver because AFAICT no one here has the AI skill to make a module that could do links better than a skilled editor and links to the individual components are probably the single most usually instructive links. The meanings of many headwords of MWEs do have a comprehensible connection to the meanings of some combination of the constituent terms, sometimes of the individual components. DCDuring TALK 21:30, 25 April 2016 (UTC)

Placing template l to all linksEdit

For the record, I dislike the recent bot-made placement of {{l}} to all synonyms and other sections. In a vote, I would have opposed. I belive there must be a better solution. This is ugly and, if tabbed interface is on, unnecessary. I guess I am in a minority and this is really just for the record. --Dan Polansky (talk) 05:50, 30 April 2016 (UTC)

Mewbot and its author are also not very careful to discriminate between correct and incorrect language labels. This is the result of arbitrary (impatient?) initiation of new classes of actions without bothering to ask about consequences. I noticed the problem when I found English vernacular names under Synonyms headers appearing with orange links. This was caused by their being entemplated with {{l|mul}}. It is quite tedious to find and selectively revert inappropriate Mewbot actions. Perhaps it should be tasked with reverting any class of such errors. DCDuring TALK 12:58, 30 April 2016 (UTC)
I agree with DCDuring. MewBot's code seems too "stupid" to do the job properly ({{l|zh|&c}}?). Given the rampant inconsistency in formatting across this project, this seems like a task that must be supervised if done at all. (Dang, at least maybe check if the language section exists on the page first; that could address what seems to be the core of the complaints I've seen about this task, namely that links are being made indiscriminately and inappropriately.) —suzukaze (tc) 14:53, 2 May 2016 (UTC)
Since words don't exist independently of language I'm in favor of always linking to the section of the language in question. But yeah, saying CodeCat is unintentionally breaking things by bot is a bit like saying rain's wet. Renard Migrant (talk) 10:49, 3 May 2016 (UTC)
One of several things I dislike about "l" is that it doesn't form links in edit summaries. It's nicer to see something like "added derived term amazingly" where it's either blue or red and can be clicked on. Equinox 12:16, 3 May 2016 (UTC)
You have to write edit summaries separately anyway; you can still use bare links in them. What I like about {{l}} even for English is that it will always take me to the English section, even when I've previously been looking at another language. In tabbed languages, at least, the software remembers what language you've been looking at, and will take you to the same language (if present) when you click on a bare link. So if I'm looking at aisling#Irish and then click on the bare link [[dream]], I will be taken to dream#Irish instead of dream#English. But if the gloss had {{l|en|dream}} instead, I will be taken to the right place. —Aɴɢʀ (talk) 12:43, 3 May 2016 (UTC)
I agree with Angr. I also like when the links use the format {{l|en}}, for the reasons he mentioned. --Daniel Carrero (talk) 12:45, 3 May 2016 (UTC)
I usually create edit summaries by copy-pasting the content I added. It's quicker. Equinox 12:55, 3 May 2016 (UTC)
https://xkcd.com/1172/ --WikiTiki89 14:40, 3 May 2016 (UTC)
Not relevant. I'm not doing something foolish; copying and pasting is a standard way to move material from one place to another. Equinox 14:58, 3 May 2016 (UTC)
But if you have added the derived term amazingly to an entry, you haven't actually written the words "added derived term amazingly" to the entry (I hope), so in the example you gave above, your edit summary isn't copied and pasted from your edit. —Aɴɢʀ (talk) 15:02, 3 May 2016 (UTC)
Right. That example is not an example of me copying and pasting. It is, though, an example of a clickable blue or red link. Equinox 15:04, 3 May 2016 (UTC)

I agree with Renard Migrant, Aɴɢʀ, and Daniel Carrero. — I.S.M.E.T.A. 16:05, 5 May 2016 (UTC)