Not to excuse Equinox's behavior, but I get the feeling that I'm being gamed here: three days ago, an admin hid an edit summary of Equinox's from August of 2022, and another from February of 2023. Today, someone created a brand-new account and posted a question on Equinox's talk page about the revdeled edit summaries, complete with diffs and quotes of the edit summaries and an exaggerated mischaracterization of the worst of the two. Now this, which is obviously more of the same.
It seemed odd that someone would be asking in detail about something that disappeared from public view before their account was created, so I did some checking. A quick search for the wording in question turned up a Reddit post by an account name I recognized immediately. It was involved in attempts earlier this year to hijack Wiktionary for use in a vendetta against a certain Wikipedia admin with no ties to us at all (the hijacking included some rather dirty hacks exploiting undocumented practices that would have taken considerable investigation to find). That post is apparently what prompted the hiding of the edit summaries.
I also find it odd that someone who professes ignorance about our procedures would just happen to ping me, specifically, out of the blue.
I'm not really sure what to do here: the comments in question are indeed pretty nasty, but they were momentary lapses that everyone had forgotten about. The complaints, on the other hand, bear all the hallmarks of extensive digging around for anything that could be used for ammunition and of careful targeting. Chuck Entz (talk) 07:20, 1 July 2023 (UTC)[reply]
This is mostly definitely not a new editor. The insistence on the UCoC makes me think of a particular editor, but I'm not really sure it is them. — SURJECTION/ T / C / L /07:36, 1 July 2023 (UTC)[reply]
No, this isn't one of our editors, and certainly not who you're thinking of, since neither I nor Equinox has much to do with their areas of interest. If it's who I think it is, there's a certain irony about someone who chose an account name stating that a certain WP admin "sucks" complaining about personal attacks. Chuck Entz (talk) 07:59, 1 July 2023 (UTC)[reply]
[...] complete with diffs and quotes of the edit summaries and an exaggerated mischaracterization of the worst of the two
How did I mischaracterize them? I quoted them exactly as they originally appeared. The context makes it worse if anything, as he was actually responding to claims of verbal abuse.
It was involved in attempts earlier this year to hijack Wiktionary for use in a vendetta against a certain Wikipedia admin with no ties to us at all (the hijacking included some rather dirty hacks exploiting undocumented practices that would have taken considerable investigation to find). That post is apparently what prompted the hiding of the edit summaries.
No such attempts have ever been made against this wiki. However, there was an attempt against French Wiktionary several months back.
I also find it odd that someone who professes ignorance about our procedures would just happen to ping me, specifically, out of the blue.
I just checked RC and looked for an admin.
I'm not really sure what to do here: the comments in question are indeed pretty nasty, but they were momentary lapses that everyone had forgotten about. The complaints, on the other hand, bear all the hallmarks of extensive digging around for anything that could be used for ammunition and of careful targeting.
These aren't isolated examples. I have about a dozen more if your interested, and that was just after surface-level digging. With more times, you could probably find hundreds of cases. He has been doing this for over a decade. Remember when he deleted the main page?
And that's just what he does in public, he probably does far worse in private. I have found several testimonials from other users of his abuses. OvskMendov1 (talk) 10:18, 1 July 2023 (UTC)[reply]
You have no "testimonials". Anyone could tell you're only here because you have an axe to grind. It is not at all normal for a completely new user account to come in and claim to be an "administrative abuse investigator". Either contribute to the project or leave. — SURJECTION/ T / C / L /10:33, 1 July 2023 (UTC)[reply]
Anyone could tell you're only here because you have an axe to grind.
Against what?
It is not at all normal for a completely new user account to come in and claim to be an "administrative abuse investigator". Either contribute to the project or leave.
No user claiming to be new on the site is in any position to carry out "investigations of abuse of power", nor would any actually new user be inclined to do so. Again, contribute to the project or get out. I'm not going to ask you to present any evidence, because it is clear you do not have any. — SURJECTION/ T / C / L /10:51, 1 July 2023 (UTC)[reply]
@Surjection: Those two diffs are not enough?? That would get any non-admin indeffed. Again, if you do not allow any outside investigation, you are a cult. Dry and cut.
For outsiders, some of the revisions are deleted, but they are about equally as "rude". Advisable behavior? Hardly, but this is half a dozen cherry-picked edits out of hundreds of thousands. We have contributors that haven't been blocked for more vulgarity than this, because incivility on its own is not a crime, especially when it is justified against editors who are being disruptive (as in the links #2, #5, #6, #8 and the "testimonial" which is from a block-evading vandal). — SURJECTION/ T / C / L /11:01, 1 July 2023 (UTC)[reply]
When directed at timewasters and people who don't know what they're doing but insist that they do, a good little "f*ck off" is nothing to be frowned at. PUC – 11:25, 1 July 2023 (UTC)[reply]
I am quite disappointed to see the Wiktionary sysop team showing a generally approving attitude towards f-words. Maybe Wiktionary is indeed quickly approaching what User:Wyang has described on his current user page so with all my respect we can name it "Wyang's curse". -- Huhu9001 (talk) 01:43, 6 July 2023 (UTC)[reply]
This is extremely disingenuous: "fuck" is not comparable to the slur which is actually being discussed, and conflating the two is the wrong approach. Theknightwho (talk) 03:05, 6 July 2023 (UTC)[reply]
They're hardly "cherry-picked": they represent what is unacceptable behavior. No one is arguing that his other edits aren't constructive. What a weird way to talk: "He literally spends 99% of his time not murdering and you focus on the tiny sliver of time when he does!" Admins are expected to be better and represent the best of the community. When you see multiple users in this thread saying that his behavior drives them away, doesn't that make it obvious that his attitude is deleterious for the project? Sheer volume of edits does not wash away someone's use of slurs or hate speech. —Justin (koavf)❤T☮C☺M☯01:47, 6 July 2023 (UTC)[reply]
Incivility does not always warrant blocking, but hate speech and sockpuppetry do. That's my perspective and if I see hate speech posted here, I will always delete it (per the WMF terms of service) and if I see a sockpuppet that I know is a sockpuppet, I will always block him (per our blocking policy). I in no way support either and make that explicit across every wiki where I participate. I know others don't have the same attitude, unfortunately. —Justin (koavf)❤T☮C☺M☯11:38, 1 July 2023 (UTC)[reply]
I don't know if this is supposed to be a joke or something, but you've posted hate speech here, which is pretty outrageous. It's one thing to be kind of surly or make ha-ha jokes that are mean, but it's totally inappropriate to post some of the things you've posted. —Justin (koavf)❤T☮C☺M☯12:33, 4 July 2023 (UTC)[reply]
I'm sadly unsurprised by that development. I once respected Equinox for his quality work but he has has a long history of grinding an anti-trans/gender-critical axe on this wiki.[1][2][3][4][5][6] I've called him out on that front multiple times. In fact, I'd be skeptical of any contribution he makes to trans-related entries, given I've seen him get basic facts wrong when not explicitly banging a GC drum. Presuming that his username on the official Discord is "Eq," then he was also the author of a disgraceful misogynistic comment about me that's half the reason I no longer actively contribute. I've learned that Wiktionary has no stomach for reining in problem admins. But this needed to be said. WordyAndNerdy (talk) 10:35, 5 July 2023 (UTC)[reply]
What business of ours is a user's off-Wiki behavior eg, at Discord? We descend into tribalism when we make it our business. In contrast, we often don't act on behavior on other WMF projects. DCDuring (talk) 14:05, 7 July 2023 (UTC)[reply]
Honestly, I'd view the Discord as an extension of on-Wiki behavior, considering that it's used very heavily to discuss policy and changes to the project, like an extension of Beer Parlour. It's literally linked in WT:Discord server. It may not be "official", but considering that admins on the project are also admins on the server and we require that people authenticate themselves based on their on-Wiki usernames, conduct there should be able to add to context for action here. A normal user would've at least been sanctioned on the server, but even that did not happen. I don't get where all these buzzwords come from, if anything is "tribalism", it's the protection of an admin who hasn't even really been commenting on this to begin with. (Also in the time I've been here, we have often used behavior on other WMF projects to provide additional context for problematic users, like with Dan Polansky on the Czech Wiktionary) AG202 (talk) 15:11, 7 July 2023 (UTC)[reply]
I kind of agree. I wouldn't go as far as to say that the Discord is on the same level as the wiki (for instance, I wouldn't block anyone based solely on the behaviour presented on Discord, and I hope that nobody would), but I definitely think that common decency should be shown both here and on the Discord server. Thadh (talk) 15:46, 7 July 2023 (UTC)[reply]
Maybe we should make it official, at least in the sense that we actually add some guidelines saying that a similar standard of behaviour is expected there. The atmosphere is certainly more laid-back than on the wiki itself - which I don't want to discourage - but at the end of the day it's still a place for the Wiktionary community. Theknightwho (talk) 17:23, 7 July 2023 (UTC)[reply]
The Discord ought to be treated as an extension of the wiki when it is officially presented as an extension of the wiki. It's literally linked from WT:Community Portal as a "means of communication." Given that Discord is basically a big chat server I don't think it's unreasonable to expect that some discussions there might be a bit more casual than what is generally accepted on-wiki. E.g. discussing sports, video games, sharing recipes, etc. But users there (particularly admins) should still be held to on-wiki levels of civility and decorum. It's very easy to DM someone on Discord if one wants to privately express something. WordyAndNerdy (talk) 19:49, 7 July 2023 (UTC)[reply]
@WordyAndNerdy If your choice is to leave then I can respect that, but I think your contributions have been (and continue to be) extremely valuable to the project, and it would be a huge loss to see you go... Ioaxxere (talk) 11:28, 5 July 2023 (UTC)[reply]
This is one of several reasons why hate speech and needless aggression shouldn't be tolerated. The argument "Well, person may be off-putting or posts slurs sometimes, but he's a good editor [otherwise]" ignores that sane persons will leave this project if we allow that kind of behavior. —Justin (koavf)❤T☮C☺M☯12:01, 5 July 2023 (UTC)[reply]
Thanks for the kind words, Ioaxxere. In response to Koavf, I do want to stress that I wouldn't deem the Discord comment about me (quoted at the end of this post) to be a slur or hate speech in the strict sense. The comment also wasn't new when I found it. It was made in February or March of this year, if I recall. I'm not on the same device as my screenshots at the moment. But I absolutely did find a grossly misogynistic overtone in being characterised as an "abused wife." Equinox is among the handful of old-guard contributors who may have been aware of my gender. It's not something I've actively advertised. Aside from mentioning it recently in a discussion of cisgirl, the last time I remember it coming up was once, in 2014, under my old account. And, again, I do want to stress that I think Equinox's overall contributions have been consistently good. It's the interpersonal conduct and axe-grinding that need to be addressed. My usual tendency is to just walk away from conflict. Launching the second vote to desysop Theknightwho was the first time I've tried truly standing my ground. We all saw how well that turned out. Interestingly, Equinox seems to have a habit of white-knighting (white-squiring?) for Theknightwho. He injected a random personal attack into the initial BP thread I opened about TKW in April. The "identity politics" jibe is presumably a reference to the multiple times I've called out Equinox for his GC axe-grinding. In June Equinox called User:LlywelynII an "absolute psychotic joker" in response to what seems to have been a simple misunderstanding over template formatting also involving TKW.[7][8][9] TKW has previously issued a blatantly punitive block against LII, and threatened a second in May.[10] Equinox's userpage notes that he and TKW have met in person. WordyAndNerdy (talk) 13:02, 5 July 2023 (UTC)[reply]
If that comment was indeed directed at you/concerned you (which I personally didn't notice, and I'm sure that's true for others as well), I expect Equinox to apologise and not add such comments about other users again. Thadh (talk) 18:03, 6 July 2023 (UTC)[reply]
I agree, and this continues a pattern I've seen here with protecting vestiges that have been harmful at points. I've recruited folks in the past to work on this project, but I won't do it again. I've already sulked back to the language communities that I feel I can make a meaningful impact and avoided things like RFVE because of issues like this. I honestly contemplated leaving entirely, but I figured that continuing to improve coverage of endangered languages is worth even some vitriol and questionable behavior that I've seen and received. I know not everyone can make the same decision, though. AG202 (talk) 14:13, 5 July 2023 (UTC)[reply]
@Koavf: "sane persons will leave this project if we allow that kind of behavior." I wouldn't go too far down that road, you know; one man's meat is another man's poison. Sure, some people will be put off by such behaviour; but just as many, I suspect, will be by mindless tone policing.
It reminds me of some inclusionists saying "we're going to drive people off by submitting their entries for deletion, we should be much more lenient!". Yeah, but other contributors will get the impression that we're allowing the dictionary to turn into trash, and will be driven off because of that.
In short, egregiously bad behaviour should not be allowed, but please don't start playing morality police any time you see comments a bit too racy for your taste. PUC – 13:34, 5 July 2023 (UTC)[reply]
You deleted rudeness and obscenity, but nothing among the many links provided in this discussion is hate speech by any reasonable definition. There are a lot of links provided which also purport to have some content which they don't (e.g. WordyAndNerdy referencing the cisgirl discussion page as some kind of personal attack). This whole thing feels very ginned up and brigade-y. My experiences with the original complainer lead me to believe that causing drama within Wiki communities is a hobby, albeit a particularly deranged one. - TheDaveRoss14:12, 5 July 2023 (UTC)[reply]
Did you actually read all the comments in this thread? I was very clear that I mentioned the cisgirl discussion in the context of being one of the two times I recall revealing my gender on-wiki. Koavf also hasn't posted any links thus far. That was OvskMendov1, and regardless of whether said user has a vendetta, etc., that doesn't excuse misconduct from Equinox and other admins. WordyAndNerdy (talk) 14:19, 5 July 2023 (UTC)[reply]
I apologize, there is a lot going on in this thread. I misread your comment, I thought you had said that Equinox brought it up in the comment, but you are saying that you brought it up. As far as Koavf's deletions, I assume that was referring to the deleted edit summaries, which I did read through. - TheDaveRoss14:27, 5 July 2023 (UTC)[reply]
Appreciate the apology here. I brought it up to highlight the fact that my gender isn't common knowledge, but that Equinox can reasonably be presumed to be aware of it either through the cisgirl discussion (which he initiated), or as a long-time contributor who may have seen the 2014 thread on my old account's talk page. Thus, I don't think calling me an "abused wife" was just an extremely poor-taste metaphor that accidentally hit the mark gender-wise. I firmly believe it was intentional. WordyAndNerdy (talk) 14:36, 5 July 2023 (UTC)[reply]
"nothing among the many links provided in this discussion is hate speech by any reasonable definition" You all are just making up stuff: what I deleted was not linked to in this thread. You don't know what you're talking about. Additionally, the initial poster may well have some questionable motives and some inappropriate behavior himself, but that doesn't justify an admin's abusive language or inappropriate behavior of his own. I am not here to justify the OP (and in fact, do not support his methods). —Justin (koavf)❤T☮C☺M☯14:23, 5 July 2023 (UTC)[reply]
You deleted the edit summaries of several of the links provided. If you are talking about other deletions, please link those. Accusations without evidence are not particularly helpful. - TheDaveRoss14:28, 5 July 2023 (UTC)[reply]
That edit summary (which I was aware of) is certainly the most problematic of those I have seen, though still well short of hate speech. I don't think it is wise to use terms which are sometimes used in hateful ways without context and in a public forum, but on its own and in context this is not hateful speech. Are there other examples? - TheDaveRoss14:43, 5 July 2023 (UTC)[reply]
The numerous gay men who have chosen to self-identify with that word might have something to say on the matter. Also, using a word is not hate speech, there has to be hateful intent. Poor taste is not hate. - TheDaveRoss15:16, 5 July 2023 (UTC)[reply]
Seconded. While the edit summary is indeed problematic, your definition of hate speech is imo quite liberal. Is someone exclaiming "those stinky Belgians!" or "those f*cking fries-eaters!" automatically engaging in "hate speech"? No, intent and context are key. PUC – 15:21, 5 July 2023 (UTC)[reply]
Come on, I would’ve expected better from you. Putting aside the issue of whether this is directed hate speech, gay men choosing to identify with that term does not mean that they are comfortable with *other* people using that term (and as a queer man, I’m very much not comfortable with people not in the community using it). Our own usage notes at words like the n-word explain that. The context in this instance is clear; let’s not bring up unrelated hypotheticals like that. AG202 (talk) 15:30, 5 July 2023 (UTC)[reply]
I sympathize, and I would not choose to use the f-slur in any context because I understand that the word has a lot of baggage. You referred to yourself as queer, and as recently as perhaps five years ago I would have said the same thing about queer that I now say about the f-slur (specifically that I wouldn't use it), but language changes, and in this area it is changing rapidly. In this case I am not willing to assume who Equinox is, and whether it is appropriate or not for him to use the f-slur. I am instead going to look at the context (what the related actions were, who it was addressed to) and determine that it might have been distasteful, but was not hateful. The fact that it is ambiguous means that it shouldn't have happened, but labeling anything we personally find offensive or distasteful as hate speech is not the path to a more accepting society. - TheDaveRoss15:39, 5 July 2023 (UTC)[reply]
Again, this is an issue that hinges on *reception* and usage. As a queer man, I am fine with using the word queer in my spaces and okay with people using it in provided that it’s not in a hateful or offensive manner. Someone saying, “all the queers are doing xyz negatively” would be considered hate speech as it’s targeted language towards a certain group. Just because it’s been reclaimed by the in-group does not give carte blanche to use it whenever and wherever. Re: creating a path to a more accepting society. Sure let’s not call it hate speech for your thought experiment, but you can agree that it’s offensive. The way to make sure it doesn’t happen again is to take direct, meaningful action. Regardless of the semantics of what it is, it still happened, and we cannot allow offensive usage like that to stay, especially from an admin. This is the paradox of intolerance in action, honestly. AG202 (talk) 16:11, 5 July 2023 (UTC)[reply]
Had I seen this when it was not-stale, I would have blocked immediately and would have started a conversation about de-sysop. Since it's older, it's not as critical, but the "Oh, I'm just a naughty boy--boys will be boys, tee-hee!" attitude is not encouraging. —Justin (koavf)❤T☮C☺M☯16:59, 5 July 2023 (UTC)[reply]
Evaluating based on reception is impossible, since people can choose to be offended by anything. As an obviously trivial example, if I decided that sysop was the only appropriate way to refer to people who could block/delete on wiki projects, and it was inappropriate for anyone to use admin to refer to such people, that does not make your previous statement hate speech, generally offensive, or even inappropriate. I am not attempting to trivialize the use of words which are broadly considered offensive due to their historical usage, or their meaning, or their use with hateful intent in other situations, but clearly my reception of your language is not enough to fully characterize your language. I am sure there are people out there who are offended by the label queer, but undeniably there are large numbers of people who have adopted queer as a positive label. The context and intent are what constitute the difference between inoffensive speech, offensive speech, and hate speech. This is, of course, not a problem which is easily resolved, or will be resolved here. - TheDaveRoss18:18, 5 July 2023 (UTC)[reply]
The context here is offensive speech, at the very least, though. And let's not ignore the point I made that even if this is not hate speech, the fact that it is objectively offensive speech means that action should be taken. The context has been provided; there's no way to talk this out of being offensive. Let's focus on that and what's to be done for the future. Focusing on whether or not it's "hate speech" honestly feels like a distraction, and I'm not sure why it really matters considering that it's offensive nonetheless. (Though personally I still believe that this is hate speech) AG202 (talk) 18:45, 5 July 2023 (UTC)[reply]
The context I mean is that the word was used in an edit summary on Equinox's user talk page, without any other words around it, directed at (presumably) a particular user (as the edit was to remove some content that user had added to the user talk page. If you look at the totality of the interactions between those two users, and at the manner of speaking that the two use consistently, I think it is pretty clear that the term was being used to be edgy and 4channy, and not out of hatred for gay people or any other group. I do think it is important to distinguish between hate speech and speech in poor taste, because I do support the notion that hate speech can result in an immediate ban from the project. Distasteful or generally offensive speech may warrant some other punishment, and long-term incivility likely should result in a ban as well, but they are worth distinguishing to me. - TheDaveRoss13:05, 7 July 2023 (UTC)[reply]
I can understand why you have that opinion, but I hope you can understand that there are other valid opinions in the world. - TheDaveRoss13:09, 7 July 2023 (UTC)[reply]
Other punishments could include reprimands, temporary or permanent removal of tools, temporary blocks, bans, and referrals to WMF for global bans. - TheDaveRoss15:29, 7 July 2023 (UTC)[reply]
Not sure my input in this matter would be helpful. But desysopping is the strongest sanction I feel would be warranted here. Equinox, to his credit, has sat out this discussion. I think that's allowed it the space it needed to proceed in a constructive and generally civil manner. Plus, unless I'm mistaken, I don't believe there have been allegations of abuse of admin powers (e.g. retaliatory blocks, or threats of such). I think Equinox's problematic behaviour has been limited to breaches of civility and what would be considered "POV-pushing" on Wikipedia. Definitely something that needs to be addressed, but Equinox has shown self-awareness in that regard in the past. WordyAndNerdy (talk) 21:34, 7 July 2023 (UTC)[reply]
@TheDaveRoss and PUC - You're both doing the exact thing you've accused Justin/Koavf of doing, i.e. insisting that Wiktionary follow your own subjective (and, I daresay, unactionably high) standard of "hate speech." There need to be baseline standards of acceptable and unacceptable conduct for a community-driven project like this to function. Using the F-slur outside of the context of discussing the F-slur itself (this is a dictionary, after all) should be unacceptable. That ought to be pretty uncontroversial. Intent shouldn't factor into the consideration. That affords too much wiggle room for people to excuse inflammatory and abusive comments by claiming, "It was a joke," "You're too sensitive," "You're a liberal snowflake" etc. WordyAndNerdy (talk) 15:42, 5 July 2023 (UTC)[reply]
I think the standards of accusation should be much higher than the standards of acquittal. This is the basis of "assume good faith" or "guilty beyond all reasonable doubt". See above about intent. I agree that it is hard, and that my stance will allow more inflammatory language than a hard-and-fast policy of merely blocking anyone who uses words which someone doesn't like, but I am not willing to cede that freedom of expression is dead, or that any one group or individual holds the moral high ground on evaluating everyone's speech. - TheDaveRoss18:26, 5 July 2023 (UTC)[reply]
I'm not saying slurs should get a free pass. What I'm saying is rudeness and vulgarity don't automatically equate hate speech; it seems to me that this is not so much my subjective interpretation of the term hate speech as a regard for what it literally means. We should not be throwing this buzzword around every time someone engages in unsavoury behaviour. PUC – 22:13, 5 July 2023 (UTC)[reply]
There are African-Americans who have reclaimed the n-word. That doesn't mean that casual "ha-ha, so funny" usage of it as a slur isn't a hateful slur. It's totally inappropriate behavior and should never have been posted. —Justin (koavf)❤T☮C☺M☯15:43, 5 July 2023 (UTC)[reply]
Deploying the F-slur is rather unambiguously hate speech. Like, unless the discussion is about a bundle of sticks or maybe 4chan users circa 2011, usage of that particular word does not warrant the assumption of good faith, in my book. Plus, Equinox's history of GC axe-grinding has been unnecessarily inflammatory, if not crossing the line into hate speech. This being the worst example that I can recall. And he did use the T-slur in that one, albeit as part of a hyperbolic fake quote. WordyAndNerdy (talk) 14:59, 5 July 2023 (UTC)[reply]
I don't know what the point of these fresh accounts (by "completely different" people, I'm sure) coming in and trying to manufacture consensus is. Anyone could see through it. — SURJECTION/ T / C / L /14:38, 5 July 2023 (UTC)[reply]
The person who initially posted is an alternate account of someone who is a critic of WMF projects. The use of alternate accounts (except in narrow circumstances and with declarations) is obviously out of bounds and as others here now, I always block sockpuppets on sight. The motivations of the OP may be at a cross purpose with the community here, but even then, his complaints are justified. —Justin (koavf)❤T☮C☺M☯15:19, 5 July 2023 (UTC)[reply]
Since I seem to be the focus of some of this, let me explain myself: everyone here has their own shortcomings. Mine is that I'm not comfortable with confrontation. I dislike having to take the initiative in disciplinary matters. In this case, the fact that the complaint came from outside the community by someone who was hiding their identity and was in reference to summaries on edits to Equinox's personal web page 5 or 6 months and more after the fact pointed to trolling. I'm not about to take the initiative to discipline a long-time, productive editor over something like that.
That said, I want to be clear that I really, really wish that @Equinox wouldn't write things like that. I'm sure he sees it as just letting off steam, but what he writes becomes public record and reflects badly on him and on the project. Yes, there are trolls, but we shouldn't be giving ammunition and incredibly easy targets for those trolls. Also, this is a very diverse community, and any random slur is likely to be offensive to someone here.
You may notice that I use my real name here. There's a reason for that. Early in my online career, it became apparent to me that anonymity can lead to bad behavior by otherwise decent people. I see a parallel between mob mentality, where people do things as an anonymous part of a group that they would never do individually, and online behaviour. It's too easy to say to oneself: "no one knows who I am, so I don't have to care what people think." I decided to use my real name and take responsibility for my actions. There are trolls who have taken advantage of that, but I don't regret it. Of course, that's my personal decision- I'm not about to judge anyone else for what they decide in such matters. For some people who are vulnerable in real life, there may not even be a choice.
I'm running out of time before work, so let me sum it up: as much as I dislike confrontation, I do take what people like @WordyAndNerdy have said seriously. Their complaints are genuinely based on their own experience, and are miles away from the fake drama that comes from trolls. I see my role here as the implementer of community consensus: if it becomes clear that the community has decided an action needs to be taken, I will take it. Chuck Entz (talk) 15:08, 5 July 2023 (UTC)[reply]
Thanks for that. As is obvious, the person who started this thread was using multiple accounts, may have ulterior motives, etc., but that doesn't invalidate all his concerns, nor does it absolve Equinox (or you or me or anyone else) of behavior unbecoming of someone who is supposed to be a leader by having admin privileges. Unfortunately, there is very uneven application of things like rules or best practices or terms of service and some users disagree about when and how to let something slide. To some degree, that's inevitable and it's the job of someone who has more user rights to have good judgement and, as you wrote, abide by community consensus. Someone who fails to show good judgement repeatedly and is hostile to others should not be an admin. I'm not proposing any revocation or a vote now for a variety of reasons, but there should be no place for hate speech and the kind of "ha-ha I'm so funny" slurs and aggression on this site. —Justin (koavf)❤T☮C☺M☯15:18, 5 July 2023 (UTC)[reply]
As someone who hates drama and tends to avoid discussions like this, I have to say, at a certain point you need to take action against the people who are causing the drama, esp. when there's hate speech involved. I blocked Dan Polansky for racism, personal attacks, disruptiveness, etc., and IMO it has reduced some of the drama. I think we should consider taking action against Equinox; this isn't a new issue and it's not likely to get better on its own. Benwing2 (talk) 21:15, 5 July 2023 (UTC)[reply]
The fact that strongly disagreeing with the dominant perspective on transgender issues is listed among the examples of "hate speech" above is deeply troubling to me. Are only socially progressive people welcome on Wikimedia sites now? Can even our most valuable editors not make politically incorrect comments? I find this genuinely frightening for what Wikimedia represents. But perhaps I missed some genuine examples of hate speech in the above discussion. Andrew Sheedy (talk) 21:39, 5 July 2023 (UTC)[reply]
"strongly disagreeing with the dominant perspective on transgender issues is listed among the examples of "hate speech" above" When did this happen? "Can even our most valuable editors not make politically incorrect comments?" Such as? —Justin (koavf)❤T☮C☺M☯21:42, 5 July 2023 (UTC)[reply]
I'm primarily referring to the diffs linked by WordyAndNerdy above. But reading some of the comments I missed, I realize you admins may be primarily referring to edit summaries that I cannot see. I initially took you to be referring to said diffs as hate speech, because you said, "This is one of several reasons why hate speech and needless aggression shouldn't be tolerated" in response to WordyAndNerdy's statement that they are considering leaving and understood you to be calling the examples he gave "hate speech." Andrew Sheedy (talk) 22:14, 5 July 2023 (UTC)[reply]
The examples WordyAndNerdy linked above in the comment beginning "I'm sadly unsurprised by that development...". They are a list of unprofessional, but not hateful, comments made by Equinox. Andrew Sheedy (talk) 22:26, 5 July 2023 (UTC)[reply]
This is the second time within the course of this discussion that someone has put words in my mouth. I have not personally characterised anything that Equinox has written as "hate speech." I remarked that it was "sadly unsurprising" that someone with a known history of anti-LGBT (or at least anti-trans) axe-grinding on the wiki would casually toss out the F-slur. At the same time, I don't think the "hate speech" characterisation is misplaced, even if it's not one I'd make myself. Everyone has different standards of what they consider to be acceptable vs. unacceptable speech within the context of this project. Turning this into a quasi-legalistic debate over the precise meaning and ramifications of the terms hate speech, hateful, etc. seems like tone-policing to dodge having to examine or address unacceptable conduct by an admin. At points I'd even say it borders on gaslighting because I ought to be the sole authority on whether a misogyny-tinged comment made about me is hateful and offensive. And what the framing of this as some kind of free speech issue – of "our most valuable editors" being allowed to make "politically incorrect comments" – is doing is telling me and everyone else offended by such hate speech/axe-grinding/political incorrectness (YMMV) that our thoughts, feelings, and contributions don't matter. Sure, we might contribute (or have once contributed) as prolifically as Equinox, and with as much proficiency, but we're not "valuable editors" worthy of the same special consideration as him. Am I supposed to meekly abide not only the off-topic inflammatory axe-grinding on LGBT issues but also a targeted misogynistic remark about me? People can go to 4chan or Twitter if they want to share the hottest of hot takes. I shouldn't have to put up with the stuff I've described in in my comments here as some kind of price of admission for contributing to a dictionary. I've held back saying anything about finding the "abused wife" comment for two months (mostly because I haven't wanted to contribute in its wake). I've let the anti-trans axe-grinding pass with only mild-to-moderate reproach for years. I did not wake up this morning and randomly decide to pick a pointless fight. It hurts having to bring this to the table, because I deeply, deeply respected Equinox. WordyAndNerdy (talk) 00:15, 6 July 2023 (UTC)[reply]
Correction: I did deem "deploying the F-slur" to be "rather unambiguously hate speech" later on in this discussion, but didn't characterise any of his remarks/edits as such in my initial post, which I thought was what was being referenced here given the mention of diffs. WordyAndNerdy (talk) 01:09, 6 July 2023 (UTC)[reply]
@WordyAndNerdy I apologize, that was indeed a mischaracterization. It wasn't intentional--I misread your comment in reply to Koavf's statement that Equinox had used hate speech. I realized by the time I made subsequent comments that that wasn't what you had said, but I failed to recall that I had conflated your comment with Koavf's in my first remark. I agree that Equinox's behaviour towards you has been unacceptable, now that I have a better understanding of what that in fact was. I hope both you and Equinox stick around. Andrew Sheedy (talk) 17:13, 6 July 2023 (UTC)[reply]
Appreciate this. Would probably make the discussion here progress more smoothly if those looking at it from the outside could try to disentagle their thoughts when appraising appraisal the overall situation. The problem isn't an isolated inappropriate comment by Equinox toward another user. It's a long-term pattern of making such comments. It's not having an opinion or even registering it on-wiki. It's expressing an opinion many find objectionable in itself in inflammatory terms and making mainspace edits in servive of said opinion. This is, of course, between a long history of consistently good work. I've had friendly interactions with Equinox as well (see my talk page and its history). Not fun having to send your friends to the principal's office, as it were. But there's a time when a pattern of problematic behaviour shifts from something that can be tolerated or overlooked to something that needs to be directly addressed. WordyAndNerdy (talk) 20:50, 6 July 2023 (UTC)[reply]
The majority of the discussion hinged around Equinox’s usage of the f-slur. Please make sure to read the full discussion. That in itself is more than concerning considering the history that Benwing2 mentioned, and I’m getting frustrated at these constant detractions. It shows to me that certain editors are more focused on random semantic nuance or comments on “political correctness”, and as a consequence, protecting certain negative aspects of the site, rather than focusing on the fact that an admin used a slur multiple times point blank. AG202 (talk) 21:57, 5 July 2023 (UTC)[reply]
I'm sorry. I found certain aspects of the discussion confusing to follow, because much of it concerns edit summaries that are only visible to admins. I also saw Equinox being accused of hate speech and within context, it seemed that the hate speech referred to was in fact certain comments he made, which I could see, as opposed to the edit summaries that I was unable to see. Andrew Sheedy (talk) 22:18, 5 July 2023 (UTC)[reply]
Also, I don't think using the f-slur is hateful in itself. Did Equinox in fact direct it at someone with the intention to insult them? That's unclear to non-admins from this thread. Equinox is known for saying provocative things. It's not a praiseworthy thing, perhaps, but I think we should be careful with accusations of hate speech. I have read the whole thread and I have not seen a single example of hate speech personally. I'm not an admin, so perhaps this doesn't concern me, but I am part of this community and Equinox is one of the most valuable editors we have and one whom I most appreciate personally, so it would be nice to know what all this fuss is actually about. Andrew Sheedy (talk) 22:32, 5 July 2023 (UTC)[reply]
It's weird to me that you reflexively find some abstract threat to your ability to wrongthink more of a threat than multiple users in this thread explicitly saying that they've been made to feel unwelcome and slurs and aggression have driven them to not contribute. —Justin (koavf)❤T☮C☺M☯00:03, 6 July 2023 (UTC)[reply]
@Koavf I apologize. I did not read the situation objectively, but came at it with my own experience of certain opinions being called "hate speech" simply by virtue of their being offensive to some people, and I misunderstood parts of the conversation. I am going to step back from this conversation, because it's clear that it involves a situation that I was not privy to anyway. Andrew Sheedy (talk) 17:17, 6 July 2023 (UTC)[reply]
I'm not an admin. I've been fielded for it multiple times, most recently by Equinox, ironically. But I remain tool-less by choice. (TL;DR - I know I'd make a bad admin.) I just knew what revdel'd edit summary was being referenced here ("def**gotisation") because I saw it before it was revdel'd and it stood out to me at that time. WordyAndNerdy (talk) 01:02, 6 July 2023 (UTC)[reply]
People don't have to share all the same concerns in equal measure, and be equally worried by the same things, though? Although I don't approve of those slurs, I personally care very little about them. I am indeed more interested in these "random semantic nuance[s]" you're speaking of. PUC – 22:30, 5 July 2023 (UTC)[reply]
Likewise. I would be far more concerned about an admin with hundreds of thousands of edits being banned over alleged hate speech that I can't even view than about forgiving the occasional slur (which is not to say that I at all approve of the latter). Andrew Sheedy (talk) 22:34, 5 July 2023 (UTC)[reply]
I would like to give you (and other autopatrollers) access to the relevant edit summaries, so that you can judge for yourself, but it seems there's no way of doing so without making them visible for everyone.
Since we already have various protection levels for pages, couldn't we ask the devs for a similar thing here? PUC – 22:48, 5 July 2023 (UTC)[reply]
An admin using a slur in edit summaries to users should not be brushed aside or summarized as just “drama”. Please be better about this. @Chuterix honestly that’s the gist, an admin using slurs and other targeted comments + a history of questionable edits. AG202 (talk) 00:31, 6 July 2023 (UTC)[reply]
The procedure can be found at WT:Votes if my memory serves me correctly, but I’d caution any vote creation until folks like @Chuck Entz & @Benwing2 give more opinions. This project is notorious for protecting the aggressor in situations like this, and I would not want a failed rushed vote to vilify this type of behavior. AG202 (talk) 00:35, 6 July 2023 (UTC)[reply]
@AG202 Not at all my intent. Per User:Benwing2: "[A]t a certain point you need to take action against the people who are causing the drama, esp. when there's hate speech involved." If anything I'm pointing out that "admin shenanigans" seem to be happening on a fairly regular basis these days. Ioaxxere (talk) 05:06, 6 July 2023 (UTC)[reply]
i dont believe its possible in MediaWiki to have a view permission that is anything other than admin/non-admin. this has come up before and i think the WMF says it would be a security risk so it just isnt enabled in the software. —Soap—23:28, 5 July 2023 (UTC)[reply]
The two edit summaries referred to at the beginning of this thread are still visible here, anyhow, so it shouldn't be necessary to unhide them; anyone can view that conversation. —Soap—08:39, 6 July 2023 (UTC)[reply]
Has anyone who’s meaningfully contributed to the conversation directly said that he be banned? Quote them directly if so. All these hypotheticals not focusing on the issue at hand are pissing me off. Don’t apologize again if you don’t mean it. I don’t care about whatever nuance xyz has. What will that discussion end up as? Nothing. He used an offensive slur. That is the objective truth. That is what I care about. Something should be done about it. If y’all are just going to focus on whatever irrelevant semantic issue is at hand to deflect, go ahead, but I hope that others actually see the problem at hand. AG202 (talk) 00:30, 6 July 2023 (UTC)[reply]
Given the pace at which this thread was moving yesterday, i think it's possible people have accidentally misread or mistyped a few things. The confusion above about the "f-words" may have been an honest mistake too; I admit when I first saw this thread I misunderstood it as well. As another example, all our questions about the redacted edit summaries could have been answered by simply going to Equinox's talk page history, but seemingly nobody (including myself) noticed until just now that the edit summaries we all wanted to see are hiding in plain sight in this diff. I think we were all just too distracted to focus, and not trying to distort the truth. I would encourage people to read carefully before they post here, but also to understand that people make honest mistakes when emotions run high. —Soap—09:32, 6 July 2023 (UTC)[reply]
A probably not important nitpick, the quotes of the f-slur edit summary are universally misquoted. It is probably a typo of the word people are quoting, but it isn't actually the word which is being quoted. - TheDaveRoss12:57, 7 July 2023 (UTC)[reply]
I agree, it is just mildly frustrating when someone is being judged for their speech and the accusers couldn't be bothered to accurately represent that speech to begin with. - TheDaveRoss13:06, 7 July 2023 (UTC)[reply]
By "universally misquoted" you mean that people have quoted the spelling with an "S" (-isation) when it was actually the spelling with a "Z" (-ization). Spelling doesn't alter the meaning of the word, nor the fact it contains an offensive slur or the context of its use. Given that I have been substantively misquoted twice in this thread – including once by you, by your own admission – I don't think it's reasonable to be presenting a simple mix-up over spelling as some kind of grave misrepresentation or twisting of words. I doubt anyone is going to judge Equinox for what spelling was used in the edit summary. Few are that partisan when it comes to the North American vs. UK spelling divide. In fact I'd hazard that this was an error borne out of hypercorrection: presuming UK spelling from a UK speaker after the revdel hid the precise text. WordyAndNerdy (talk) 20:49, 7 July 2023 (UTC)[reply]
That isn't actually the typo, it doesn't have an "ot" in it but it does have a "z". Nobody said it was grave. Again, and as clearly stated, it isn't actually important. - TheDaveRoss21:14, 7 July 2023 (UTC)[reply]
IMO the problem here is that many people are thinking this is an isolated case, or a few such cases, of using a slur. If that were the only issue I would not be raising my voice here, as I am also willing to give people the benefit of the doubt. The situation as I see it, rather, is a long history of abusive/intimidating/insulting/etc. behavior, along with little or no understanding on the part of Equinox of why this behavior is problematic or indications that they're likely to change in the future. Such behavior has real harms for the project, contributing to a toxic environment that causes good contributors to leave and turning off potential new contributors. I understand that Equinox has contributed quite a lot of good edits, but they can certainly continue to do that without being a sysop. People have been desysopped in the past; e.g. that happened to both Rua and Wyang as a result of a protracted wheel war that the two of them engaged in (and IMO this behavior is less problematic than Equinox's behavior). Benwing2 (talk) 00:55, 6 July 2023 (UTC)[reply]
The thing is that verbally abusing other contributors is not an issue relegated to one administrator; it is systemic and it has been a problem for more than a dozen years now. I have suffered verbal abuse from administrators like Mglovesfun, Qehath, and elsewhom. I myself have exploded at others (including at Equinox himself). Users on this project—and perhaps especially admins—have rarely suffered any serious consequences as a result of verbally abusing others. I remember when I first joined the project I tried to act formal and professional, but when I saw that it wasn’t paying off and others were getting away so easily with being rude, I gave up and sunk myself to their level. This has long been a common issue and I don’t see it going away anytime soon. Prove me wrong, if you wish. — (((Romanophile))) ♞ (contributions) 23:01, 9 July 2023 (UTC)[reply]
"Something should be done about it." If I see Equinox using abusive language or slurs, I will block him and make a vote for desysop. Many users have made it clear that his repeated behavior is driving them away from the project and have been explicit that his actions are unacceptable, even those who are making some apologies or excuses to contextualize why it's not "hate speech" as such, etc. It is clear from the WMF Code of Conduct and terms of service as well as the community's will that this behavior should not stand. See also our Wiktionary:Blocking policy: "Abuse, plagiarism, persona non grata type blocks, based on community consensus". I'm hopeful that Equinox can take this as constructive. —Justin (koavf)❤T☮C☺M☯01:02, 6 July 2023 (UTC)[reply]
You've seen Eq use slurs, I think. Have you considered blocking him and making that vote? Or am I misinterpreting something? cf (talk) 01:28, 6 July 2023 (UTC)[reply]
I've considered it previously, but it seemed fruitless. Now, there is a clear community consensus against these actions and I'm not a "rogue admin". Note also that as a fellow admin, he can unblock himself, but I'm hopeful that he wouldn't do that for optics. Similarly, if I ever use slurs or hate speech on this forum or any other, please do block me for an appropriate length. —Justin (koavf)❤T☮C☺M☯01:29, 6 July 2023 (UTC)[reply]
I want to be conservative about piling on and making this a bigger deal than it already is because 1.) Equinox is a valued contributor outside of his harsh and unnecessary language and 2.) I don't want this to be a product of someone just trying to cause chaos and discord. If he's going to be sanctioned or have his rights removed, I want it to be for valid reasons going forward rather than retroactively. That said, if you make a vote, I will respond to it. @Equinox: whom I'm sure is reading this and I hope taking it to heart. —Justin (koavf)❤T☮C☺M☯01:42, 6 July 2023 (UTC)[reply]
I trust Equinox to employ language well-measured, including slurs. By not recognizing his contributed passion, that is still moderately intimidating and insulting, considering the increasing likelihood of offensive incidents for everyone who is not doctrinary after dedicating more lifetime to the project, you yourself are more toxic than he ever could be in half a decade—not least to yourself: hunting down slurs or hate-speech is self-harm; how do you feel after discussing this matter for days? It is that rumination, or tantamount to it, as a sublimated form, that fulfils a prime criterion of depression. “Abuse” is a polysemous term, and you are doing the weightier one than Equinox. Argument functions normally if it goes on the offensive occasionally, but the offensive should not take, not to say exhaust, the energy of days against individuals one cherishes. Virtually anyone came home and maligned one of his family, where it is of little concern afterward whether formally a slur or hate-speech was employed: today you are a goat, tomorrow a hon. Much stays inexplicable, and must so. Distinguish a “history” from a pattern, which may exist as guided by systemic error in attitude or only hyperfocussed on by you rather than being a personality trait essential to the person. Or fiat iustitia et pereat mundus? Fay Freak (talk) 12:14, 8 July 2023 (UTC)[reply]
Just ignore them. They tend to add nothing to these types of conversations except wasted time due to not being able to understand what they say. AG202 (talk) 17:18, 8 July 2023 (UTC)[reply]
You both should be ignored. You only pretend to explain things in spite of everyone being fully aware and woke already. Fay Freak (talk) 17:28, 8 July 2023 (UTC)[reply]
For the record, it's been stated that this user has a history of linking to Neo-Nazi websites in entries, so I'm really not surprised at all that they'd say slurs when angry, hence why I don't think it's even worth it to engage. @Koavf, @CitationsFreakAG202 (talk) 17:54, 8 July 2023 (UTC)[reply]
While I also have some trouble understanding the substance of Fay Freak's message, nowhere have I seen proof that he "says slurs when angry". Your way of using another participant's inference as if it were an established fact already is almost as disgusting as the slurs you're decrying. PUC – 19:04, 8 July 2023 (UTC)[reply]
You all cannot actually be serious. They didn’t even deny it above when asked about it directly. Y’all will do anything, literally anything at all to derail. I’m not putting up with this BS anymore, and I don’t think anyone serious about actual action should. You as an admin (who took three votes to become one) have yet to even suggest any meaningful action for the person that sparked this issue and said slurs in the first place. You even said that you care more about semantic arguments and “care little about [slurs]” in the first place. Why even comment on this thread? If there’s an eventual vote on any action, just vote oppose and leave it at that. Knowing your past attitude and interactions with me and others on issues like these and other problems (re: “inclusionism”), I question whether this reply is even worth it in the first place, but it will be at least helpful for the people who do care. AG202 (talk) 20:31, 8 July 2023 (UTC)[reply]
This will likely be my last intervention in this thread.
You see, I care about those semantic arguments because they have an impact. Case in point: you're distorting my words. I didn't say I "care little about [slurs]" in general, I said I "care little about those slurs", i.e. the ones Equinox used in his edit summaries. While they're in bad taste, I find them to be of little consequence. If Equinox were throwing slurs at users to actively and purposefully make them feel unwelcome, that would be another matter; but I don't think that's the case here (at least I haven't seen evidence to that effect.) If people choose to get offended anyway , much good may it do them. Two remarks here:
I'm saying "choose to get offended" because it's imo a mindset, but I can see why some people might approach this differently.
This doesn't actually mean I would vote oppose if someone requests disciplinary measures be taken against Equinox, since I do think it would have been better for him not to use said slurs.
As for the rest, I fail to see how the number of times it took for me to become an admin has anything to do with this, and I also fail to see what you referring to with "Knowing your past attitude and interactions with me and others on issues like these or other problems (re: “inclusionism”)". I don't remember discussing this topic with you or anyone else, though my memory might be failing me here. PUC – 17:41, 10 July 2023 (UTC)[reply]
@AG202 This is highly disingenuous argument. Citing a source does not imply agreement with a source, and arguing that it should is extremely foolish. If a term such as "tradthot" has a particular meaning among neo-Nazis, then the evidence for that meaning will de facto come from neo-Nazi sources. I am sure that I have pulled citations from thousands of authors over the years, and I have no idea what their opinions are on almost anything in most cases. - TheDaveRoss13:37, 10 July 2023 (UTC)[reply]
There is a standard of quality (also I'm almost certain that Fay Freak has agreed with those sources at some point) that comes with citing and quoting websites here. If we can cite a word without linking to a Neo-Nazi websites (which tend to be low in quality anyways), then that's much better. I don't even think about frequenting Neo-Nazi websites or adding Neo-Nazi words. The words in question are not related to Neo-Nazi communities at all if you actually took a look at them. Honestly, I'm very concerned that you're making this type of argument when the context was clear and already linked to. CC: @-sche, @WordyAndNerdyAG202 (talk) 13:48, 10 July 2023 (UTC)[reply]
There is not a standard of quality, see UseNet. Accusing Fay Freak of neo-Nazi sympathies is a probably hate speech according to your views on the matter above. I disagree that anything in this conversation is clear, but of the few terms I looked at some were likely alt-right or worse, and some were likely not, regardless, my point stands, citing a word from a source does not imply agreement with or support of that source. - TheDaveRoss14:38, 10 July 2023 (UTC)[reply]
I've literally gotten messages from admin about how this user is likely Nazi-aligned. The senses at Talk:smash had nothing to do with Nazis, yet Nazi websites were linked. They also said, "It would give them more street cred – neo-Nazis are also customers to put it bluntly, and here possible contributors", to which an admin had to tell them publicly that Nazis are not welcome as contributors here. Literally called themselves a Neo-Nazi expert in Talk:ᛦ, toning it down after they told that Nazis are not welcome on Wiktionary. Please spare me the BS that what I'm doing is hate speech, when even the user themselves has not objected to the matter and has embraced it. I do not wish to continue engaging with you, as it's clear that you're only here to detract and distract. AG202 (talk) 14:49, 10 July 2023 (UTC)[reply]
Admins are not gods, their opinions carry no extra weight. You should work on your ability to engage with thinking which is different from your own, the world is a much better place if you are willing to be slightly open minded. - TheDaveRoss14:56, 10 July 2023 (UTC)[reply]
I was. This is the second or third time in this conversation where you have trivialized and dismissed my thoughts and opinions on a matter merely because they didn't exactly match your own. We agree on so much, but because we don't agree on every single thing your conclusion is that my purpose in the conversation is to distract and detract. The logical conclusion of this manner of engagement is that it will always end with you against the world, perhaps you prefer that, but I prefer to find common ground and agreement, as this is a community-based project. - TheDaveRoss15:11, 10 July 2023 (UTC)[reply]
I've always tried to find common ground on so many topics on this project, but issues involving slurs and hate speech and Neo-Nazi sympathizers crosses the line for me. I refuse to give way for defending people who use offensive language and sympathize with people who would rather have me dead. I do not appreciate conversations related to semantic arguments or deciding if someone is truly a Nazi or not when they admitted it themselves. I do not appreciate being told that I myself am participating in hate speech for calling out the fact that someone called themselves a Neo-Nazi sympathizer and said that we should be more welcome to Neo-Nazi contributors. We may agree on so much, but if this is truly the hill that you want to die on, then that's a dealbreaker for me. If that is a problem with you, so be it, but it's surely not me against the world when there are others who have agreed with me here and have focused on the actual problem at hand. AG202 (talk) 15:16, 10 July 2023 (UTC)[reply]
There is an awful lot of mischaracterization going on there. It's fine with me if you don't wish to discuss this topic with me further, but I am utterly unconvinced of most of the claims you are making. - TheDaveRoss15:21, 10 July 2023 (UTC)[reply]
Why are you so determined to die on this hill? There's a huge difference between A) quoting neo-Nazi sources in entries for neo-Nazi jargon (e.g. Holohoax) and B) shoehorning neo-Nazi propaganda into random entries. The majority of readers would likely agree that B) is actively detrimental to the quality of a dictionary. People would stop trusting Oxford, Collins, etc. if they looked up cactus and found Daily Stormer bile among the usage examples. Which is precisely what happened with smash back in 2020. The sense in question was a synonym of hottie (i.e. "attractive person"). In the end, it doesn't matter whether the Stormer quote was inserted by 1) an actual neo-Nazi evangelising their ideology, 2) a 4chan-style edgelord trying to stir up trouble for the lulz, or 3) a free-speech warrior blindly committed to the most self-sabotaging interpretation of "Wiktionary is not censored." The effect of all three is functionally the same, in that it involves Wiktionary becoming a platform for the dissemination of hate propaganda to the detriment of its reliability, accuracy, and neutrality as a dictionary. I've tried to follow advice given to me years ago – by Equinox, I should note – to ignore everything Fay Freak says. Nothing is to be gained from giving a serious audience to what is, at best, tedious pseudointellectual bloviation, and at worst, neo-Nazi apologism poorly cloaked by a veneer of r/iamverysmart. WordyAndNerdy (talk) 00:58, 11 July 2023 (UTC)[reply]
How do you know that the people you suspect are neo-Nazi sympathizers aren't normal folks seeking knowledge about neo-Naziism, possibly for helpful or benign reasons or for mere curiosity? DCDuring (talk) 15:35, 10 July 2023 (UTC)[reply]
He doesn’t even care even if he knows for certain that folks are adequate and assiduous to maintain balanced and moderate views, even if he is presented the psychological factor that his opponent admitted to have it difficult to sympathize with anyone or form passion due to alexithymia. It’s perseverance effect maintained by continuous petitio principii: Somebody loaned a term from a group he opposed and is now a poisoned well. Nazis worked the same way. “The Jews are behind everything so it’s the Jews again,” likewise if I or Equinox make an example from neo-Nazis with something containing taboo language the mere formality suffices to reject the whole person, and his utterings, apparently because of not including sufficient trigger warnings or hedges not to affect the poor person. It’s an actual reason why one should avoid swearing, in order not to stir passion, but from a quarter of a century on the internet we have learned that people get offended by more and more things and have constructed false systems of victimization, which are worse than the causes of offence. Self-sabotage due to valuing oneself lowly exists, and is also the cause why people can feel injured by a bad nickname. Conspiracy theories also work that way. “They are only doing that to … us” – and in astrology the stars only move in reference to some naked apes; no, actually people here mix a lot of reason in case into their posts, and the question should stay where there are more, rather than speech policying, which becomes the main motivation for some people here—that is wrong. I am little concerned with and rather bored by being called slurs as opposed to someone being untruthful—that’s why the internet persona exists, to be objective. Supercilious it is of course from me then to assess offences of others by the same yard-stick, but the standard is not wrong and soothfast beneficial for our scientific pursuit, as outlined. Fay Freak (talk) 16:35, 10 July 2023 (UTC)[reply]
Your mileage may vary. I can visit Neo-Nazi pages and be entertained, by their systematic error. Like when one listens to Attila Hildmanns messages one already knows that everything related back to “the Jews” in a few minutes. This is certainly supported by me lacking neurotypicity. But on the other end it is still worthy of concern why you feel threatened by some louts being wrong in public. “Nazis everywhere!, nazis everywhere!” – it is, aside from being a song of The Yardbombs, literally Kremlin propaganda, half of their casus belli. They don’t tell us though why Nazis exist, lest one trace how their own hunts upon culture have created it: it is mostly reactance—humans are primarily driven by effects that eighty years ago one wasn’t equipped to consciously conceptualize, to exert one’s endbrain in avoidance of being driven by them. Reductio ad hitlerum is still a fallacy even if applied to known and outspoken Nazis. The Epimenides paradox applies to them and the internet is the perfect place for them to be wrong, and consequently corrected, while at the same time providing examples of illustrative material, not necessarily being wrong, even less likely so if the agenda or motivation is more clear than on average: as you might know, definition hails from finis(“fine, end”), hence extremists show us how language is stretched and abused. Appeasement does not work and encapsulation neither: Only understanding deconstructs the adversary. Thinking outside the box is necessary if conformity rests on fallacies of essentializing false dichotomy, so, as the modus vivendi of scientific investigation, a choice regular habit for me, who isn’t really drawn into social interaction in the first place, and there must be people who are more “based” for this thing to work: often ratiocination mixes with emotion and hence tribal marks again, when there is lacking respect for the effort based on pure tribal considerations. In this cataloguing people as some who “spread hate” use “slurs” or otherwise “associate” with hate-groups is fundamentally flawed and actually reinforces their patterns by means of ascribing excessive value to collective action, which is the very destructive force that created all known totalitarianisms. So paradoxically, by seeking to be more “inclusive”, doing it by a superficial, affective scheme, you support what you oppose—as in those cases of medical error or security breach where the treatment amplifies the ailment. What you suspect to do or have is not always what you actually do or have. Quaternary prevention is underrated. Fay Freak (talk) 16:35, 10 July 2023 (UTC)[reply]
Read through all of this and everything I'm able to view. Not interested in the OP. This thread has brought up a mountain of highly problematic behavior on the part of Equinox. My opinion regarding them has changed. Their use of degrading slurs and insults is more than enough to warrant desysopping and we should move forward with that now. Yes, they've made a really incredible amount of valuable contributions, yet that doesn't give a free pass to this type of behavior. If they launch another attack like this one, they should be blocked. Calling someone the c-word and telling them to go die in the same breath? Why on earth are we tolerating this, especially when it keeps happening?! Megathonic (talk) 07:24, 14 July 2023 (UTC)[reply]
@Chuck Entz Since the conversation has generally died down and it at least looks to me that there's a consensus for some kind of action, what action can before an action such as desyopping? (I'd like to avoid putting that to a vote due to tensions here) AG202 (talk) 18:36, 20 July 2023 (UTC)[reply]
Some time ago, I was asked, as an uninvolved administrator, to review this discussion, I gather to implement an appropriate reading of any consensus and closure thereof. Although it is now on the stale side, I intend to do exactly that in the next week or so. I think, however, that an initial point that has the general agreement of the community is that Wiktionary is a dictionary, not a venue for pressing social or political preferences, and that quotes or citations that appear to primarily be intended to press such preferences should be avoided. I would also note that, although it may seem counterintuitive, a sysop can be blocked—even for an extended period—without adressing the question of desysopping at all. bd2412T03:34, 4 December 2023 (UTC)[reply]
@BD2412: Is there any follow-up on this? You may not have gotten a ping from the IP, so I'm pinging you directly. It's important that there's some consensus on this matter. AG202 (talk) 05:06, 18 January 2024 (UTC)[reply]
I can't say that I am any closer to finding a consensus or resolution to this discussion that I was when I first looked at it. bd2412T22:51, 21 January 2024 (UTC)[reply]
Freedom of speech ONLY means freedom to say the stupidest most ignorant comments imaginable in your social context. Freedom to say smart things everyone agrees with is meaningless. It is the freedom to speak comments considered bona fide evil by the society you are in which is the most meaningful. --Geographyinitiative (talk) 00:03, 22 January 2024 (UTC)[reply]
Don't post off-topic screeds and jeremiads. This is not a government-run site and there are terms of service. If you want to post hate speech, make your own wiki and post it there. —Justin (koavf)❤T☮C☺M☯01:53, 22 January 2024 (UTC)[reply]
Wiktionary is first and foremost a project to build a comprehensive dictionary of the English language. Examples of the use of a word should first and foremost provide a context that highlights the meaning of the word, without adding distractions to that purpose. For example, it would be an unhelpful distraction to provide as a citation for valet the example: Daniel MacIvor, Cul-de-sac (2005), p. 73: "And flipping burgers and driving your fucking car around the block and valet parking it up your fucking ass". bd2412T02:07, 22 January 2024 (UTC)[reply]
I'm not sure what I should be doing with regard to populating category:Pali causative verbs. The spiel at the start of the category page says, "Use this only for separate verbs (as opposed to causative forms that are part of the inflection of verbs)". It appears to be meant to be populated by invoking {{causative}} in an etymology section or {{catlangname|pi|causative verbs}} elsewhere.
Now, in Pali, causatives, like passives, appear to sit on the borderline between inflection and derivation. Pali even has, both morphologically and semantically, double causatives, as in 'to make someone get a craftsman to repair the wheel'. (Some double causatives are semantically single, but I haven't reached that bridge yet.) Is it in order to categorise a verb as a causative and have a definition line defining it as the causative inflection of another verb from the same root? --RichardW57 (talk) 10:51, 2 July 2023 (UTC)[reply]
That's almost what I've been doing, except that I list them explicitly next to the conjugation tables, rather than tucking them away under 'derived terms'.
I think the answer to my question is that in definitions I should use {{causative of}} rather than {{inflection of|pi|<base verb>||caus}}. I will, however, need to make use of its undocumented parameter |nocat=. I'll take this up at the Grease Pit. Thanks for showing me the parallel. --RichardW57 (talk) 20:42, 2 July 2023 (UTC)[reply]
@Maas555, Rua, Thadh, Tropylium I'd like to hear your feedback on some possible changes to our Proto-Finnic coverage:
The (back) vowel *ë, now more often considered part of Proto-Finnic proper than not (contra the traditional theory that there was only an *e). (To my understanding, this notation isn't valid UPA, which would be *e̮, but that form appears to be used less.) YSuS, now the closest thing we have to a Proto-Finnic reconstruction database, uses it consistently ([11], [12]).
If we do split these, there may also be a case to separate Northern Finnic into its own subfamily with at least the sound change *ë > *e, although the levels may be too close and just result in duplication.
Notation of semi-long consonants. The most common way seems to be e.g. *k̆k, but this is inconvenient. I think apostrophes as we do now (*k') are fine here, but there could be a case for generating and showing the other forms automatically with the headword template.
Lemmatization of verbs. Currently we use the *-tak infinitive form for this, but most other sources use the verb stem. Both approaches have their pros and cons:
*-tak
Matches most descendants
Avoids mixing up verbs with nominal roots
Verb stem
Matches with Proto-Uralic
Sidesteps the issue of different forms as lemmas in Estonian and Finnish
I'm fine with the *ë thing and I don't have anything against the geminate proposal, but I prefer lemmatising Proto-Finnic verbs at infinitives. Thadh (talk) 13:36, 5 July 2023 (UTC)[reply]
In fact, Northern Proto-Finnic might make more sense than I thought. There are plenty of words only found in Northern Finnic and a few more changes (like *ai-a > *ei-a). There would still be quite a bit of duplication, but probably not as much as I first assumed. If we do not create a separate proto-language, we could at least show it in the descendants and generate a reconstruction to display. — SURJECTION/ T / C / L /21:30, 9 July 2023 (UTC)[reply]
Subgrouping would take a separate discussion really. North Finnic as it stands is mainly distinguished lexically, while the phonologically most significant split is South Estonian vs. Livonian vs. Gulf-of-Finland Finnic (*ai > *ëi ~ *ei happens here; I've called it "Core Finnic" at times); arguably also, Finnish-Karelian-Ingrian is on the net even further distinct lexically than it would be with Ludian-Veps included. Under Wiktionary's current data model and current stock of Finnic editors it seems clear we wouldn't want to distinguish four separate subfamily stages though.
Semi-long consonants are only allophonic and they also remain that way pretty long… they might not strictly even need to be indicated!
Verb lemmata: we could split the difference by instead reconstructing *-ta infinitives also for Proto-Uralic /j
I'm getting annoyed with this user, who has been repeatedly warned about adding semi-irrelevant pages to categories but continues to do it. All his contributions involve adding pages to categories, often wrongly or irrelevantly (e.g. reacción en cadena = "chain reaction" getting added to "Nuclear warfare"; in this case he edit warred to get it this way, and I just reverted again). He seems obsessed with categorizing (on his Wikipedia page he is a self-admitted Aspie, probably has something to do with it ...). Should we block him for awhile to send him a message? Benwing2 (talk) 08:51, 4 July 2023 (UTC)[reply]
He's already been blocked five times for edit warring and ignoring warnings and discussions. The few responses to warnings on his talk page show no understanding of why certain edits are unwelcome. I think a longer block is appropriate. Ultimateria (talk) 18:37, 4 July 2023 (UTC)[reply]
Formerly, Template:en-past of displayed simple past and past participle on a single line. However, it's usual when displaying forms that correspond to multiple inflections to list each inflection on its own line. This is consistent with how all other languages work, e.g. in Portuguese falamos; German and Romanian gute; etc. User:Ioaxxere doesn't like this and thinks we should display it on a single line. What do other people think? Benwing2 (talk) 22:22, 6 July 2023 (UTC)[reply]
For one thing, no one calls them "inflections"; most English speakers probably wouldn't even recognize the word. Second, having three lines bloats the page for no obvious benefit. Third, it's inconsistent with our headword template which groups "simple past and past participle" together. Finally, the template doesn't work at all with quotations, usage examples, etc. (see left#Verb).
One line: Expanding on Ioaxxere's comment, having multiple lines doesn't work with explanations either, such as the nautical sense at payed, which is "{{lb|en|nautical}}{{en-past of|pay}}{{gloss|to cover with a waterproof substance}}" and is meant to be displayed as a single paragraph. If I were to use{{inflection of}}for it, I would use an ampersand tag or an 'and' tag, not a semicolon tag, to keep the inflection information on one line. (The semicolon tag in {{inflection of}} is very nice, but is a pitfall when the definition gets complicated.) --RichardW57m (talk) 10:40, 7 July 2023 (UTC)[reply]
I apologise for what feels a pedantic question. The ever-increasing number of scrapers may, however, make it relevant.
The documentation of{{root}} says its invocation should be placed at the first line of an etymology section. Can one however use it elsewhere, e.g. under a PoS header, if there is no relevant etymology section? Or should one create a visually empty etymology section just to house it?
The use case is the non-finite parts of a Pali verb that are nominally formed directly from the root rather than the present stem. I am currently giving them their own entries, partly as a way of housing their quotations. When I document a Pali root, I want to be able to gather them together using the category for the root. However, there rarely seems any point in writing an individual etymology section for most such parts. --RichardW57 (talk) 00:46, 7 July 2023 (UTC)[reply]
@Romanophile @Koavf Thanks for the attention. I personally dislike the curlies, and I have deleted them under my usual principle of "DRY" (don't repeat yourself) or minimalism. But it seems that, as Koavf suggests, there might be cases where it matters. Even so: we need to find a better solution than including a "curly" version of every apostrophe headword, which is insane. Equinox◑00:42, 14 July 2023 (UTC)[reply]
We have 107 admins currently per Special:ListUsers/sysop, which IMO is a fair amount. The following admins have contributed little or nothing in 5 years:
Per Wiktionary:Votes/pl-2017-03/Desysopping for inactivity, admin status can be removed "without further ado" for any admin who hasn't used their admin tools in 5 years, which probably applies to all of the above (although I didn't check their recent contributions to make sure). I recommend de-sysopping them without prejudice, i.e. they can have their admin rights restored upon request. Benwing2 (talk) 03:20, 9 July 2023 (UTC)[reply]
Five years ago is July 2018, so on the basis of the vote, there is absolutely no doubt that the users who have not contributed in any way since then should have the admin flag "removed without prejudice to reinstatement":
Since we voted specifically to desysop users who have not "used admin tools" in 5 years, the following users who have contributed in some (very minimal) way - mostly just a scattering of edits - but haven't used administator-specific tools since July 2018 should also have the flag removed:
Yes, it seems he was elected in a low-stress era and it didnt matter to the voters that his edit count was already well into decline by that time.
In theory, I like the idea of restoring admin rights upon request, but strictly speaking, our policy does not offer that option, and some people might object if one of these long-inactive people comes back and is promoted to admin again without further discussion. —Soap—08:39, 9 July 2023 (UTC)[reply]
One of the rationales for desysopping inactive users was the reduction of the risk of hijacked accounts with sysop powers. Has any admin account been so hijacked at English Wiktionary, at any Wiktionary, at any sister project? DCDuring (talk) 16:32, 9 July 2023 (UTC)[reply]
There was an incident in 2012 on English Wikipedia when an inactive admin came back and started editing the Main Page, seemingly unaware of its importance or of what he was doing. The details were not made public ... we don't know if the account was compromised or if he somehow forgot where he was. However, he had only been inactive for about six months, not five years, so strictly speaking this isn't the same situation. I don't know if other inactive admin accounts have been involved in similar incidents. —Soap—17:40, 9 July 2023 (UTC)[reply]
I'm not sure about hijack situations but it seems risky to me to have so many inactive admins; password requirements have gotten stricter over the years so people who haven't logged in in awhile might have lax passwords that are relatively easy to guess. Benwing2 (talk) 17:56, 9 July 2023 (UTC)[reply]
That rationale, while valid, isn't particularly important since it is easy to undo anything an admin can do, and we have plenty of admins active around the clock to take care of such things and summon a steward to remove tools. I do support keeping the list "clean" for many practical reasons, not least that it is helpful to have a somewhat accurate count of how many admins are around. - TheDaveRoss13:22, 10 July 2023 (UTC)[reply]
I strongly disagree with this thinking. I know you've expressed this opinion before, but a malicious user with admin access can fuck up all sorts of shit in a way that would take a lot of work to undo, esp. if they're knowledgeable about bot operation (there's no way to prevent someone with admin access from using that account in conjunction with a bot). Benwing2 (talk) 19:06, 10 July 2023 (UTC)[reply]
Malicious users who know how to run bots can cause lots of damage without being admins, they just don't seem to do that. It is fine if that is why you choose to prioritize removing admin rights from inactive admins for that reason, I agree that we should just for a different reason. - TheDaveRoss21:08, 10 July 2023 (UTC)[reply]
We've talked about this in the past, and (as a half-arsed past sysadmin) yes, I think we should minimise the powers of anybody who doesn't seem to be logging in. On the off-chance that anyone doesn't understand why: it's a "surface area" issue. If somebody decides to attack our project, they might well start with some old unused admin account, and try to guess the password with pet names or wife names or etc... They might get in, or not. But if we have lots of essentially "spare unused" accounts with admin rights, we are giving them much more chance to do so to do harm if they do get in. Equinox◑00:49, 14 July 2023 (UTC)[reply]
The current policy should be enforced; if they meet the criteria, they should be desysopped without a vote or any further fuss. (Personally, I wouldn't mind decreasing the required 5-year timeframe.) If any of the individuals return to the project after being desysopped, the policy doesn't automatically grant them admin rights back; they've been gone for too long and must go through the nomination process again. A message can be left on their talk pages informing them of the reason for the desysop and that they are welcome to reapply for adminship should they return. Megathonic (talk) 02:37, 14 July 2023 (UTC)[reply]
@Benwing2: let me throw in another factor that needs to be considered: COVID. A couple of years starting sometime in 2020 should not be treated as just any old random stretch of time. During that period, everything changed and people had other things to deal with. Whether that means that everything before 2020 should be treated like a prehistoric epoch too remote to connect to modern times, or whether the years in question are lost time that should ignored in our calculations is up for discussion- but it should definitely be discussed. Chuck Entz (talk) 23:37, 14 July 2023 (UTC)[reply]
@Chuck Entz I'm really not sure why COVID has anything to do with it at all. None of the above people have contributed significantly since 2018, which is well before COVID, and if anything I'd expect more contributions during the COVID years rather than less, consistent with the boost that online everything got during the first pandemic year or two especially. IMO this is a not insignificant security risk, and the policy is clear, so we should not be making excuses to not implement it. Benwing2 (talk) 23:49, 14 July 2023 (UTC)[reply]
The way I see it, adminship is a service to the community, not a perk or honour for a user. Someone should only have that position if they can perform the service. There might be any number of reasons why they can't, but there's no use in inactive users being admins, regardless of the circumstances. If they become active again, then it's not a big deal to have a vote to reinstate them. Andrew Sheedy (talk) 20:43, 16 July 2023 (UTC)[reply]
I also don't think the Covid pandemic has any bearing here. The policy is to remove roles without prejudice, we should just follow the policy. If it turns out that this policy causes problems for some reason (and they would be easily remedied) we can alter the policy. - TheDaveRoss23:01, 16 July 2023 (UTC)[reply]
This has come up several times before. I don't understand the apparent reluctance of our bureaucrats to act on our policy. I also don't think rights should be reinstated automatically if a desysopped admin does come back, but that's another issue. PUC – 20:52, 16 July 2023 (UTC)[reply]
This echos my thinking. I find it annoying that we're even having this conversation. The line "without further ado" was put there so we didn't have to keep litigating this. What should have happened:
Step 1) Those who haven't used the admin tools in 5+ years are brought to attention.
Step 2) A bureaucrat verifies the list and sysops those who qualify.
All the current "Chinese spelling" entries are created by User:Lvovmauro, solely based on one source 英語集全 [Chinese English Instructor] ({{RQ:cpi:Instructor}}). Generally speaking the chracters used in Chinese sources depend heavily on the Chinese dialect used by the author, since they represent different sounds, and this often differs between different books even when they're based on the same dialect as they may employ special characters or systems to represent the foreign sounds in CPE. After glimpsing over the preface and some of the pages in the Chinese English Instructor, it appears to me that even within the same work the characters are used inconsistently, e.g. piecee can be written as either 卑時 or 卑士, while gentleman is written as 毡地文 although ⿰口毡地文 represents sounds that are closer to the English pronunciation.
I reckon that we should not allow Chinese spelling of CPE entries, unless there is a spelling that is (at least somewhat) consistently used between works. Nevertheless, the Chinese sources are a valuable resource in attestation, so I think they should still be allowed to be used as quotations on the corresponding Western spelling of the CPE terms.
Hmm I can say for sure that the Shanghai-Ningbo pidgin is certainly not standardised and pretty much every single book uses a different romanisation, but how it should be treated is something I'm not certain I would know how to deal with. The simplest option would just to not allow them but the sinographic sources do have some significance. The first thought I had was to make a module that includes some of the major publications and what forms they use (similar to dial-syn) but whether that is something reasonable would be a different question (I mean, we haven't even come to clear consensus as to how we ought to treat the Shanghai-Ningbo pidgin, uh, thing) — 義順 (talk) 08:52, 9 July 2023 (UTC)[reply]
@ND381: Thanks for the reply. I think listing all the forms and variants is going too far, since (a) we only have access to a handful of sources, and we don't even know how many there are out there; (b) different dialects (?) of CPE may correspond to the same English word but spelt differently in the Western spelling (e.g. "sell" is <sellum> in Canton but <sell> in Shanghai), plus (c) putting them in the quotations already does the work and at the same time provides better context as to how the word is used. – Wpi (talk) 10:01, 11 July 2023 (UTC)[reply]
As I understand it, there is a strong desire that alternative forms that are synonymous with a 'main form' should be categorised as forms. For example, the Pali prefixes paṭi-, pati- and pacc- have no discernible differences of meaning, but the preferences for them are determined, to varying degrees, by the phonetics of the rest of the word. The choice between the first two is fairly free, and therefore I am too uncomfortable with treating the third as a phonetically conditioned variant of the first of them, and link to it in etymologies. I have accordingly made the first a prefix, and made the other two 'prefix forms'. So far, so good, if a little arbitrary.
Note, however, that I revolt at doing this for inflected words that have their own inflection tables, and would also balk at doing this when headwords contain inflections and similar.
@Benwing2 However, when I then look at the category for the alternative forms, namely Category:Pali prefix forms, I find its contents described as "Pali prefixes that are inflected to express grammatical relations other than the main form". That is definitely wrong. Most if not all Pali prefixes are not modified to show grammatical relations. The offending text is automatically customised translingual boilerplate, and shared across parts of speech. So, what to do?
Hold one's nose, and do nothing.
Globally generalise the text to something like "Pali prefixes that are inflected to express grammatical relations other than the main form, or are alternative forms"
The text was already dubious - precision has been sacrificed for snappiness.
Fix them on a language by language and part of speech by part of speech basis. For example, the notion of Pali suffixes being inflected is quite reasonable.
@Benwing2: I did mean templates. The usual practice is one headword template for each PoS-like category. However, while the categories can be mass-produced, the headword templates are generally hand-crafted. There might be some simplification by instead adding a variant flag |variant=, but then the language-specific headword module would still need editing. --RichardW57 (talk) RichardW57 (talk) 20:34, 9 July 2023 (UTC)[reply]
@RichardW57: I don't really understand your objections. For suffixes we'd need to make the distinction in any case, even for Pali, where inflected forms of suffixes are different from alternative/variant forms of suffixes and both can exist. Furthermore, the two categories have different parents; suffix inflected forms are non-lemma forms whereas suffix alternatives/variants are lemmas. Adding a new template or param seems hardly a big deal. Benwing2 (talk) 20:38, 9 July 2023 (UTC)[reply]
As an example that is none of my doing, Latin ir-, il- and im- are prefix forms and do not occur in Latin etymologies, while their English descendants are prefixes and occur in English etymologies. --RichardW57 (talk) 03:26, 10 July 2023 (UTC)[reply]
@RichardW57: We don't need hundreds of templates. We don't even need one. You can write {{head|pi|prefix variant}} (using your terminology) and all will be well. Also you never addressed my concern that a distinction needs to be made in any case between 'suffix forms' and 'suffix variants', if you prefer that terminology. Benwing2 (talk) 03:51, 10 July 2023 (UTC)[reply]
@Benwing2: So not true, though it was only recently that I noticed the failure to categorise by script. It is a pity, though, that writing this gets the transliteration option wrong by default. And, I think it at least complicates the categorisation of Pali verbs by present stem formation ('conjugation').
I said I didn't like your suggestion - I would rather simply distinguish lemmas and inflected or derived forms (except in so far as they themselves become lemmas) - not that the distinction needed to be made. --RichardW57 (talk) 06:31, 10 July 2023 (UTC)[reply]
@RichardW57 Ughh, I'm getting frustrated. Honestly, you are one of the most annoying people I've ever tried to carry on a discussion with. You criticize without presenting alternatives; you make opaque suggestions without giving clarifying examples; you write walls of text, often full of irrelevant asides, and expect me to wade through and figure out the gist of it; and you misunderstand a lot of the status quo leading to your creating incompatible language-specific ecosystemlets. I also sometimes feel you resist and act contrarily just for the sake of doing so, not for any sound, principled reason. Finally, I have never seen you thank me for any of the things I've done on your behalf (apologies if I've overlooked one). This sort of behavior makes me less and less interested in helping you any further. (Pinging again @RichardW57 as my signature got botched.) Benwing2 (talk) 07:11, 10 July 2023 (UTC)[reply]
A lot of what you do goes off at half-cock. The way you treated Pali conjugations missed the point that they are largely etymological in nature, with much less connection to the type of conjugation than one sees in Latin or even Old English. I still think it needs rework, but thank you for showing me how I can do it myself. Incidentally, having some of the language-specific infra-structure modules show up amongst the language-specifc modules should be a great help to those who are happy to build their own structures using their knowledge of the languages. Thank you for that.
Adding |caus-of= for Pali uses of {{inflection of}} may have looked like assistance, but on its own it would just make things more difficult; invocations that work for a single script don't work well when one has the concept of light ('subsidiary lemma') and full ('primary lemma') entries and wants the synchronic derivation path in the script of the entry with cross referencing to the full entry at every step. We have discussed a way of making this work for Sanskrit, but until that is working, I think it is, alas, actually a hindrance rather than a help.
Blindly suppressing, unannounced, the suppression of default transliteration of non-Roman Pali headwords caused a lot of bafflement. --RichardW57 (talk) 12:48, 10 July 2023 (UTC)[reply]
@RichardW57 I have no idea what your last sentence refers to. As for the remainder, I don't want to implement something completely bespoke for Pali in a general module like Module:form of. Whatever is implemented needs to be done in a general way that works and has use cases for other languages. I've asked you to help brainstorm this but haven't gotten any support. Benwing2 (talk) 21:12, 10 July 2023 (UTC)[reply]
The 'last sentence' refers to setting the customisation of transliteration in the headword support in Module:links so that if the transliteration was not given, no transliteration would appear. The problem is that transliteration does not always yield the Roman script equivalent. Incidentally, transliteration is one of the reasons that {{head}} is awkward to use for Pali, even though it allows full control. It's counter-intuitive that the usual case needs more parameters than the unusual case.
I don't know where you think brainstorming help is needed. We've discussed how to support multiple scripts - I though we had covered what should be achieved. Were you looking for suggestions on how to achieve those effects? I had assumed that you would happily work that out for yourself.
If you were looking for more parameters like |caus-of=, you should have said so. What you were offering doesn't actually help with Pali as it is, because such chains of quasi-inflectional derivation for a single script won't translate easily if one tries to put the fuller derivations one click away. The reason for chains is the intention that the senses of the base form will undergo corresponding transforms in a manner that the reader can evaluate for himself. At present Pali is, in my opinion, overly light on quotations. And if I do come to find them useful, you have given an example that we can build on easily as we come across occasion to use them. --RichardW57 (talk) 22:40, 10 July 2023 (UTC)[reply]
Don't confuse topical categories with lemma/non-lemma categories. Topical categories are for navigating between terms that have something non-lexical in common and are outside of the normal organization of entries. If I click on a category link from petit pois, having both snow pea and snowpea in the category is just clutter, especially since snowpea has nothing but a link to snow pea. I generally don't include alternative forms in the same category as the main form, though I may include synonym-of entries if they're different enough that someone might not know that they're the same thing. Categories for terms that have something lexical in common are something else entirely. Chuck Entz (talk) 04:29, 10 July 2023 (UTC)[reply]
I should add, this is the majority handling of prefixes and suffixes, at least in the Romance and Germanic languages I've examined in depth. Benwing2 (talk) 06:29, 10 July 2023 (UTC)[reply]
I wonder if @Urszag would care to explain why he's converted at least one Latin prefix (ac-) from a prefix to a prefix form. Would there be any objection to my converting it back? --RichardW57 (talk) 21:08, 10 July 2023 (UTC)[reply]
I was simply not aware that "The intention behind 'FOO forms' where FOO is a part of speech is that the "forms" of FOO are inflected forms" applied to prefixes, since Category:Latin prefix forms already contained things like col- when I made that edit. I don't have much objection to making a consistent change to the categorization of such forms, although I do find categories like Category:Latin prefixes slightly more readable when they aren't full of contextual allomorphs that have no difference in meaning, so it would be nice if there was some alternative way to deal with that. Therefore, Benwing's suggestion of using something like CAT:Pali/Latin alternative prefix forms or CAT:Pali/Latin prefix alternative forms sounds good to me.--Urszag (talk) 22:35, 10 July 2023 (UTC)[reply]
@Urszag: So I would ask the question of @Rua, though we may have to wait long for an answer. Of course, for Latin, one solution would be to customise the description. And, @Benwing2, where is the notion of 'prefix form' defined? Does it solely lie in the boilerplate for the category text, and if so, how long has it lain there? The glossaries seem distinctly misleading on what a form is. --RichardW57 (talk) 22:59, 10 July 2023 (UTC)[reply]
@RichardW57 I don't see why you're harping on this point. All "POS forms" are non-lemma inflectional forms; that is standard across Wiktionary, not just in the glossary. Trying to redefine this on a language-by-language basis is going to lead to endless confusion. A prefixing language can potentially have inflecting prefixes, just like a suffixing language like Pali has inflecting suffixes. Benwing2 (talk) 23:27, 10 July 2023 (UTC)[reply]
Could someone please translate the second sentence of this into English, please. I am particularly having difficulty with the word 'All' (does it mean 'all that don't contradict the utterer's thesis'?} and the phrase 'not just in'. All I could find in the two glossaries, in WT:Glossary, was 'form-of entry – A non-lemma entry; that is, an entry that contains a term that is a modified form of another term'. Most, possibly all, of the 21 entries in cat:Latin prefix forms are prefixes modified by the phonetic environments, with no conversion I can discern of phonetic conditioning becoming grammaticalised conditioning.
Now, one might suspect that cat:Form-of templates would help explain the entry in the glossary, but that includes {{alternative form of}}, which does not support the inflectional interpretation. Perhaps the documentation of template {{head}} will help; but for |2= it merely says 'part of speech', which sits ill with the prohibition of 'noun form' as a part of speech in WT:EL. The documentation references WT:Part of speech, but that sheds no light on the signification of 'noun form' etc.
There are two categories of prefix forms in Cat:Prefix forms by language which are grammatical modifications - both are subcategories of prefixes that have undergone a Celtic mutation. The rest, including Indonesian, follow the model used for Latin. It looks as though the model of Latin has only also been implemented systematically for Indonesian.
Now, if 'communities' have been coordinating internally, it looks as though 'form' has been interpreted conveniently, with no acknowledgement of the claimed rule that 'alternative of' preserved the 'part of speech' as used with {{head}}. --RichardW57m (talk) 15:17, 11 July 2023 (UTC)[reply]
@RichardW57 Your question displays the confusion quite well. Topical categories aren't based on the concept of lemmas at all, though adding nonlemmas is a bad idea for practical reasons. Category:Pali lemmas is supposed to contain all Pali lemmas, but Category:pi:Religionshould not contain all Pali lemmas related to religion- just the more representative ones. Chuck Entz (talk) 12:35, 10 July 2023 (UTC)[reply]
So in his question, what did @-sche mean by 'the lemma'? Or are people omitting the italicisation when writing 'the lemma' meaning the 'main lemma'? --RichardW57 (talk) 12:51, 10 July 2023 (UTC)[reply]
In the context of "alternative forms vs lemmas", I mean by "lemma" the form we lemmatize and put the content on (the main spelling, as contrasted with the others which are just {{alternative form of}}s). (In the context of "inflected forms vs lemmas", the lemma is the form we lemmatize and put the content on, the "uninflected" form, which can be an alternative form.) - -sche(discuss)15:15, 11 July 2023 (UTC)[reply]
Thank you for the explanation. Does this mean that you would not expect an alternative form of a lemma to be categorised as a non-lemma? --RichardW57m (talk) 17:04, 11 July 2023 (UTC)[reply]
Following a debate, largely with @Kwamikagami, over whether the entry for the 'translingual' Chakma letter 𑅇 should remain, we now have instead an entry for the Pali Chakma letter 𑅇(va). This raises several questions:
1. Should the Chakma script letters used in Pali remain translingual, or should we create Chakma and Pali entries for them instead? --RichardW57 (talk) 20:02, 10 July 2023 (UTC)[reply]
2. As to letters used in Pali that have do not have translingual entries but do have entries for other languages, should they have Pali entries? I can fairly readily clone them from the entries for the translingual letters of the Tai Dharma script. --RichardW57 (talk) 20:02, 10 July 2023 (UTC)[reply]
The answers to Questions 2 and 3 affect whether further work should be done to enable function alt() exported from Module:pi-headword and accessed via {{pi-alt}} to support letters. Currently letters are being linked to as Pali letters, but some links may need to be tweaked so as to link to a translingual letter. The handling could get a bit complicated for non-Burmese Burmese script letters primarily used in Pali! --RichardW57 (talk) 20:02, 10 July 2023 (UTC)[reply]
I may be wrong here, but my understanding is that 'translingual' should be used for letters or symbols with a substantial presence across languages. So for example the letters of the Armenian alphabet are all listed under 'Armenian', despite most having once been used for Kurdish and Turkish. We could add additional sections for their use in Kurdish and Turkish, but AFAIK that wouldn't justify creating an additional 'translingual' section. Greek letters have both a translingual section for their international use and Greek sections for their use in Classical and Modern Greek. In the case of 𑅇, AFAICT it's only used for Pali, not even being part of the Chakma alphabet, so how would it be "translingual"? As for the other letters of the Chakma script (i.e. the Chakma alphabet), I would think we'd want them listed under 'Chakma' and then, if you wish to, an additional section for Pali would flesh out usage. Pali and Sanskrit are written in a lot of scripts, and we often don't list those languages in the articles on the letters of those scripts, so I think it would just be a matter of how complete you wish to be.
BTW, I got here because I've been removing unjustified claims that letters or symbols are "translingual". For some letters, we have no evidence that they're used for any language at all, yet they're listed under [mul] "translingual". (I've changed some of those to [und] "undetermined" and tagged them for deletion. Unicode has created bogus letters that have never actually been used, except as mentions of their presence in Unicode, so I'd like to see some evidence that a letter exists, and have an actual definition for it, rather than blindly copying from Unicode under Wiktionary's voice.) kwami (talk) 20:31, 10 July 2023 (UTC)[reply]
Some of the current decisions are unclear. Why are Burmese letters generally translingual, while Thai letters are only per-language? I'm inclined to let Pali rely on the multilingual entry where it exists, but otherwise go for having a Pali letter. There are, however, some Pali letters from Burma that are difficult to attest, and for which I suspect a band of enthusiasts managed to sell Unicode the idea that they were different from other people's letters and were 'used'. --RichardW57m (talk) 16:11, 11 July 2023 (UTC)[reply]
Generic use of 'translingual' is uninformative and in many cases we might as well not bother with an article at all. For major scripts like Latin, Cyrillic, Arabic and Ethiopic it's clear enough, but most scripts are used for one principal language plus a few minor ones that happen to be in the same geographic area.
As for Burmese, many of the "translingual" sections are mislabeled: They're simply Burmese. They even have pronunciations, which generally isn't possible for translingual entries, and the pronunciations are specifically Burmese (both the sound files and the IPA transcriptions, though the latter are miscoded as [mul]). There's an alphabetical order, again specific to Burmese. And in most cases there's no Burmese entry, even if there are sections for Mon or Pali! For other letters, there's no information at all apart from the Unicode box, and those should be deleted. If there is a translingual section in an article, it should contain some actual information that would be of use to readers, beyond usage for a specific language. kwami (talk) 18:03, 11 July 2023 (UTC)[reply]
@Kwamikagami Please convert all the single-character entries you've marked for speedy deletion to use {{rfd}}. Speedy deletion is only to be used for non-controversial deletions, which these are evidently not. The above discussion should be part of the RFD discussion. Benwing2 (talk) 07:36, 13 July 2023 (UTC)[reply]
@Kwamikagami: I think you've misread. Pali has a lot of stems (which is the form we use for lemmas, where identifiable) that consist of a single consonant with the vowel 'a', which goes unwritten in abugidas. Additionally, Mon has a fair few one character words, though certainly most of the Mon alphabet has its own letter. Now, Pali letter names, which are declinable nouns, have stems that are generally indistinguishable from the letter, and are prone to being used as bare stems. However, we haven't entered many of them, and orange link removal has not been pursued, so I believe there are currently none. --RichardW57m (talk) 09:02, 13 July 2023 (UTC)[reply]
Now, there has been a recent trend to claim that the various nations have their own version of the Pali alphabet, and Unicode has often accepted these claims, so there may be a flood of Pali letters to come. But not amongst the empire-wide characters. --RichardW57m (talk) 09:02, 13 July 2023 (UTC)[reply]
Standardising dial-syn character usage for v-/f- Negators across Wu - 勿/弗, 𧟰/覅/𫸻, 㬟/朆/ ⿰弗曾, 甮/⿱弗用, 𣍐/⿰弗會 etc?
There are a bunch of f-/v- negators across Wu, and currently in the dial-syn for 不, the v- negators seem to mostly be represented with 勿, and the f- negators 弗. However, partly because this explicit separation does not exist in most regions' orthographies (and since preferred characters vary across varieties), there seems to be a lot of inconsistency on lists for vocabulary items such as 不用, Module:zh/data/dial-syn/不是, 別, 不會 and such - certain f- negators are written with 勿 in both compound forms and ligature characters, e.g. 勿會 and 𧟰 instead of 弗會 and 𫸻 are currently used for Suzhounese /fəʔ ɦue/ and /fiæ/.
As we add synonyms for more localities going forward, I fear that this inconsistency might pose some problems. The two most straightforward solutions I see right now are:
write all with 勿, OR
standardise using 弗 for f- negators and 勿 for v- negators
though both introduce certain problems which I feel might be worth discussion.
With the first one, if we opt for 勿 for everything, there leaves us no immediate way to tell which localities have f- and which have v- type negators. This might not be ideal since whether the negator has a voiced initial is arguably an important distinction to note when comparing Wu varieties, and certain localities use both a f- and a v- variant (though having checked sources for both Chongming and Huzhou the two's usage seems to be complementary and dependent on the negator's position in a phrase). This does make it less of a fuss to standardise everything, and certainly with ligature characters such as 𧟰, 甮 or 𣍐, the 勿 variants are much more likely to have proper unicode support than the 弗 variants, but has problems.
The second approach (which I personally tend towards) does capture the difference between f- and v- negators, but then there are still certain problems. For one, certain composite characters e.g. ⿱弗用 aren't supported by unicode. Depending on the locality certain 弗-ligatures might also be orthographically quite rare - it is very rare to see 𫸻 or ⿰弗曾 in written Suzhounese, Wuxinese, or Huzhounese, and much less in their typesets. If we were to use the 弗-ligatures for those localities, I'd imagine that it might be necessary to add some usage notes on individual pages to address that the 勿-forms might be orthographically preferred/more common with certain localities, and that 弗 sometimes is also used in localities with v- type negators, under certain circumstances.
Also on a semi-related tangent - is there any reason why all the information for 覅/朆 are at 𧟰/㬟? I do understand how it might make more sense to have the latter two as standard forms, though I do feel that the former, being quite a bit more prevalent in Ming-Qing literature, might perhaps deserve their own pages? — Musetta6729 (talk) 07:46, 11 July 2023 (UTC)[reply]
There's a method to indicate different etymologies of the same character in the dial-syn table (see Template:zh-dial/documentation#Etymology), which should be a better solution than what we have now, but I do think this still doesn't address the problem properly as (a) there isn't an exact correspondence between f-/v- and the etymology numbers, and (b) it is unclear whether or not they consititute two distinct etymologies (from what I gather from your description, f-type and v-type negators derive from 弗 and 勿 respectively, but I don't really know if that's really that case).
Sorry also for the confusion - in terms of etymology, both of these types are from 不 according to Pan (2002), and v-type negators are most likely voiced variants ultimately developing from the f-type negators. In practice the two characters are somewhat mutually interchangeable, and sometimes used by certain people to distinguish between the voiced and unvoiced forms. My concern is mainly about whether the two phonemic forms should be marked with different characters or entries in dial-syn tables, despite sharing the same etymology. — Musetta6729 (talk) 11:05, 11 July 2023 (UTC)[reply]
This is an interesting discussion and I do think that distinguishing them by sound could be useful, but I worry that it would be too prescriptivist rather than descriptivist. My understanding is that if written forms are generally written with 勿 (as you say, 弗 can be quite rare), then Wiktionary should reflect that. ChromeGames (talk) 17:36, 11 July 2023 (UTC)[reply]
That is true - though the use of 勿 (instead of 弗) for checked f- type negators seems to be more common use in (non-scholarly) Northern Wu orthography than it is in Southern Wu, so maybe this acknowledgment could be achieved through usage notes on individual pages (if needed)?
I have asked for some brief opinions from some experienced people/speakers of Southern Wu. One feedback I have received was that certain impressions of 勿 being more reminiscent of "formal" texts such as books of family disciplines, and 弗 being more "colloquial" and tending towards emphasising "regionality". The Jinyun and Qingyuan Wu speakers who replied to me also responded that they mostly see 弗 being used for their varieties' f-negators (certain people do apparently use 伐 for the v-variant also existing in Jinyun, but no mention of a 勿 is made). However, they generally seem to agree that using 弗 for checked f- negators and 勿 for checked v- negators is a common approach amongst linguists dealing with Southern Wu (which is not absent from Northern Wu ime).
As to the 弗-ligatures I've not managed to get much feedback on - certain Southern Wu sources do seem to use them quite liberally, though I'm unsure of their role in traditional/common Southern Wu orthography, so maybe some care needs to be taken with those? — Musetta6729 (talk) 07:08, 12 July 2023 (UTC)[reply]
I would just split the negator into 勿 and 弗. If a character is unrenderable we can always split the ligature or find an alternative spelling or something — 義順 (talk) 00:14, 14 July 2023 (UTC)[reply]
@Ioaxxere recently added a new sense 1 to the entry acid test: "(chemistry) A test involving the reaction of an acid with another chemical". This sense is in the OED (and is the original sense of the term), but since it is sum-of-parts I think it can be omitted and just mentioned in the etymology ("[f]rom the use of nitric acid to test if a metal is gold; true gold is not affected by the acid"). Should such a sum-of-parts sense be included in entries of this kind? (The entry is WOTD for 15 July 2023.) — Sgconlaw (talk) 21:54, 11 July 2023 (UTC)[reply]
I actually had no good idea what an acid test was in chemistry. I suppose I believed that it would be much closer the usual popular sense of "rigorous test". In addition, a literal sense could be a test of the acid, not of a second material. So, I would have needed that definition. DCDuring (talk) 22:17, 11 July 2023 (UTC)[reply]
I actually don't think it's sum of parts, either - "acid test" cannot mean "the test of an acid", which it could if the term was genuinely SOP. Theknightwho (talk) 14:45, 12 July 2023 (UTC)[reply]
This template requires one effectively to split any circumfix into a prefix and suffix. For example, the etymology for Georgian უკლასო has the following code:
Nothing about the displayed etymology suggests that we are dealing with a circumfix, as opposed to a prefix and suffix, except for the URLs linking to უ- -ო(u- -o, privative circumfix).
That the displayed etymology splits the circumfix into უ-(u-) and -ო(-o) is also synchronically problematic. This may be plausible as far as უ-(u-) is concerned, since there does exist such a prefix with negative and privative usages, but certainly not -ო(-o), which does not seem to exist as a suffix with any usage relevant to უ- -ო(u- -o), at least not in the modern language. There are other examples of Georgian circumfixes that cannot synchronically be split into combinations of prefixes and suffixes, and no doubt the problem exists for other languages as well.
Another problem is that whenever an etymology tries to explain what a circumfix does the resulting gloss can only attach itself to either the (apparent) prefix or the suffix, but not both. For example, see the etymology for უსარგებლო:
There are a number of ways to fix all this, but perhaps the simplest may be to retire {{circumfix}} in favour of {{affix}}. This has been done, for example, in the etymology for უსინდისო:
From სინდისი(sindisi, “conscience”) + უ- -ო(u- -o, privative circumfix).
If anyone has another suggestion, do please share.
Agreed - the implementation of {{circumfix}} seems to have been poorly thought-through, and given it's redundant to {{affix}} (which came later), we may as well just retire it outright. No point spending time fixing it when it's no longer necessary. Theknightwho (talk) 14:05, 12 July 2023 (UTC)[reply]
I agree, {{circumfix}} has always been awkward. I would actually like to eliminate {{prefix}}, {{suffix}}, {{confix}} and {{compound}} in favor of {{affix}}; it's annoying working with existing entries that use the former templates. I have written a script to do it and have run it for certain languages, but not globally. Benwing2 (talk) 01:15, 13 July 2023 (UTC)[reply]
BTW @Nicodene, Theknightwho: I am in the process of rewriting Module:compound (probably to be renamed Module:affix) to support language-specific mappings between display affixes (e.g. '-käs') and corresponding link affixes (e.g. '-kas'), and in the process I can implement any necessary changes needed to completely support circumfixes in {{affix}} (and also write a script to convert the c. 3,500 current uses of {{circumfix}} to {{affix}}). {{affix}} already recognizes circumfixes by the presence of a hyphen-space-hyphen sequence in them but this part of the code hasn't been extensively tested and maybe could use some tweaks. If one or both of you could compile some test cases (i.e. just invocations of {{affix}} with circumfixes in them, but with as many different edge cases as you can think of), I'd be grateful. Benwing2 (talk) 05:56, 13 July 2023 (UTC)[reply]
Are there any 'circumfixes' or any uses of {{affix}} that apply 3 or more parts (instead of just 1 or 2)? That's one possible edge case that comes to mind (and also just an interesting question!). - -sche(discuss)07:28, 13 July 2023 (UTC)[reply]
@-sche{{circumfix}} doesn't support 3-way or N-way circumfixes; only circumfixes composed of a prefix and a suffix. {{affix}} doesn't either currently, although it could be taught to do so if there's a standard way of recognizing such N-circumfixes (do you have any examples?). Benwing2 (talk) 07:38, 13 July 2023 (UTC)[reply]
Some ablaut processes come to mind, e.g. Old English past participles of strong verbs with ge- WEAK -en, and Pali aorists from the root by a- LENGTHEN -i. The latter could be seen as a- -a- -i. --RichardW57m (talk) 12:07, 13 July 2023 (UTC)[reply]
@Nicodene Do you mean I should run my script and you will clean up any issues? I can do that, definitely; let's wait a couple of days to see if there are any more comments, and then I'll go ahead. Benwing2 (talk) 04:20, 16 July 2023 (UTC)[reply]
@Nicodene I added the ability to convert {{circumfix}} to the script (which previously handled {{prefix}}, {{suffix}}, {{confix}} and {{compound}}) and ran it on all 2,587 pages with {{circumfix}} on them (without saving anything). Warnings output here: User:Benwing2/convert-circumfix-warnings There are 142 warnings; most of them come from using ont- -en in Dutch and adding a gloss for the ont- part. The script (understandably) doesn't know what to do in this case so it outputs a warning and does nothing. In this case I'm not completely sure the circumfix analysis is correct but assuming it is, I guess for this particular situation the gloss should be moved to the circumfix as a whole without change. (In some cases this isn't possible because e.g. both parts are glossed separately or the gloss really is intended only for one of the parts, meaning it can't be done automatically.) There were only around 5 or 6 warnings for Georgian. Benwing2 (talk) 03:37, 17 July 2023 (UTC)[reply]
@Nicodene I deprecated it, which makes it display with a "deprecated template usage" message. If you find people still using it, we can add an abuse filter to block this. Benwing2 (talk) 19:17, 26 July 2023 (UTC)[reply]
As I parse it, the word "affix" implies that one or all or the constituent parts have a POS of "affix" -- which, in Japanese entries at least, is very much not the case: many terms are basically noun stacks, where the individual chunks are defined with a "noun" POS, or the so-called "infinitive" or "continuative" verb stems (which by some analyses are also nouns) plus nouns. Examples include 交通信号(kōtsū shingō, “traffic signal”, literally “traffic [NOUN] + signal [NOUN]”), or 食べ物(tabemono, “food”, literally “eat [VERB STEM] + thing [NOUN]”). So I have been using {{compound}} pretty extensively in my work on Japanese etymologies.
Ultimately, I don't have terribly strong opinions one way or the other about how we ultimately name this template -- mostly I'm just curious why you'd prefer the more-specific term "affix" over the more-general term "compound" as the name for a very-general-use template. ‑‑ Eiríkr Útlendi │Tala við mig18:14, 17 July 2023 (UTC)[reply]
@Eirikr I didn't actually design these templates or choose the names. What happened was that originally we had only the more specific *fix templates ({{prefix}}, {{suffix}}, {{confix}}, {{infix}}, {{interfix}}) and {{compound}}. User:Rua created {{affix}} as a generalization of the *fix templates, and then (possibly on someone else's suggestion) extended {{affix}} to handle compounds as well. So the name is a bit historical and maybe should be renamed, but I don't know what a better name would be. {{compound}} is a bit strange in that it only recognizes interfixes as such and otherwise treats all terms beginning or ending with a hyphen as compound components. (If you want that behavior in {{affix}}, precede the term with a circumflex, ^.) Benwing2 (talk) 18:25, 17 July 2023 (UTC)[reply]
Thank you for the background, that's useful information.
Is it appropriate to record pronunciations for translingual items? If so, how would one record the language? Can a labelling such as {{a|Burmese}} be appropriate? In the case I have added this label, Translingual ဣ(i, letter), {{a|Burma}} would be inappropriate because the contrast is with the pronunciation in different languages of Burma, with overspill to present-day Thailand. @Kwamikagami has already expressed the view that pronunciations are inappropriate for translingual items. --RichardW57m (talk) 09:54, 13 July 2023 (UTC)[reply]
This is like claiming under j#Translingual that J is the tenth letter of the English alphabet and is pronounced /ʤ/, and then deleting the English entry as redundant. English entries belong under English, and Burmese entries belong under Burmese. kwami (talk) 18:20, 13 July 2023 (UTC)[reply]
RichardW57m claims that because of the military in Burma ("As to what is special about Burmese, it's the Burmese Army"), the Burmese language is some sort of translingual entity and that, by implication, Mon and Shan are only secondary to Burmese. Why not say that, because Mon takes historical precedence, Mon is the translingual entity and Burmese is secondary? There's nothing special about Burmese that it should be treated differently than every other national/official language of the world. kwami (talk) 18:20, 13 July 2023 (UTC)[reply]
That is a lie.
Because of the Burmese army, the Burmese script is known as the 'Burmese' script, not the Mon script. Moreover, it is usually more useful to know Burmese rather than Mon or Shan. And indeed, because of Burmese empire, the Shan writing system has diverged so far from the Tai Le script and somewhat converged with the Burmese style to the extent that it was not a total travesty to include the Shan alphabet within the Burmese script. The Burmese alphabet therefore makes sense as a reference point for the Burmese script as a whole. --RichardW57 (talk) 23:40, 13 July 2023 (UTC)[reply]
RichardW57m has repeatedly deleted entries for Burmese, even though they are specific to Burmese, with the letter order of the Burmese alphabet and the pronunciation of the letters in Burmese. I've tried correcting the translingual section to say that it is a letter of the Mon-Burmese script, but he's reverted me there too. (Because of the links to WP, it's clear that he's presenting these as specifically letters of the alphabet of the Burmese language, and not of the translingual script. It's also clear from his comment that the Burmese alphabet is "the best documented, and possibly stablest, alphabet using the Burmese script" that this isn't a matter of confusion over the scope of term "Burmese alphabet".) The nationalism and ethnic assimilation policies of the Burmese military have no place on Wiktionary. kwami (talk) 18:20, 13 July 2023 (UTC)[reply]
Clearly, the Burmese letter should have its own, Burmese, entry. Translingual can be used to say that it is a letter in a certain script, but should not have language-specific information. Andrew Sheedy (talk) 18:45, 13 July 2023 (UTC)[reply]
So most languages using the Roman script should have their own entries for most of the 26 letters of the Roman alphabet? That'll do wonders for Lua memory errors. Now, it might make sense for exceptional features, but for every letter of every language? --RichardW57 (talk) 23:28, 13 July 2023 (UTC)[reply]
Regardless of what you think of the outcome, there was no consensus to do anything otherwise in the vote we had on this. You participated in the vote, so this shouldn't come as a shock, though it may have slipped your mind. Andrew Sheedy (talk) 23:38, 13 July 2023 (UTC)[reply]
Also, as was mentioned on the talk page of that vote, merging the entries wouldn't actually help much with the Lua errors. A different solution is needed. Andrew Sheedy (talk) 23:41, 13 July 2023 (UTC)[reply]
No-one said that every letter needs to have an entry for every language that uses it. That's why we have 'translingual' sections in the first place. But when a script (or letter) is only used by a handful of languages, it's not an issue to provide that information to the reader. kwami (talk) 23:43, 13 July 2023 (UTC)[reply]
Wrong question. The question is how many have durably archived materials written in them. And that is a tricky question, without fieldwork. As I understand it, published books count as durably archived by default. --RichardW57 (talk) 00:41, 14 July 2023 (UTC)[reply]
@Kwamikagami: You should only be asking the second question. And the answer is all of them, since we aim to include every language, regardless of whether they are standardised or not. Thadh (talk) 15:16, 14 July 2023 (UTC)[reply]
I looked up the first five 'established' languages listed in Ethnologue (skipping Bengali). Only one (Akha) was noted as having a Mon-script orthography. If representative, and if we count Burmese "dialects" as Burmese, that would suggest a score of languages use Mon script. But good luck attesting to many of them.
Anyway, if there are too many to list them all (assuming anyone ever wanted to go to all that effort), that's what the translingual section is for. But that's still no excuse to promote one language at the expense of others. kwami (talk) 00:27, 14 July 2023 (UTC)[reply]
If we're going to define letters by their position in an alphabet - and that mostly works only for base consonants - we need something with a stable alphabet. With the exception of NYA v. NNYA, Burmese works well for that up to the letter WA - so long as we stick to base (Unicode sense) consonants. You're fairly free to improve the definitions. For the Tai Tham consonants, I defined them with respect to the traditional Pali consonants, which is a culturally-relevant and stable reference point. I do not believe the positions of the 'extra' consonants added for Tai languages are stable. After all, the Tai languages seem not to agree on the repertoire. --RichardW57 (talk) 00:57, 14 July 2023 (UTC)[reply]
So please don't delete translingual sections, at least other than via due process, which is via {{rfd}} in these cases. (Unicode mistakes may fall to {{rfv}}.) Where I need a Pali letter entry, I was planning to cross-link to the multilingual letter entries where they existed. --RichardW57 (talk) 01:49, 14 July 2023 (UTC)[reply]
I have no problem with translingual sections. But I will delete false claims, whatever they're masquerading as. This isn't a matter of deleting something just because I can't confirm it, which is what rfd is for, but because it's intentional falsehood. I even tried creating a translingual section, only for you to delete it. kwami (talk) 02:51, 14 July 2023 (UTC)[reply]
@Kwamikagami: I've searched for such a change, and cannot find it. Are you possibly referring to this provision of content where I changed the L2 from 'Undetermined' to 'Translingual'? Should I have created a new section instead? There may be a valid objection that I changed the PoS from letter to ligature. People, please advise. --RichardW57m (talk) 17:19, 14 July 2023 (UTC)[reply]
Should we ultimately have a vote so WT:EL can prohibit pronunciation sections for translingual items? We may need to refine it to just exclude audio clips - it's possible that in this case a pronunciation section for ဉ like '/ɲ/ as the onset of a syllable' would be both appropriate and acceptable. (Whether that letter is translingual is another issue. It is used in Mon Pali.) A statement like 'See individual languages for pronunciation' would invite a lemma entry on the letter for each language that uses it, which is currently in the realm of the permitted rather than the preferred. (A majority, but not a supermajority, favoured removing individual languages letter lemmas in the failed vote references above.)
If we put a prohibition in, then an HTML comment could warn users not to add sound clips in the translingual section and reference WT:EL. RichardW57 (talk) 07:01, 15 July 2023 (UTC)[reply]
There's no reason we can't have a pronunciation, it just needs to be translingual. For example the IPA and other phonetic transcriptions. As you've been told, 'translingual' is beyond any particular language. Burmese is not 'translingual', nor is English.
As for ဉ being pronounced /ɲ/, there are two issues:
First, the transcription claims this is a phoneme. Is it actually phonemic in all languages that it appears in? Or in some cases might it be a phonemic /nj/ because the language has no /ɲ/?
Second, if we solve that by transcribing it [ɲ], is it actually phonetically [ɲ] in all languages that use it? In Burmese, for example, ရ is /j/, not /r/. It would be inaccurate to claim ရ is a translingual letter pronounced [r] (or /r/, for that matter). What we could do is say that ရ is a letter transliterated⟨r⟩. Under the Burmese entry it would be clarified as being /j/ (/r/ in some dialects), under the Mon entry it would would be transcribed as /r/ in the breathy register, etc.
Let me see if I get you right. We can't have regional pronunciations for English words, because they're regional, not universal English? That's the analogy I see. I started this topic to ask about how to label them, to which I got the answer best translated as 'Bugger the user who wants such information'.
I was talking about the general principle. We would have to mention exceptions. I'm not aware of any exceptions for ဉ, but we would have to mention their existence (better: list them) if there were. Describing the coda will get horribly complex. I do wonder if someone misinterpreted the pronunciation need as one for the syllableဉ - there's a whole lot of reworking needed on Burmese entries for the consonants if one chooses to have them; Mon is easier.
Transcribing it as [ɲ] is inviting a lot of trouble. We want a broad brush description with caveats for translingual statements. Phonetic transcription belongs to languages, or perhaps even idiolects.
Yes, Translingual ရ, even trans-Burmese ရ, gets complicated. I wouldn't bet on Mon ရ always inducing breathy voice - loanwords might have broken that association, just as tonal implications break down in Tai languages, whence the extra tone marks added to Thai to deal with having a lot of Chinese immigrants. --RichardW57 (talk) 08:22, 15 July 2023 (UTC)[reply]
This isn't difficult: translingual means translingual. You seem to be purposefully misunderstanding in order to make your argument.
If under 'translingual' you wanted to give the pronunciation of the various languages that use the letter, that would be technically correct, but that's why we have sections for individual languages.
Take a look at j#Translingual and j#English. Can you tell me what the translingual pronunciation of the letter J is? Not IPA or NAPA or any other phonetic system, but of the letter in general? I sure can't.
(The pronunciation given in there belongs to the IPA definition, which is why it's dab'd as IPA; personally I would be more comfortable if it were a subheader under the IPA entry, as IMO that would be less likely to be misleading.) kwami (talk) 08:39, 15 July 2023 (UTC)[reply]
Thank you for acknowledging that giving the pronunciation for various languages is correct. Now can we please agree on the labelling. Are {{a|Burmese}} and {{a|Burma}} technically correct, or should I be using something else for labelling pronunciations by language? --RichardW57 (talk) 12:14, 15 July 2023 (UTC)[reply]
@RichardW57: A vote might be a good idea. I'm inclined to think that pronunciation is language specific, except for things like IPA, and only the symbols themselves can be considered Translingual. I seem to recall there being a discussion or two about prescriptive pronunciation sections for taxonomic names. I think the consensus was that there was no point including them, but I'm not sure. I would be in favour of prescriptive pronunciations for Translingual that are determined by some central authority, but since Translingual isn't an actual language, I personally prefer relegating language-specific pronunciations to entries for that language. As you point out with Spanish, if a letter can have a plurality of different pronunciations in a single language, then including pronunciations for every single language within the Translingual entry would clutter the entry too much (though I certainly sympathize with the argument that having all the languages separate just creates a different sort of clutter). Andrew Sheedy (talk) 18:18, 15 July 2023 (UTC)[reply]
The article for Spanish z brings attention to another point. Looking at the Burmese script examples, the pronunciation gives the pronunciation of a whole syllable, and that reveals far too many irrelevant details. Just giving the pronunciation associated with the symbol itself, as in the Spanish example, would give a much simpler description, so that one could say of the Burmese script near-analogue သ something like
As the sole element of the onset, for most languages it represents /s/, but Burmese has /θ/ and Shan has /sʰ/ contrasting with /s/ represented by ၸ(tsǎ).
(There's something weird going on with Shan initial sibilants or their descriptions that I haven't got to the bottom of.) One would not need to give phonetic samples for all the languages. One could use the quotations to reference a word from every language, and let the very curious reader progress from there. --RichardW57 (talk) 19:47, 16 July 2023 (UTC)[reply]
It seems rare for letter sections to address what happens in the coda, but in Indochina I believe there may be a lot of variation in that across languages. Standard Burmese has collapsed to two terminal consonants, with the written consonant often determining the vowel. That is possibly too complicated for letter sections to handle. Thai goes into more details for most, if not all, consonants. --RichardW57 (talk) 19:47, 16 July 2023 (UTC)[reply]
I agree on both points: focus on the consonantal sound and overtly mention the realization as a coda -- even for cases where they're identical, because the reader can't know they're identical or even if it occurs as a coda if we don't mention that. kwami (talk) 23:33, 16 July 2023 (UTC)[reply]
I've been reverting your out of process replacements of translingual entries by Burmese entries, for which you should have been blocked. Where I noticed an improvement as far as the Burmese information was concerned, I made the effort to leave separate translingual and Burmese entries rather than do the simpler revert. --RichardW57 (talk) 23:51, 13 July 2023 (UTC)[reply]
So far as I am aware, 'Mon-Burmese' script is not the standard term here. You should ask for formal renaming, or else striking out the 'Mon-' would be a reasonable measure to take. I haven't been doing that explicitly; if I've done it, it would be collateral damage due to your out-of-process deletions of translingual entries. --RichardW57 (talk) 23:51, 13 July 2023 (UTC)[reply]
I should note for the readers here that you have not been making an honest attempt to consolidate the language entries into a single 'translingual' section, which actually I wouldn't care all that much about. Rather, you've been specifically promoting Burmese to 'translingual' status. kwami (talk) 00:40, 14 July 2023 (UTC)[reply]
Dear imitator of Johnson, I have been reverting your replacements of translingual entries by Burmese entries. Additions are not a problem I care enough about. If I were aiming to replace originally Burmese entries by translingual entries, I would also have been consolidating the Mon entries. I am not the one who added Burmese-dominated entries to translingual entries. Indeed, the definition of the translingual entries by reference to the Burmese alphabet goes back, at least in the cases I've looked into, to the pages' original creation by @Visvisa. --RichardW57 (talk) 01:41, 14 July 2023 (UTC)[reply]
I'm having trouble understanding Wiktionary:Votes/2011-06/Redirecting_combining_characters. Why don't ႅ(“U+1085 MYANMAR VOWEL SIGN SHAN E ABOVE”) and ᩖ(“U+1A56 TAI THAM CONSONANT SIGN MEDIAL LA”) have hard redirects to ႄ(“U+1084 MYANMAR VOWEL SIGN SHAN E”) and ᩃ(“U+1A43 TAI THAM LETTER LA”) respectively? Is this merely a failure of implementation? A list of redirects was created by @Daniel Carrero, so he might know the answers. --RichardW57m (talk) 17:08, 13 July 2023 (UTC)[reply]
Hello. Maybe you would like to create the redirects for some entries like those you mentioned? Yes, we can redirect the combining character to the non-combining version when possible. Yes too, it seems like we have not yet implemented this for all possible cases so far. --Daniel Carrero (talk) 11:37, 15 July 2023 (UTC)[reply]
@Daniel Carrero: I was thinking of doing them, but I then wondered if there was some reason for not doing them. There are some awkward cases around, as the Shan and Mon uses of Translingual ဵ U+1035 MYANMAR VOWEL SIGN E ABOVE make me wonder if they are mere homonyms; it was encoded for Mon without any understanding of its function. --RichardW57 (talk) 12:37, 15 July 2023 (UTC)[reply]
Personally I don't speak those languages. As far as I can see, it looks like Unicode originally intended to have the "same" character encoded in two codepoints, but people actually use the separate codepoints for different purposes.
If that is correct, then I would suggest that real-life usage has priority, not what Unicode thinks. Perhaps we can keep them as separate pages on Wiktionary, unless I'm missing something so far. The "Usage notes" sections can also help by explaining how the characters are used, possibly including notes about the Unicode implementation too. --Daniel Carrero (talk) 13:09, 15 July 2023 (UTC)[reply]
@Daniel Carrero: My plan is to split them by 'etymology', and then have a 'soft redirect' for the Shan usage, rather than a hard redirect as mandated. One can't have a hard redirect for half a page! --RichardW57 (talk) 16:46, 15 July 2023 (UTC)[reply]
(This is part one of a number of proposed changes)
Rationale
When a word is borrowed into Chinese, or vice versa, the pronunciation carries a very important role in influencing the spelling and/or orthography (except in calques and orthographic borrowings). The entries currently reflect very little of this, or when it does, it is often quite messy, the extreme cases being 茶 and 咖啡. This problem is exemplified by how Japanese and Korean borrow words from various periods/lects of Chinese such that sometimes there are so many readings to organize them properly. {{CJKV}} is also often despised of due to it awkwardly taking precedence over other descendants.
Additionally, Unified Chinese has long been a contentious issue. I hope this proposal would at least provide a practical solution by unmerging Chinese in places (i.e. etymologies) without compromising the duplication of contents, given that we already sometimes do clearly distinguish the Sinitic languages in etymologies.
I propose that for etymology purposes, we treat the Sinitic lects as distinct languages with their respective pronunciations, with [zh] (Chinese) as the version of Sinitic languages unified under their written form of Chinese characters.
Actual suggestions
Some of the proposals have multiple options listed here; the proposals may be implemented separately if possible. Feel free to put forward other suggestions as well.
[zh] should not be used for all etymology purposes except for calques (this includes partial calques) and orthographic borrowings (which basically concerns Japanese only), whether or not the direction of the derivation is into, out of Sinitic.
The correct etymology code should be used instead of [zh] in such cases.
For etymologies that are less clear (e.g. those from Sinitic into Zhuang and other MSEA languages), [zhx] Sinitic may be used.
Note that if calques are used in only a lect (or some lects) with vocabulary grammar/sentence structure unique to that lect, [zh] should not be used as well, e.g. 唔係我杯茶／唔系我杯茶 should use [yue] whereas 一日一蘋果，醫生遠離我／一日一苹果，医生远离我(yīrì yī píngguǒ, yīshēng yuǎnlí wǒ) should continue to use [zh].
For internal etymological relationships (some of these are already done in practice - the proposal aims to rectify such practice):
Internal borrowings based on sound, e.g. Cantonese 冷(“Teochew person”) from Teochew 人(“person”), should use {{bor}}.
Internal borrowings based on phono-semantic matching, e.g. Mandarin 買單／买单(mǎidān) from Cantonese 埋單／埋单(maai4 daan1), should use {{psm}}.
Inherited words that are written with the same characters (or orthographic variants) need not be mentioned; words that are written differently will be marked with {{inh}}, e.g. Cantonese 夾／夹 from MC 合.
{{dercat}} (or something similar) may be used where appropriate.
For words originating from a Chinese lect and subsequently used in the rest of Chinese using the same characters (whether or not it has a further external origin), should be considered as orthographical borrowings from that lect into the concerned other lects. This does not need to be explicitly use {{obor}}, but should be categorised appropriately, and the original lect should be mentioned in the etymology.
When it is unclear which lect was a loanword first borrowed into (as often is the case for e.g. loanwords in Taiwanese Hokkien and Hakka from Japanese), one may treat such loanwords parallelly borrowed into each Chinese lect.
When a loanword has multiple origins (e.g. 巴士(bāshì), 咖啡(kāfēi)), they should be listed under multiple etymology sections accordingly.
The descendants will list all descendants by explicitly explaining which lect (including OC/MC where appropriate) a word is derived from, either in (a) tree format (similar to reconstruction entries), (b) flat list where the first level bullet are the Sinitic lects and their respective descendants listed under them, or (c) subheaders similar to (b) but each lect is listed as a header. The former will likely lead to many levels of ident while the latter two requires; all of them will likely cause problems with {{desctree}}.
{{CJKV}} will be deprecated - they will be listed along with the rest of the descendants, with a |qq=Sino-Xenic if appropriate.
The Proto Sino-Tibetan reconstruction entries will only list Sinitic lects up to either (a) Old Chinese, or (b) Old Chinese, Middle Chinese, and Proto-Min. The information will be hosted in the descendants section of the relevant character instead, as mentioned above.
Since [zh] now only serves to represent the language written with Chinese characters, non-Chinese character entries (including Latin script entries, Hokkien POJ, Bopomofo, emojis, etc.) will be split. This facilitates easier treatment and organization of such entries, e.g. Q and K which reach into absurd amounts of etymology sections. There will unavoidably be some duplication, but I believe such impact would be minimal compared to any other more drastic approach.
The existing mess evidently reflects a gap between reality & our model of it. I’ll be as concise as I can here, esp. b/c the “thread” is young and I hope to steal none of your thunder. If there are gaps in my presentation, or holes in my reasoning, I hope you & others will let me know.
It seems evident (but maybe not obvious) that we need to acknowledge the existence (actual or constructive, as each case may be) of hybrid interlanguages that pair the phonology of each respective “dialect” with the non-phonological elements (incl. the orthographies!) of Standard Chinese.
Socially, these interlanguages — euphemistically called 書面語 in the Hong Kong Cantonese context — appear to be high (& “modern”) registers of the adjoining vernaculars. Linguistically, they are simply situational dialects, or versions, of Std Chinese. This is — rather — the only clean, efficient way to conceptualise them.
Except phonologically, the various interlanguages are identical to Std Chinese & to each other. In the context of projects like Wikt (or words.hk, where there was a period of deep discussion), busywork is generated ad nauseum if these interlanguages are viewed as mutually separate registers of the vernaculars. Hence on Wikt we’ve attempted to duck the busywork by formally denying the separateness of the vernaculars, which defies reality.
The interlanguages are not “versions of Sinitic languages unified in written form with Chinese characters”. Rather, they’re outgrowths of Std Chinese, itself a modern, political phenomenon (like many or most national languages). They only began to form when China started modernising, in the 1850s or later. The interlanguages are local or geo-ethnic modules for interfacing with the modern phenomenon that is Std Chinese. They're part of Std Chinese, whether we want to see them as semi-stable dialects, or as sets of (phonological) rules that generate momentary phenomena at “runtime”.
What’s [zh]? Logically, [zh] is the modern, nationalistic Std Chinese. A version of the traditional, cosmopolitan Book Koine is built into it, just as another version of that Book Koine is built into Std Japanese. For practical reasons, and in line with global practice, it seems safe to say that [zh] also includes the modern & near-modern stages of the dialect(s) of Mandarin that Std Chinese is based on. And, very importantly, [zh] includes Mando-Cantonese & all the other nationalistic interlanguages.
This seems to be the only way to pack [zh]. We get in trouble (with logic) if we try to pack the other vernaculars into [zh]. We also get in trouble (busywork) if we try to unpack the nationalistic interlanguages from [zh] and throw them in with the vernaculars.
The one vernacular component of the interlanguages is the phonological matrix — largely closed — that specifies how each Sinograph is to be voiced. All other (incl. orthographical) elements come from Std Chinese, which is dynamic & boundless. In fact, the interlanguages are largely insulated from lexical (“slang”, etc.) & grammatical evolution in the adjoining vernaculars — although unrequited convergence towards Std Chinese may currently be the greater part of evolution in the vernaculars anyway. This, again, illustrates that the interlanguages are special modules attached to [zh], while the vernaculars are not.
So we’ve addressed “duplication of contents”, but — on the flip-side — there’s also systematic omission of info on the vernaculars under the pretense that all non-phonological aspects of Std Chinese automatically apply to the vernaculars. (Of course, all non-phonological aspects of Std Chinese do automatically apply to the interlanguages, as a matter of socio-political reality.)
What does 桌 mean in vernacular Cantonese? Say I’m a learner and I go to
The layout implies that Cantonese is a sub-part of “Chinese”, sharing in the two meanings of 桌 laid out in the Chinese section. Scrolling down, I find that 桌子 has a Cantonese “dialect synonym” 檯. So I guess I could say 買咗張檯. But what about 買咗張桌? The dialect synonym table, in its entirety, implicitly suggests that 買咗張桌 would be ill-formed, but the entry is evasive on this.
What about 杯子在桌上 and 一桌菜? The Cantonese readings for these examples are just a few clicks away, but nothing facilitates the basic & frankly all-helpful realisation that 杯子在桌上 is not well-formed conversational Cantonese.
The entry also can’t seem to decide if it’s dealing with words or with a graphic variant. A pair of readings for 桌 are given for Hokkien. The Std Chinese example sentences are even less relevant for Hokkien than for Cantonese. In “dialect synonyms” I find 桌仔. Again, I’m left to wonder if both 桌仔 & 桌 can be used to mean TABLE. Meanwhile, in the native Sinographic script, the Hokkien word TOH-Á is written 棹仔 or, less commonly, 卓仔. Wikt omits these facts (arguably unintentionally) while non-native 桌仔 is (arguably intentionally) presented as the representative form, for reasons having more to do with [zh] than with Hokkien.
Again, 桌仔 is essentially an interlingual form, yet it is presented as being fully vernacular, & fully representative of the vernacular — for reasons that most users of the site could not be expected to even begin to understand.
We hope the user will somehow do their guesswork correctly. But doesn’t that defeat the purpose of a dictionary?
We implicitly acknowledge the existence of the interlanguages, and correctly (although maybe for the wrong reasons) conclude that they’re part of [zh]. Now we need to explicitly acknowledge their existence while de-conflating the vernaculars.
What would this look like on a word like 取消, or 便當?
Much (not all, not by a long shot!) of the content for [zh] would be fine as-is, for Std Chinese…. For 便當, for instance, the Pronunciation box under Chinese is perfect for showing the readings in various interlanguages.
The vernacular Formosan Hakka word 便當 (for LUNCHBOX) — a Japanese or Hoklo loan (or both) — is another matter. The entry for vernacular Hakka should be parallel to C, J, K, & V, entirely separate from [zh].
This does not conflict with the reality that, presented with a Std Chinese text, many elderly Formosan Hakka speakers would be able to voice it using Mando-Hakka interlanguage. 便當 (or 取消) as it occurs in that context is not the vernacular Japanese or Hoklo loanword, but rather a Std Chinese cognate (regardless of its origins) being smoked through a filter of Hakka phonology. (This point would be clearer & perhaps less counter-intuitive in the case of Std Chinese words that don’t have cognates in vernacular Hakka. Hakka speakers tend to feel that every element occurring in Std Chinese — but not, say, Japanese or Cantonese — must have a reflex in Mando-Hakka interlanguage.)
This is a straightforward “reading” of the linguistic & socio-political reality. A hundred details follow from it. Rather than try to flesh it out more ( = sprawl), let me simply invite you & others to discuss as you see fit. 釆 (talk) 02:47, 26 July 2023 (UTC)[reply]
@Wpi To briefly address a related axis of inquiry, why do we treat Vietnamese TRÀ as a “descendant” while Teochew TÊ is treated as a variant of the “ancestor”? What’s the indication that the two words didn’t just evolve in parallel?
The distinction between “Chinese” & “non-Chinese” languages is artificial, isn’t it? It doesn’t enlighten, but it does create busywork. It doesn’t correspond with rigor to any aspect of diachronic reality, and the unquestionable synchronic modern reality of “Chinese” is accounted for with [zh], as addressed above.
> For words originating from a Chinese lect and subsequently used in the rest of Chinese using the same characters (whether or not it has a further external origin), should be considered as orthographical borrowings from that lect into the concerned other lects.
This may have merit on some level. In practice, though…. Take Hokkien CHOÁN, meaning “to earn” (often through improper means). Or CHAU-THAT, meaning “to mistreat”. These were pre-modern loans from Mandarin. The words are written 轉 & 糟撻 in the native Sinographs, but Wikt prescriptively has them written as 賺 & 糟蹋. The reason for this traces back to Neo-Sinology & (thus) [zh] — namely, it’s that that’s how the Mandarin cognates are written in (orthodox) Std Chinese. So — although mileage may vary for Cantonese (given its unique socio-political 條件) — the etymological inquiry becomes circular if based on a pseudo-orthographic inquiry carried out Neo-Sinologically. As much sense as your proposed shortcut may make, there doesn’t seem to be enough water (accessible research) on the ground yet — for ecological reasons — to float it for any (socio-politically) Chinese vernacular besides Cantonese & maybe 吳越語 (& Mandarin).
Alternatively, whether CHAU-THAT (or CHOÁN, etc.) is an orthographical borrowing is a question of fact. It might not be so bad for such micro-questions to be examined in turn with no presumption either way — and with no prejudice (& no pretenses, or make-believe answers) if mankind just happened to be too busy to look into the matter. 釆 (talk) 03:20, 26 July 2023 (UTC)[reply]
Strong support. This will make life so much easier for showing descendants into the likes of Korean & Japanese. Descendants at entries like 北京(Běijīng) or 點心／点心(diǎnxīn) are a mess. We should also focus on having proper usage examples for different lects and make sure that categorization works properly. Etymological information like for Cantonese at 圜(won1) should not be hidden away in the expanded pronunciation section. I had to search around to find out where the categorization was coming from. Similarly, entries like 㗎(jià) should not have Mandarin categorization if they are not used in it. (Also hopefully this will make some push towards separating out historical lects as well.) AG202 (talk) 04:20, 26 July 2023 (UTC)[reply]
"Wiktionary:Word of the day/Nominations" advises: "Abbreviations, initialisms, acronyms, symbols, and the like are unlikely to be selected without extremely convincing reasons." We have, in the past, occasionally featured abbreviations as WOTDs—a recent example is TEFL for World Teachers' Day. @Algrif nominated e.g., i.e., or viz. back in 2020, and @Ioaxxere recently nominated a number of prefixes and suffixes for featuring as WOTD. What does everyone think about that? I guess I'm OK either way. — Sgconlaw (talk) 07:31, 14 July 2023 (UTC)[reply]
I seem to remember I also suggested tl;dr previously. If a person hasn't come across this before, it definitely is not obvious. I think some abbreviations etc. would be quite acceptable -- now and then. IMHO -- ALGRIF talk09:23, 21 July 2023 (UTC)[reply]
Pronunciation indicated as "Deliberately not telling"
I commented out the “Pronunciation” section at ဉ, added by RichardW57m, the content of which was “Deliberately not telling.” My edit summary was “RichardW57m: I do not think a dictionary should indicate a pronunciation as “Deliberately not telling.” Did you think this was suitable?” This is the discussion from my talk page:
Deliberately Omitted Spelling in ဉ
Well, as the translingual Burmese script letters used to contain pronunciations, I thought users and other editors should know that no pronunciation is given by policy, rather than being missing simply because we haven't yet got round to it. --RichardW57m (talk) 14:04, 14 July 2023 (UTC)
@RichardW57m: “Deliberately not telling” seemed that we are deliberately hiding this information from readers. J3133 (talk) 14:06, 14 July 2023 (UTC)
I believe that “Deliberately not telling,” without any explanation, is not appropriate in a dictionary. What do others think? J3133 (talk) 14:35, 14 July 2023 (UTC)[reply]
Their adding "Deliberately not telling" seemed quite clearly directed towards another user, as opposed to useful dictionary information, and was thus inappropriate to list there. I removed it. Megathonic (talk) 17:28, 14 July 2023 (UTC)[reply]
@Sgconlaw: I've now suggested changing WT:EL. Until then, all we've got is the discussion in the topic above. In principle, I'd have preferred listing more languages, though mostly but not always the Burmese onset should be representative enough where I undid the out of process deletion of the translingual section. As a first step, I was going to label the pronunciations by language, but realised there was a formal issue with the labelling. For example, some well-meaning editor could have come along later and destructively converted {{a|Burmese}} to {{a|Burma}}. However, the explicitly expressed references are now running 2:0 against providing the information where the user would first look for it. --RichardW57 (talk) 07:27, 15 July 2023 (UTC)[reply]
@J3133 @Sgconlaw just to let you both know I changed the trans-lingual entry from "letter of the burmese alphabet" to "letter of the mon-burmese script" and changed the redirect to the scripts' wikipedia page. If it's a translingual entry, it didn't make sense to direct to the wikipedia page exclusively about Burmese, especially since the Burmese-specific wikipedia page is already linked in the Burmese entry. But since you guys are discussing that page i thought i'd let you both know. سَمِیر | sameer (talk) 20:47, 20 July 2023 (UTC)[reply]
It's called a temper tantrum, because people have told him that he's not supposed to promote Burmese above other languages, but rather to confine it to the Burmese header. kwami (talk) 04:38, 15 July 2023 (UTC)[reply]
I have changed a number of entries from [mul] 'translingual' to [und] 'undetermined' when the language is undetermined. User:Benwing2 has reverted me. (78 reverts and counting for this and deletion tags for empty and spurious articles.) I've done the same with compound characters (letters with diacritics) with are not confirmed to exist in any language, such as subscript kana with the voicing diacritic -- subscript kana are intended for Ainu, but they don't occur with the voicing diacritic in that language). Is this what the 'translingual' header is for on Wiktionary, just 'unidentified/unconfirmed'? Or is it intended to have a positive meaning, such as used in multiple languages? kwami (talk) 04:31, 15 July 2023 (UTC)[reply]
Yes, I know translingual is 'mul'. The question is what to do with items where we don't know the language, because we have no attestation it exists, except for mentions which do not identify the language. This is the case i.a. for letters that are not attested in any language. It's possible they are used in multiple languages, they may be custom letters for a single language and not used for anything else, or -- in a few cases -- they are Unicode inventions that are not used in any language. For some reason, people are edit-warring to keep these. I figured [und] 'undetermined' would be appropriate when the language is undetermined. Of course, if someone is able to ID the language, it should be changed to that language's header. kwami (talk) 04:44, 15 July 2023 (UTC)[reply]
Well, I'm just answering the question you asked. I looked thru a few of your contribs and can't find an example of what you're talking about. Some diffs or links would be helpful. —Justin (koavf)❤T☮C☺M☯04:49, 15 July 2023 (UTC)[reply]
Take for example ㇲ゙. The letter ㇲ is used in Ainu, and the diacritic indicates voicing. But these letters are never voiced in Ainu. So what is ㇲ゙ for? Is there any language that uses it, or did someone just make it up? And how can we justify calling it "translingual"? Shouldn't it at the very least be labeled "undetermined" until (and if) someone is able to identify a language, at which point the header would be changed to that language? kwami (talk) 04:50, 15 July 2023 (UTC)[reply]
Sure, but there are a lot of characters that are not used by any particular language, such as 😋 or ☬. Those characters aren't linguistic data from an undetermined language: they are just characters that aren't for any particular language. Per ISO 639-3, und (undetermined) is intended for cases where the language in the data has not been identified, such as when it is mislabeled or never had been labeled. I don't know that this applies to various characters like these, so do you have a proposal for which kinds of characters you think are mul and which are und? —Justin (koavf)❤T☮C☺M☯07:22, 15 July 2023 (UTC)[reply]
Sure, emoji and many other symbols are [mul]. But, again, ㇲ゙. That's a letter, not a symbol, so it must be a letter of some orthography, but the language has not been identified.
IMO, in order to qualify as multilingual, we need to verify that the character is actually multilingual. Let's say I create an entry for the supposed letter ʞ̋, with the header "translingual" and capital form Ʞ̋. I "define" it as "a turned k with double acute accent", which provides no actual information or insight. It sits unnoticed on Wikt for a decade, then you come along and wonder what are all the languages that use it and make it "translingual", assuming it exists at all. So you tag it for verification (or deletion, since there's no content). Meanwhile, until the discussion is resolved, should it really stay under the 'translingual' header? Since the language is undetermined, shouldn't it be listed as 'undetermined' and given the code [und]? Then, if it's found to be Foonese, the code can be changed to [foo], or if no evidence is found for its existence it can be deleted, but meanwhile at least we're not falsely claiming it's found in the alphabets of multiple languages. kwami (talk) 08:11, 15 July 2023 (UTC)[reply]
If you would tag it for verification, everyone would be happy, but we don't use speedy deletion for these cases and there is no consensus to delete these entries through RFD. Thadh (talk) 08:42, 15 July 2023 (UTC)[reply]
I'm not talking about deletion here, I'm asking whether something should be listed as 'translingual' if the language is undetermined. kwami (talk) 08:55, 15 July 2023 (UTC)[reply]
Yes, this is a good discussion, I'm very interested to hear the opinions of other users and see what we collectively decide on this. I'm just saying that your point that "you tag it for [...] deletion since there is no content" is not good advise, and Benwing removing the deletion tags on all those articles is reflective of the fact that that's not how we handle these things. Thadh (talk) 10:39, 15 July 2023 (UTC)[reply]
Note that as it's a character in a block of a phonetic symbols plus an appropriate diacritic, it is inherently likely to be translingual, just like ƛ. There's no implication that is used in any orthography. --RichardW57 (talk) 09:00, 15 July 2023 (UTC)[reply]
I don't understand your logic. The letter ㇲ without a diacritic is, AFAIK, only used for Ainu. So what makes you think that the letter ㇲ with a diacritic is "inherently likely" to be translingual, especially since it does not occur in Ainu? The letter ƛ is different: it was created for a phonetic transcription system, not for a particular language, though it's since been adopted into several orthographies. kwami (talk) 09:05, 15 July 2023 (UTC)[reply]
Writing a foreign name in your script does not make that script translingual. If I write "Tchaikovsky", that doesn't make "tch" a letter of the Russian alphabet.
Do you have reason to believe that ㇼ is not used in Ainu? If you file a bogus rfv in bad faith just to score points, then yes, that would be disruptive. kwami (talk) 07:02, 16 July 2023 (UTC)[reply]
Just as a heads up, ㇲ゙ is apparently used in Miyako, e.g. ぴなㇲ゙, 捨てぃㇲ゙, 明かーㇲ゙, みㇲ゙. Though this is quite suspicious since I can't find stuff using it on {{R:JLect}}.
It could also be used to achieve stylistic effect in Japanese, as a small form of ズ), so there's certainly an element of translinguality to it. – Wpi (talk) 13:42, 15 July 2023 (UTC)[reply]
I would accept that as confirmation. The alt spelling uses html to make the kana small. If nothing else, it identifies a likely language so that the reader can investigate further.
As for being translingual, lots of characters can be used for stylistic effects, but I don't know if that should count. Just because someone used Armenian or Cyrillic look-alike letters in a text doesn't mean they're now English. kwami (talk) 20:38, 15 July 2023 (UTC)[reply]
@Koavf: Some of them have arisen because people assumed that graphic characters defined by Unicode should be recorded by Wiktionary. Additionally, those just filling in the gaps (e.g., I think, @Atitarev), had more productive things to do with their time than scour through Unicode proposals. And one won't find proposals for the earliest additions to Unicode, when they were hurriedly putting a useful system together. --RichardW57 (talk) 08:36, 15 July 2023 (UTC)[reply]
@Koavf There are situations where an ancient language in an inscription may not yet be identified - I believe we have a handful of entries like that (I've definitely seen one, but can't remember what it is). Theknightwho (talk) 15:44, 15 July 2023 (UTC)[reply]
An IP created Translingual entries for R12 and R134a, with translation hubs containing English translations "Freon 12" and "Freon 134a". This seems fishy but I don't know enough about refrigerants to say for sure. Do we allow Translingual translation hubs and if so are these entries (a) CFI-worthy, (b) properly Translingual? Benwing2 (talk) 07:41, 15 July 2023 (UTC)[reply]
If it has an "English translation" shouldn't the hub just be there? Both of these already have serviceable English entries: dichlorodifluoromethane (R12), norflurane (R134a). From a dictionary perspective I am a bit sceptical in general about the value of creating entries for stuff like refrigerant codes. —Al-Muqanna المقنع (talk) 09:14, 15 July 2023 (UTC)[reply]
User Theknightwho cancels edits for no reason and is also not an editor for the Nakh languages. Because of this user, the Chechen entry was deprived of declension. ɶLerman (talk) 17:04, 15 July 2023 (UTC)[reply]
@ɶLerman (aka Gnosandes) You changed a template in a way which broke most of its current uses, I told you this when I reverted you, and you have refused to acknowledge why this is a problem. Running to the Beer Parlour to lie about what happened just reflects badly on you, and you have a long history of this kind of behaviour. Theknightwho (talk) 17:07, 15 July 2023 (UTC)[reply]
@Theknightwho: I didn't break the template, but created a new one. And this is no longer a problem. I did not lie, but showed that you cancel the edits made completely unreasonably. This does not affect me in any way, I inform you. ɶLerman (talk) 17:14, 15 July 2023 (UTC)[reply]
@ɶLerman I got a notification that you undid my edit, the wikitext looked identical, and I didn't spot that you'd created a new template with an almost identical name. This is after you'd already reverted me and while you were repeatedly lying that I reverted you for no reason. It's no surprise that I missed that it had changed by 2 characters. If you had explained what you were doing and acknowledged that there was a problem with your previous approach then this wouldn't have happened. Theknightwho (talk) 17:31, 15 July 2023 (UTC)[reply]
@Theknightwho: It is impossible not to notice this. I think that deleting a template is a whole procedure, during which you can notice a different template name. In addition, you are the face of Wiktionary and are the administrator. I demand that you immediately stop accusing me of lying and don't violate human rights (my rights). This is provocative behavior, from which I asked user:Imetsia to protect me and from you. ɶLerman (talk) 17:49, 15 July 2023 (UTC)[reply]
Every single time you get into a dispute with an admin, you start making things up like your human rights are being violated or that you're being harassed. It's intolerable. Theknightwho (talk) 17:57, 15 July 2023 (UTC)[reply]
User:Theknightwho is an administrator, one among many. I am sorry that you find his behavior reflects badly on Wiktionary and that the discussion of the substance has devolved into interpersonal conflict. DCDuring (talk) 18:01, 15 July 2023 (UTC)[reply]
@DCDuring: Excuse me, please, I use a translator for English. Yes, I think in 5 years in Wiktionary I would be able to learn English, but I won't do it. I don't think he's doing anything bad in Wiktionary, he's making comments to me, but these comments turn out to be some kind of uncontrollable. ɶLerman (talk) 18:15, 15 July 2023 (UTC)[reply]
The template was edited and broken on many pages? So you broke the template, someone undid that and you're tattling on them undoing your breakage? Vininn126 (talk) 17:12, 15 July 2023 (UTC)[reply]
@Vininn126: At first, I may have done it wrong. This template is rarely used. But then I made a new template responsible for a certain paradigm. However, this user deleted the new declension template on the Chechen record, but did not add the old one. In addition, he deleted the new template. This is provocative behavior. ɶLerman (talk) 17:25, 15 July 2023 (UTC)[reply]
@Vininn126: Yes, I have nothing against this, the first 2 cancellations were correct, but the rest are not, he are canceling a new template and deleting a new template. ɶLerman (talk) 17:36, 15 July 2023 (UTC)[reply]
It is particularly appropriate that careful diction is one way to reduce the apparent rancor on discussion pages. One principle is to avoid using words that have imply something about the (to be assumed unobservable) intent of another person, especially a participant in a discussion). An example of such a word is lie ("To give false information intentionally with intent to deceive."), also deceive. That there is a sense of lie "To be mistaken or unintentionally spread false information." does not prevent the 'intentionally' interpretation. Another principle is to simply never discuss a person at all, but rather to discuss the subject matter. This is a bit simpler than AGF and has the virtue of excluding deniably sarcastic praise.
It is understandable that these basic practices are not universally followed by bookish folks like as as interpersonal behavior is not usually considered a subject of study, except in business schools. DCDuring (talk) 17:54, 15 July 2023 (UTC)[reply]
I'm all in favour of assuming good faith, but not when a user repeatedly demonstrates that they don't have it. If someone has something explained to them several times and then they claim that they were never given an explanation, then it is reasonable to conclude that there is intent to deceive. Theknightwho (talk) 17:59, 15 July 2023 (UTC)[reply]
Consider the possibility that are least one person in any given discussion might be mistaken. It might be even better to assume it. DCDuring (talk) 18:04, 15 July 2023 (UTC)[reply]
@DCDuring If someone gives me a bad explanation, I'll explain why I think it's a bad explanation. What I won't do is pretend they never gave me an explanation at all and then say that while complaining about them. In this particular case, Gnosandes very clearly does understand what I was saying, going by the above discussion. Theknightwho (talk) 18:07, 15 July 2023 (UTC)[reply]
Obviously, I was mistaken in believing it difficult to understand the intent of someone else in a discussion. I must be overgeneralizing from my own inability to gain such insight. DCDuring (talk) 18:13, 15 July 2023 (UTC)[reply]
@DCDuring That user has a history of being difficult otherwise and many, many people have had a less than cooperative experience with that user. It's not just the bad English, I assure you, as can many other editors. Vininn126 (talk) 18:27, 15 July 2023 (UTC)[reply]
Notwithstanding that, I don't think it ever helps a discussion to either call someone something derogatory (even when truly warranted) or to make a statement about their intentions. DCDuring (talk) 18:33, 15 July 2023 (UTC)[reply]
@DCDuring Calling a lie a lie isn't derogatory, and thinking that way only enables the small minority of users who do act in bad faith. I'm not saying it should be done lightly (because it definitely shouldn't be), but when a user has a long history of dishonest behaviour then we shouldn't shy away from that. Theknightwho (talk) 18:42, 15 July 2023 (UTC)[reply]
Even when you are 'sure' that you understand the intent of a given user, it is nonetheless the wiser course of action to AGF and refer to the 'lie' as a 'mistake' — no matter what. DCDuring (talk) 22:06, 15 July 2023 (UTC)[reply]
It's because time and time again you refuse to have a normal discussion and always take the opposing stance. I tried being more tolerant in the beginning but once I saw you had no intention of finding compromise or anything I gave up after the fifth or sixth time. Perhaps you could be less contrarian. Vininn126 (talk) 18:41, 15 July 2023 (UTC)[reply]
@Vininn126: I have never refused to have a normal discussion and there is nothing wrong with taking opposite positions and I do not always do this. I have the intention to seek a compromise (or anything), otherwise just look at what you have done with the Silesian language. I have no contradictions, I would have already apologized. ɶLerman (talk) 18:49, 15 July 2023 (UTC)[reply]
BTW, this topic was not introduced solely for the immediately preceding discussion and even less the chastise you. There have been a few discussions that have reminded me of what I learned trying to teach Principles of Management and Interpersonal Behavior to business undergraduates using case method and role-playing. DCDuring (talk) 18:28, 15 July 2023 (UTC)[reply]
Yes, there have been quite a few heated discussions lately. Maybe it's the summer heat, lol. I myself have spoken rashly in discussions lately, despite usually being pretty calm. I might also add that in the context of a written discussion, taking breaks between replies can be helpful, rather than going back and forth every few minutes. Otherwise emotion can take over. Andrew Sheedy (talk) 18:34, 15 July 2023 (UTC)[reply]
I will say that I am personally in favor of actual freedom of speech on the website, within whatever limits the WMF and the law itself may have. But I recognize that people will feel that wild conversations don't build the dictionary, and hence they will want to prevent or stop those conversations. --Geographyinitiative (talk) 19:49, 16 July 2023 (UTC)[reply]
I am a big fan of freedom and toleration. But I also would encourage all regular contributors to exercise their freedom of speech to avoid modes of speech that are viewable as attacks on persons. This has nothing to do with identity politics or political correctness. It is a matter of civility in support of frank and productive discussion. DCDuring (talk) 20:48, 16 July 2023 (UTC)[reply]
Isn't administrative tools already enough? Why did several admins in this thread express that, in addition to the administrative tools, they still need to retain the choice to verbally abuse other users they confront?
I had been thinking admins are like a support team of Wiktionary. And a support team should not keep yelling at and insulting perhaps ignorant and stupid customers even when they feel frustrated. This particular behaviour does not solve much problem and is instead the worst thing a customer may expect. If admins are not well trained in or are too stressed to be talking in a decent way that contribute to a friendly atmosphere, can they at least talk in a more bureaucratic way like "according to this rules, I will.../according to the procedure, you should.../please find other users who support you" that will not risk escalating the tension? 恨国党非蠢即坏 (talk) 09:23, 17 July 2023 (UTC)[reply]
That would help. Acknowledging one's own fallibility or at least that others might honestly find one fallible also might help. DCDuring (talk) 19:12, 19 July 2023 (UTC)[reply]
Do world/regional events affect implementation of our policies?
In the "Inactive admins" thread above it was mentioned that we should discuss this, specifically whether we should exclude the years of the covid pandemic in calculating whether an admin has been inactive for 5 years or how that should be handled. I'd like to expand this question and ask: Do world or regional events (e.g., war in Ukraine and resulting chaos) have any relevance on extending waiting periods in our policies if the policy says nothing about mitigating factors? I think the answer is clearly no and we should not have future enforcement of our policies delayed by questions of whether event XYZ in the year XXXX affects policy implementation.
Of course, if in the future someone wants to change a policy and how it's implemented, they can always initate a vote for that. Megathonic (talk) 03:30, 16 July 2023 (UTC)[reply]
@Megathonic I am inclined to agree. It's a slippery slope if we start making all sorts of policy exceptions "just because". IMO if there's ever a world event big enough to count (e.g. another World War?), it will be obvious because it will significantly disrupt Wiktionary itself (and in such a case probably we will have more pressing issues to worry about in any case). Benwing2 (talk) 04:24, 16 July 2023 (UTC)[reply]
This is a poor example of a force majeure event, tho: the COVID-19 pandemic left millions of persons without daily jobs and stuck indoors, online for hundreds or thousands of hours extra. If anything, it should have increased the availability of your average admin. —Justin (koavf)❤T☮C☺M☯19:26, 16 July 2023 (UTC)[reply]
The WHO declaration of pandemic was from March 11, 2020 - May 11, 2023. With that kind of more concrete period in place, 3 years and 2 months, I would suggest applying it if it feels meaningful for a given situation. --Geographyinitiative (talk) 19:43, 16 July 2023 (UTC)[reply]
I think this is an impossible question to answer in general, because it massively depends on the scenario in question. In this case I think we probably have been lenient due to the pandemic, but at this point I think that leniency should probably come to an end. Theknightwho (talk) 08:35, 17 July 2023 (UTC)[reply]
Do we have a standard way of transcribing the pronunciation of circumfixes? The only three "inhabitants" of Category:English circumfixes have three different solutions:
I want to add/harmonize entries for some major Austronesian languages, but want to confirm about what would be best practice here. Personally, I don't like version 2 at all (which e.g. is currently used in per- -an) Austronesier (talk) 11:41, 16 July 2023 (UTC)[reply]
I must confess I'm drawn to number 4 - /ɛnˈ- -ən/. It does have the disadvantage of seeming to place the stress on the first syllable of the base word - I don't think the stress shifts. Of the three in use, I prefer the first. --RichardW57 (talk) 13:02, 16 July 2023 (UTC)[reply]
I've noticed that the anagram bot hasn't run in a while. I just had to edit the pages for furanose and fursonae to point to each other; since the latter has been created in April 2022, the bot has been inactive for at least that long. Besenj (talk) 15:23, 16 July 2023 (UTC)[reply]
In his ignorance, @Kwamikagami has been categorising (to use the vernacular) most characters of the Thai block as translingual or not according to what actual languages the entries on the characters' pages were in. (This was before I quenched a few orange links this morning (16 July 2023) by adding a few definitions for Pali letters.) Thus a few classifications as translingual have actually been converted to classifications as Thai or Sanskrit, while having an entry for Pali as well as for Thai has reasonably led him to classify the letter as translingual.
Does anyone object to characters used in both Thai and Pali being considered as therefore translingual? I note that @Kwamikagami does not consider usage in both Chakma and Pali to demonstrate translinguality - or else he wasn't looking at quotations! --RichardW57 (talk) 21:34, 16 July 2023 (UTC)[reply]
I've been placing mislabeled info under the header of the language being described. If I made a mistake, my apologies.
Personally, I'm not sure that a language such as Pali or Sanskrit, which will be written in whatever the local script happens to be, is reason to call that script "translingual". It is Thai-speakers who write Pali in Thai, Lao-speakers who write it in Lao, etc.
Let's take a hypothetical: a script that is used only for its ethnic language X and for Pali/Sanskrit. Should we have 3 sections, for X, Pali and Sanskrit, or add a translingual section and have 4 sections, even though the 4th adds no information?
In any case, a description of how the Thai script is used for Thai does not belong under 'Translingual'. A translingual section needs to be about translingual usage. kwami (talk) 21:55, 16 July 2023 (UTC)[reply]
@Kwamikagami There are a number of languages local to Thailand, and some of them use Thai characters as at least one of their scripts. A quick scan turned up the Northern Thai language and Southern Thai language, but I would be surprised if there weren't others. That was very easy for me to find, so you should have already known about that.
You've set yourself up as the arbitrar as to whether any single-character entry or language section should exist. With that comes the responsibility to know all the relevant facts and considerations before you act. It looks like you've failed miserably in that respect. If you don't want to perform that kind of background investigation on every single thing you do, follow the process so the community can take that responsibility for you. Chuck Entz (talk) 22:46, 16 July 2023 (UTC)[reply]
@Chuck Entz Yes, I'm aware of that the Thai script is used for other languages. That's not the issue. The issue is that we have very clear consensus that a translingual section should not be about a specific language. If we say that "X is the nth letter of the Thai alphabet", then that is specifically about Thai. As such, the section header should be Thai.
Also, it would be odd to have definitions of letters of the Thai script and not cover Thai. Looking at other scripts, we cover Georgian for Georgian letters, Japanese for Japanese letters, etc. If someone wants to create translingual sections in addition to primary ethnic ones (i.e. the language the script was originally designed for / developed in, and that often gives its name to the script), that's great, but we've already had a discussion on deleting the e.g. Georgian and Japanese entries and rolling them into a translingual section, and that option was voted down. kwami (talk) 22:59, 16 July 2023 (UTC)[reply]
A lot of what you may take to be specifically about Thai applies across languages. That is by design; the people of Thailand don't seem to cope well with Thai and their local language having different reading rules. That is probably why the more official writing system of Pali that uses implicit vowels seems to be being replaced by the one that uses explicit vowels. --RichardW57 (talk) 23:18, 16 July 2023 (UTC)[reply]
@Kwamikagami: The comment in my second paragraph is independent of the issue in the first, which wasn't really a good example. My point was that you're making a lot of decisions without doing the factual and conceptual groundwork first. You're also forcing people to immediately come up with justifications in order to keep you from doing more damage, with the result that people aren't having time to think things through and are making mistakes. The result is a certain percentage of nonsense coming from both you and the others. This is a really lousy way to make basic decisions about the architecture of our site- measure twice and cut once. Chuck Entz (talk) 01:32, 17 July 2023 (UTC)[reply]
@Kwamikagami: Please give a link to this discussion. I cannot find any Georgian letters outside the Georgian script on Wiktionary, so I worry that your memory is at fault. Or is this part of your presenting the inconclusive vote on hacking back the number of letter entries as a consensus for proliferation? --RichardW57m (talk) 10:28, 18 July 2023 (UTC)[reply]
Can you be more precise, please. I'm looking for the discussion you referenced. I had found nothing under Turkish letters. and I've now found it was 'Kurdish' not Georgian (my memory problem), but we actually have a massive dearth of non-translingual letter entries for Kurdish languages. --RichardW57m (talk) 11:20, 18 July 2023 (UTC)[reply]
The only discussion I referred to was yours, where you wanted to merge all languages into a translingual section and say there is no consensus not to do that. kwami (talk) 01:49, 19 July 2023 (UTC)[reply]
There is more than one way: if the contention is that the existence of a Translingual entry is based on invalid interpretation of the rules or the agreed-upon facts, {{rfd}} is the best choice. If the contention is that there's nothing categorically wrong with the entry, but that it should be under a different language header, {{rfm}} might be better. {{rfv}} isn't so useful in this case, because it would hinge on whether there is usage for the term in the language in question, and translingual usage is hard to work with for the Symbol or Letter POS- to establish usage, you first have to answer conceptual questions that are best dealt with via {{rfd}}. Chuck Entz (talk) 01:32, 17 July 2023 (UTC)[reply]
@Kwamikagami Will usage of a character in both Thai and Northern Thai establish it as translingual? If not, may it contribute to the evidence for a Thai character being translingual? Incidentally, if Thai characters are used to transliterate (as opposed to transcribe) Northern Thai in the Tai Tham script in a book whose main language is Thai, is it being used as a Siamese character or as a Northern Thai Thai-script character? I have lots of durably archived transliteration of this nature. --RichardW57m (talk) 11:47, 17 July 2023 (UTC)[reply]
Will usage of a Thai-script character in both Thai and in words in a Mon-Khmer translation of the Bible establish translinguality? --RichardW57m (talk) 11:47, 17 July 2023 (UTC)[reply]
IMO, use of Thai script for the orthographies of multiple languages would count as translingual as we currently use it (i.e. as a synonym for multilingual). That would include Isan, Pattani Malay, Karen, or anything else (presumably but not necessarily in Thailand.) It may or may not be useful to create a translingual section, but that's a judgement call.
Thai words in Thai script, embedded in a text written in another language and using a different script, would not establish it. Take a Western parallel: Hebrew words are often written in Hebrew script in English texts, but they're Hebrew, not English. They're not even Hebrew loans into English, they're explicitly Hebrew. In such a text, Hebrew script is used for Hebrew and Latin script for English. Thus Hebrew script is not used for more than one language in that text. It is of course elsewhere, e.g. Yiddish and Ladino, but the English text is irrelevant to a judgement of whether Hebrew script is translingual. The same for an English text with Greek words written in Greek script: Greek script is translingual, but not because of that. Similarly, if a text written in Khmer and using Khmer script had Thai words written in Thai script embedded in it, that would not establish Thai as a translingual script because it's only being used for Thai. kwami (talk) 22:49, 17 July 2023 (UTC)[reply]
@Kwamikagami: Is the Thai script for Pattani Malay actually used? I've got the manual, but I have a feeling it might actually be life-threatening to use it, but no hard evidence. --RichardW57 (talk) 03:44, 18 July 2023 (UTC)[reply]
I have no evidence that it is used for Malay. I was just saying that if it were (and not just for the occasional ad hoc transliteration, but as an orthography), then that alone would IMO count as translingual. Not that I personally would find a 'translingual' section useful in such a situation, but that's a different matter.
Thai isn't as easy a case to make as Mon-Burmese, but it shouldn't be difficult to find attestation. Hell, we can attest to Armenian script being used for Turkish. And of course there's the matter of Southern Thai being considered a distinct language by some sources. kwami (talk) 07:41, 18 July 2023 (UTC)[reply]
There is a government-issued Thai script orthography for Pattani Malay; as I said, I have the manual. It's the actual use I'm not so sure about.
Still, at least for Northern Thai, there are accounts of interpreters being brought in for dealings between Central Thai and Northern Thai speakers. --RichardW57 (talk) 08:47, 18 July 2023 (UTC)[reply]
So why declare Cyrillic letters translingual? The people are just using the local script, i.e. the script of the rulers.
As a matter of history, in Thailand, Pali used to be written in the Tai Tham script in the northern areas, both Northern Thailand and northern North-East Thailand, and in the 'Khom script', i.e. a style of the Khmer script, in the rest of the country. In Laos, Pali used to be written in the Tai Tham script, as did Lao for matters related to Buddhism, such as sermons and notes in religious manuscripts. Thailand switched to using the Thai script for Pali in the 19th century.
In Northern Thailand, what happened is that there was a profane script, basically Fakkham, used for the local Tai dialect, and a sacred script, originally part of the Mon writing system, for Pali. The sacred script was extended to support the vernacular, and this system was borrowed by the Lao when they became literate. In Northern Thailand, the vernacular system died out and was replaced by the Tai Tham script, until the government in Bangkok decreed that schools should teach the national alphabet. However, the men of my wife's parents' generation still also(?) learnt the Tai Tham script.
Stop playing stupid. You've been doing it a while, and it's an annoying tactic that no-one respects.
Cyrillic is used for multiple languages by the speakers of those languages. That's what "multilingual" means -- you can look it up in this dictionary if you need to. If a script only only used by speakers of a single language, then it isn't multilingual. If it's only used for their language and, say, for Latin or Greek scripture, then I'm not sure if we should call it 'translingual' or just list the ethnic-language, Latin and Greek uses separately. We'd need to do that anyway.
As for your last claim, you've already tried to get agreement on that and lost. Give it up. At the very least, you should start a second discussion because you're dissatisfied with the results of the first one, and maybe this time you'll succeed. If you have consensus, fine: we'll remove the English and Latin sections from the letter a, the Greek sections from the letter α, the Japanese section from the kana articles, etc. kwami (talk) 22:52, 16 July 2023 (UTC)[reply]
@Kwamikagami: I did look the word 'translingual' up on Wiktionary, and its glossaries, and it was disappointing. It doesn't help, and doesn't include the requirement for multiple L1's. --RichardW57 (talk) 23:11, 16 July 2023 (UTC)[reply]
Are you aware that many translingual Burmese script letters are in the Burmese alphabet? (Not all of them - there are Shan tone marks that are used in multiple languages, though possibly only in the Shan linguistic area.) --RichardW57 (talk) 23:11, 16 July 2023 (UTC)[reply]
That's why you should respect the consensus here on Wikt.
Am I aware that the Burmese script is used for Burmese? No, I had no idea. What an interesting concept, to write your language in your alphabet! kwami (talk) 23:16, 16 July 2023 (UTC)[reply]
@Kwamikagami, RichardW57 I have no strong opinions about how to handle Translingual vs. language-specific characters, but Kwami you really need to stop unilaterally making changes while this discussion is happening. Furthermore you seem to be trying to manufacture a consensus about this matter that to me doesn't obviously exist. Some characters have language-specific definitions, some have Translingual definitions, and there do not appear to be clear guidelines as to how to handle this. If you want to resolve this, the proper way is absolutely *NOT* to edit war to get your way, but instead to make a specific, detailed proposal and get consensus through discussion. This may take awhile but there is no fire here that needs to be put out yesterday; building consensus takes time and that is OK. (I know edit warring is often the norm on Wikipedia, but it's not tolerated around here.) Benwing2 (talk) 23:58, 16 July 2023 (UTC)[reply]
Actually, there is a real question in there about Cyrillic. I could see restricting 'translingual' to items that exist across languages but that do not occur in their alphabets. Greek qualifies, because it's the international norm in mathematics and many scientific disciplines, and Greek letters are therefor used in many languages that are not written in the Greek script. Latin qualifies. The IPA and Hindu-Arabic digits. I'm not sure Cyrillic or Arabic qualify, but if, say, a language that has a Cyrillic or Arabic-script alphabet uses letters for section headers despite those not occurring in its alphabet, then that would be an instance of translingual use. That of course would change quite a few entries, where something is called "translingual" because it occurs natively in multiple languages, but it would be a possible approach. kwami (talk) 00:20, 17 July 2023 (UTC)[reply]
That brings up something that's missing from Wiktionary - the use of Thai letters for numbering. It may need some evidence that is awkward to quote - I have a suspicion that there's variation in what letters aren't used, as some automated Thai numbering by letters didn't act as I expected. It's the sort of thing that Microsoft can accidentally change. --RichardW57 (talk) 00:34, 17 July 2023 (UTC)[reply]
What is the process for changing the language of an entry from 'Translingual' to something else? @Kwamikagami is still doing it directly (within the past hour) rather than going through any process. Four of the latest victims are Chakma letters 𑄢, 𑄥, 𑄖 and 𑄚, all of which already include Pali homonyms. --RichardW57 (talk) 12:09, 16 July 2023 (UTC)[reply]
If you wish to add sections for Pali, great, but it's weird not to have a section for the Chakma alphabet. Also, should the label be changed if there is no section for Pali yet, but not changed if there is? That inconsistency would also be weird. kwami (talk) 22:02, 16 July 2023 (UTC)[reply]
@Kwamikagami: You're not reading the pages properly. In abugidas, Pali has quite a few one letter words other than the names of letters. The homonyms I am referring to are neither letters nor names of letters. Therefore it is not weird to have an entry for the translingual letter, and none for the individual languages. The only translingual letter which might also need a Pali entry is Translingual 𑄤, which has the peculiarity of only appear as a subscript letter in Pali. (I think the monks were stupid to create Pali 𑅇(va).) --RichardW57 (talk) 22:19, 16 July 2023 (UTC)[reply]
You have already asked if all languages should be conflated into a translingual section, there was a formal discussion/vote about it and the consensus was no -- something you were reminded of recently -- so why are you still trying to do it? kwami (talk) 22:33, 16 July 2023 (UTC)[reply]
Alright, no consensus to change the current consensus, which is established by the thousands of articles that follow it. (Though I'm not sure how that differs from simply calling it "consensus".) kwami (talk) 23:41, 16 July 2023 (UTC)[reply]
@Kwamikagami, RichardW57 Can we (a) please move this discussion to the Beer Parlour, where it belongs, and (b) not edit war while this discussion is going on? It would be best to not touch any existing single-letter entries until some sort of consensus is reached, even if this takes awhile. Benwing2 (talk) 23:43, 16 July 2023 (UTC)[reply]
@Benwing2: Do you mean move it bodily? I was hoping there would be an uncontroversial answer to my question. If so, who is to perform the move?
Alerting @Kwamikagami, Chuck Entz. I think we have a digression here, starting at 'if you wish to add', so one could move that instead, but then I'm not sure what to call it - perhaps 'Translingual' v. 'Entry per language'. --RichardW57 (talk) 00:17, 17 July 2023 (UTC)[reply]
@RichardW57 There is no specific procedure for changing an entry from Translingual to another language. IMO the answer to that question would answer the overall issue about how to handle single-character entries, or at least is intimately tied in with that issue. By move, yes it is fine to cut and paste the whole discussion and replace the text of this section with an indication that it's been moved to the Beer Parlour. I have done that in various instances when discussions e.g. related to a particular module have started somewhere else e.g. on the Talk page of a particular entry, so that the discussions are consolidated and all the relevant info can be later found. Benwing2 (talk) 01:15, 17 July 2023 (UTC)[reply]
@Benwing2: Where should I assemble the evidence of translinguability pending the determination of the meaning of the term? I suggest User:RichardW57/mul_evidence, which are others are invited to create or contributive positively to, even by adding challenges to the existence of terms in specific languages.
I don't know what you're asking. If you mean adding a proper dictionary definition to the entry, then yes, that's what we're here for. But if you mean adding more material under the wrong header, or trying to pass off a language as something other than what it is, then no, that's what I have a problem with. kwami (talk) 08:38, 22 July 2023 (UTC)[reply]
@Kwamikagami, @Benwing2: I unthinkingly added some explanation ('trivia') to Mon ဣ(adverb) to shore up the entry and provide some sort of external dictionary back up. Should I undo that change pending resolution? Silence will be taken as consent to leave it. I myself have no particular objection to others enhancing the entry for the adverb, despite its being spelt with a single character. --RichardW57 (talk) 08:28, 22 July 2023 (UTC)[reply]
I don't think 'trivia' is appropriate. Halliday's analysis is relevant IMO but should be under some other heading, such as 'user notes' or perhaps as part of the definition. kwami (talk) 08:33, 22 July 2023 (UTC)[reply]
It was correct, but double quotes make it more obvious that 'verb.' is part of the quotation. 'User notes' is not a valid header, so WT:EL puts it under 'trivia'. As I said, I don't mind improvements. One might reduce this adverb to a soft redirect to the whole affix. --RichardW57 (talk) 09:10, 22 July 2023 (UTC)[reply]
If you mean "verbal affix", then IMO you should spell that out and move the quotation mark to just before "denoting". We generally try to avoid abbreviations in explanations.
Do you mean an {{rfd}}? That did occur to me, and that would be consonant with the principle of 'bugger the user'. We could also rearrange the adverb's entry to move the comment on the definition and the 'trivia' into 'usage notes'. What I have in 'trivia' doesn't count as usage notes on its own.
By "lying", I can only assume you mean you don't know what "verb." stands for, as there is nothing even approaching a lie. But if you don't understand what you've read, you shouldn't quote it. kwami (talk) 20:39, 22 July 2023 (UTC)[reply]
This question originated on Koavf's talk page, under the section entitled "WF". See attached talk page discussion for more details regarding this. If you have any other pieces of talk in regards to this topic, link them and I'll put them into the Discussion section.
I'm not including the Short vers., because I don't think it impacted this topic as the other pages. There could be some relevance in the vote, but not enough to where it merits a link in the main post. cf (talk) 05:48, 17 July 2023 (UTC)[reply]
In my experience Wonderfool means well but is sloppy; he is trying to get his edit count up so he works quickly, fairly often makes mistakes, and can't be bothered to fix them later even if asked (although in truth it is no worse than User:SemperBlotto, who is a bureaucrat but has produced a lot of shoddy content, rarely or never cleans up even when asked to, and has continued in the same vein year after year [although he has been inactive lately] -- were he not an admin I might have blocked him by now for this sort of behavior). Sometimes Wonderfool gets frustrated or burnt out and goes on a vandalism rampage (although personally I only experienced this once so I don't know how common it is). Blocking him has no effect, but per User:Chuck Entz reverting might get his attention as (presumably) it lowers the edit count. Benwing2 (talk) 05:41, 17 July 2023 (UTC)[reply]
Reverting would not, but mass deleting pages he creates would. I have been reluctant to do that because some of his edits are of quality. Note also that I have not seen his edits be wrong, but I have seen many that are incomplete and could easily be stronger. If Wonderfool wanted to collaborate on Spanish (from one account), then I would actually be really happy to do that. He also adds a lot of links that are things a clever bot could do, so his insight could be really helpful. All that aside, I don't see it as valuable to actually remove his useful edits even if he gets a kick out of a high edit count: that would be detrimental to the project. I think it's also detrimental to the project to allow someone to hop around hundreds or thousands of accounts. I'll try to not insert myself much more, but I did want to make the point about the quality of his Spanish edits as I have some competency there. I will also point out that I have been told by a native Bosnian/Croatian/Montenegrin/Serbian editor that Wonderfool has made completely wrong Serbo-Croatian edits, but I am too ignorant of Slavic languages to even comment. —Justin (koavf)❤T☮C☺M☯05:55, 17 July 2023 (UTC)[reply]
I think they are mediocre, but I've only ever see them be accurate. Accurate-but-imprecise or accurate-but-incomplete are valid entries to make. Which is why I hope we could collaborate, because working together, we could be more effective. :/ —Justin (koavf)❤T☮C☺M☯06:03, 17 July 2023 (UTC)[reply]
I spent a lot of effort cleaning up WF mistakes in Spanish entries. Generally WF doesn't understand templates well or can't be bothered, so a lot of adjectives had incorrect feminines and/or plurals. Similarly noun plurals were often wrong. Since then I revamped the templates to have much smarter defaults, so the number of errors is likely to go down, but if WF doesn't understand the template operation he'll still make mistakes in cases where the defaults need to be overridden. Benwing2 (talk) 06:26, 17 July 2023 (UTC)[reply]
@Theknightwho Yup, I know that, in fact I did a bot run recently to change the description of all of his audios to "Southern England" per discussion with User:Equinox and User:Sgconlaw; seems that WF was often intentionally mislabeling the location of his audios. In some cases I removed the audio because e.g. it claimed to be a Northern England or US pronunciation but was obviously in a Southern England accent. Benwing2 (talk) 09:30, 17 July 2023 (UTC)[reply]
I had no idea anyone had done this. Thank you! (To clarify: WF seems to speak like a south-easterner, or Londoner, but he seems to have a habit of randomly tagging his stuff as Berkshire, Hampshire, etc. depending on troll mood.) Equinox◑05:33, 26 July 2023 (UTC)[reply]
Not to get off the original topic, but won't DerbethBot just re-add any audio files which have been removed? Did the bot stop running and/or did we get anywhere with a blacklist of known-bad files? IIRC Metaknowledge took point on the latter idea for a while. - -sche(discuss)06:04, 26 July 2023 (UTC)[reply]
OK, looking at DerbethBot, it says the bot operator prefers not to use a blacklist and will readd the files every week; his preferred solution is that you RFD the files on Commons. (If Commons keeps them because "it's a valid pronunciation, even if mislabelled", then we have to decide what to do.) - -sche(discuss)08:21, 26 July 2023 (UTC)[reply]
Hello. I no longer run my bot every week, there aren't enough changes, I run it every couple of months, depending on the number of updates I detect on Commons. The bot keeps a record of visited pages. A page is not re-visited unless a new audio file appears. This is to speed up bot execution (otherwise, just matching Dutch audio files with pages here will take several days). I can manually clear the 'record' to force the bot re-visit pages and add missing audio files. But I need hint from you what happened, because the wall of text here looks intimidating. --Derbethtalk09:22, 26 July 2023 (UTC)[reply]
First, please note that I'm active at other wikis, and haven't yet decided to become active here. Sockpuppetry, which is what Wonderfool has been engaging in for a long time, is not allowed at any wiki that I know of. Sockpuppetry is an abuse of normal editing processes, and any user who abuses multiple accounts should be blocked on sight. This isn't to say that sockpuppeteers at Wiktionary shouldn't ever be allowed to edit here; that's not correct. However, to edit, they need to respect the concept that abusing multiple accounts isn't acceptable, and they need to either request an unblock or edit only though a single account. Compromising and allowing them to edit—since, apparently, their sockpuppetry is difficult to handle—shouldn't be an option. There are many problematic implications to sockpuppetry, and when a person has clearly demonstrated that they couldn't care less for this policy, they shouldn't be trusted to do so. Nythar (talk) 06:38, 17 July 2023 (UTC)[reply]
It's important to remember that this has been going on for well over a decade, and that (in the past) consensus for militantly blocking WF was much stronger. It failed then, and I don't think we have many new tools at our disposal now. WF is a problematic contributor at times, but looking at the last couple of years I can't see any massive problems that would warrant a permanent block. I just don't see what this is in aid of anymore. Theknightwho (talk) 07:35, 17 July 2023 (UTC)[reply]
@Theknightwho: The abuse of multiple accounts is good enough to be the only reason to block their socks. But for some reason I can't find any relevant policies; do you know where to look? Nythar (talk) 07:38, 17 July 2023 (UTC)[reply]
@Nythar I know, but there is also a wide tolerance for WF among the community, which very clearly changes the circumstances since many editors (including many admins) do not feel particularly strongly about blocking WF, or (like me) just think the whole thing's a bit pointless by now. There has been a somewhat steady status quo for a long time, and this recent campaign against WF has been solely driven by one admin, with lacklustre support from a a few others at best, and quite a few people opposed.
Perhaps we can put this to bed by having a vote on whether to unblock WF, but I'm not entirely convinced it'll work, since the guy thrives off being an outsider. Yes, it's unorthodox, but fundamentally I just want the project to thrive. Theknightwho (talk) 07:46, 17 July 2023 (UTC)[reply]
@Theknightwho: I understand that the community is generally not opposed to Wonderfool's edits, and I understand that they've made many helpful contributions. But from what I've heard, that user doesn't engage collaboratively at times, and they have a tendency to make mistakes, in addition to their other long-term abusive behavior. It wouldn't set a good precedent to allow such a person to edit. Benwing2's proposal below is the only reasonable possibility I see here; otherwise, I suppose it would be wise to implement Megathonic's proposal. Nythar (talk) 07:57, 17 July 2023 (UTC)[reply]
@Nythar Being realistic, it doesn't set any kind of precedent - this has been going on for over a decade now. The recent threads are not because of anything new that WF did, either.
@Theknightwho: Thanks for explaining. I think the only reasonable path forward is to seek consensus on whether Wonderfool should be allowed to edit from one account, in line with Benwing's proposal. Following such a proposal (whether or not they are unblocked), stricter measures should be enacted to DENY any troll-ish activities. @Benwing2: and @Megathonic:, are we in agreement? Nythar (talk) 08:12, 17 July 2023 (UTC)[reply]
Who are you, sorry? Seems quite strange to me to pop up in a community you've never participated in before in order to make these sorts of proclamations, but I might be missing something. —Al-Muqanna المقنع (talk) 08:16, 17 July 2023 (UTC)[reply]
I was hoping they'd respond here, but since this hasnt been answered yet, I'll reply. Nythar has over 66,000 edits on the English Wikipedia, and although I can't find it now, I think they said that Wonderfool's name came up in a recent sockpuppet investigation there. Just to be clear, I think the context was a different user being compared to Wonderfool, not an investigation involving Wonderfool himself. —Soap—10:16, 28 July 2023 (UTC)[reply]
@Nythar I'm tentatively supportive, but I do want to make it known that this does carry the risk of WF intentionally trying to get blocked again and coming back with a vengeance if we try to get more militant. It's hard to know, because some people thrive off being like this, and we need to take a practical approach to managing them if we want them to be productive contributors. Theknightwho (talk) 08:18, 17 July 2023 (UTC)[reply]
I would vote to unban him, but I agree with you that it may not work, as Wonderfool has hinted at us that he feels perfectly ready to go on the way he has been, even if it means a new account and a new block literally every day. And he has at least twice has asked us to start enacting rangeblocks, reminding me of a disobedient child who begs for ever greater punishments to humiliate the parents who think they can change him. —Soap—07:57, 17 July 2023 (UTC)[reply]
Maybe we should simply ask Wonderfool to "come in from the cold" and start using a single account, for which we will agree to apply standard blocking policies, i.e. they don't get permablocked the first time they do something blockable, but get blocked according to standard rules, which double progressively if the blocking behavior continues. I don't personally see much blockable in WF's contributions, but if we don't at least give them a chance to behave sensibly then of course they'll continue what they're currently doing. If they use this "normal" account to start doing abusive things, that's a sign they need to be blocked more aggressively and their sockpuppets blocked strongly. Benwing2 (talk) 07:44, 17 July 2023 (UTC)[reply]
I wouldn't oppose trying this, but I think it should be put to a vote to clearly demonstrate consensus before we "locally" unblock a globally blocked user. Megathonic (talk) 08:06, 17 July 2023 (UTC)[reply]
Tolerance of Wonderfool's nonsense needs to end. All of their sockpuppets should be blocked on sight. The page User:AryamanA/Wonderfool should be nuked, or if the community really wants a centralized list of WF's sockpuppet accounts, then at the very least the "# of edits" column should be abolished. Wonderfool thrives off driving up their "net" edit count and there's no reason for us to play into that and document it. We should adopt Wikipedia's practice of allowing all edits performed by a blocked user to be reverted by anyone for any reason. Any page that WF creates should be speedily deleted, and in his case that should include useful edits in order to attack his edit count. To the maximum extent possible, WF should be erased from this Wiki going foward until they decide to follow policy. Given that they've been sockpuppeting and vandalizing since 2005, I'm not sure if that will happen, in which case WF should not feel welcome here. Megathonic (talk) 07:11, 17 July 2023 (UTC)[reply]
I think Wikipedia needs that rule because many of its banned and blocked editors are people who have used Wikipedia to push political viewpoints or to vandalize. Those things happen here as well, but much less frequently, and I think even Wonderfool's outspoken critics agree that most of his edits are helpful. To my knowledge, the issue has never come up for discussion here, but if we were to explicitly adopt a rule that all of Wonderfool's edits could be reverted by any uninvolved user, it would change very little, as many other editors would defend those edits and make them their own. I say this assuming that we have no intent of blocking editors who accept the pages and edits created by Wonderfool ... on Wikipedia, people who edit on behalf of blocked editors can themselves be blocked, but as I stated at the outset, Wikipedia's blocked editors are very often political operatives or vandals. —Soap—07:35, 17 July 2023 (UTC)[reply]
If another editor recreates one of Wonderfool's deleted edits/pages and have reasons of their own for doing so, then that's perfectly fine. But if they are editing on behalf of a blocked user, thus turning themselves into a proxy, then I'd say that should be grounds for a block (assuming they don't stop when told to) and clearly stated in our policies. Megathonic (talk) 07:49, 17 July 2023 (UTC)[reply]
I don't really understand how this approach would benefit the project - there are no other users who are in a remotely similar position, and the tolerance of WF up till now hasn't set any kind of precedent for other permanently blocked users. Theknightwho (talk) 07:55, 17 July 2023 (UTC)[reply]
There's no policy against having multiple accounts, only abusing multiple accounts. Given that WF accounts are immediately recognizable and WF never tries to hide the fact that he's WF, there's no deception involved. Whether it's one account or 100, We all know who they are. Multiple accounts don't mean the same thing in this context. It's not been uncommon for WF to hint at or even outright ask to be blocked so he can take a break- not as a mind game, but as a sincere request. Yes, they did some terrible things in the early part of their career, but they gave that all up when they came out in the open.
And yes, WF is globally blocked, but so is the main IP-address range for the third largest mobile provider in the US- a company with a tenth of a billion subscribers. I'm not about to tell the stewards who they should block, but that doesn't mean I have to blindly follow their lead, either. Each wiki has its own policies and practices. Chuck Entz (talk) 08:13, 17 July 2023 (UTC)[reply]
Ban evasion is an abuse of multiple accounts. Multiple accounts shouldn't be used for the purpose of evading a ban or block. Nythar (talk) 08:19, 17 July 2023 (UTC)[reply]
Accounts are blocked so that they can't edit. When another account is created in order to edit, that's block evasion. It's the same case with ban evasion. Nythar (talk) 08:23, 17 July 2023 (UTC)[reply]
I think this is mostly an academic point - I don't think anyone's disputing whether WF is in breach of policy (which he obviously is). Theknightwho (talk) 08:33, 17 July 2023 (UTC)[reply]
Then in that case, the "# of edits" column should be deleted, since it's not necessary for seeing what types of edits WF has done and instead just stokes their ego/encourages their behavior. Megathonic (talk) 08:35, 17 July 2023 (UTC)[reply]
@Megathonic I'm not sure it's entirely fair for you to suddenly start deciding that we need to change the page becauase you want to take a hard line. It's been the way it has been for a long time. Theknightwho (talk) 08:39, 17 July 2023 (UTC)[reply]
I'm not interested in getting into a shouting match with you. I've stated my view & am well within my rights to seek consensus to change the page. If you disagree, that's ok, but please don't imply that I'm possibly being unfair. Megathonic (talk) 08:44, 17 July 2023 (UTC)[reply]
If you disagree on my edit, feel free to start another discussion about whether that should be there and is appropriate. As for a shouting match, I don't mean that one exists now, but rather that I don't care for one to start and for us to get into a back-and-forth. So I'm disengaging. Megathonic (talk) 08:54, 17 July 2023 (UTC)[reply]
I strongly agree here. I honestly still do not understand the support for WF here, and it's similar to my concerns that there's a tendency to protect vestiges of the website even when they do direct harm to the project. Didn't WF literally blank the Main page at some point? He became admin on multiple occasions and has nominated admin, some of which are participating in this current discussion. That doesn't read to me as someone who's been instantly blocked and is essentially treated even better than the average user. (We couldn't even fully support a vote to block blocked users from creating votes because of WF.) I truly do not get it, but let's not pretend that admin have been instabanning him for years. I agree with @Chuck Entz's assessment that it's not just a few rogue admins. AG202 (talk) 12:51, 17 July 2023 (UTC)[reply]
To be honest I struggle to care about WF's antics as an admin in the mid-2000s and don't think it's particularly relevant to anything given that he's unlikely to ever become an admin again. I very much disagree with the "rules are rules", "policy is policy" mindset: the rules should serve the project, not the other way round, and without any concrete examples of WF's behaviour spreading to other users over the last two decades it's hard to see what specifically was wrong with the pre-Koavf approach beyond insinuations about bad precedents and so on. —Al-Muqanna المقنع (talk) 13:21, 17 July 2023 (UTC)[reply]
"Fairly often makes mistakes, and can't be bothered to fix them later even if asked" if this were repeated behavior from a normal user and they refused to change, they'd be blocked, as I've seen multiple times even from yourself. AG202 (talk) 13:48, 17 July 2023 (UTC)[reply]
@AG202 Sure, but someone also pointed out that SemperBlotto was often like that as well, who I don't think anyone would seriously want to permablock (if he were still active) - his Italian contributions are semi-regular features in RFV. I've not personally seen egregiously bad edits from WF - mostly just lazy ones, and while they're annoying, it doesn't stand out to me as terrible. The real problem editors are the ones whose contributions I know I can't trust at all, because they're consistently bad - those definitely should be blocked if they keep at it. Theknightwho (talk) 13:53, 17 July 2023 (UTC)[reply]
And that's why folks like me have to keep playing cleanup and waste our time cleaning up entries because people don't want to take action on bad edits. I try to make sure that my edits are of high quality to help the project and avoid having people cleanup after me, but then at the same time I have to monitor to make sure that those entries and others don't end up broken because of repeated frankly careless edits that lead to no action taken. Even if people are necessarily more trusted, if they are making disruptive edits they should at least be blocked if they refuse to change. It adds to the feeling of not wanting to contribute if people who've been here longer have free rein to do whatever, while I have to keep pointing out disruptive changes and essentially have to beg for months for things like language code updates. While I'm here, I'll note that once again, I've recently noticed Korean links have been broken: circumflexes (^) are appearing in links once again, see the Usage Notes at: 한성(hanseong). AG202 (talk) 14:05, 17 July 2023 (UTC)[reply]
There is a line - I'm not saying WF has free rein to do whatever, but equally there are sloppy editors who also bring a lot of good, and while we try to get them to improve we also don't want to drive them away. It's a balance. Theknightwho (talk) 14:08, 17 July 2023 (UTC)[reply]
Making admins dedicate time to playing whack-a-mole with WF isn't going to make language code updates happen any faster either. —Al-Muqanna المقنع (talk) 14:27, 17 July 2023 (UTC)[reply]
That was not the implication, and is also not what I said. I'm more frustrated with the fact that I have to deal with repeat offenders AND at the same time, wait months for actual meaningful changes to go through. Much more effort is put into these type of discussions instead of the actual changes that have to go through. If the project can put this much effort into honestly protecting a permabanned user and tracking their changes and treating them like some kind of legend, then at the very least when our small community editors request that necessary changes be implemented, we shouldn't have to wait MONTHS on end each time after pinging multiple people. At this rate, I don't care if he's getting blocked every time or not, but he shouldn't have special pages made for him nor should he still be able to make large changes to the project. AG202 (talk) 15:17, 17 July 2023 (UTC)[reply]
@AG202 I get it - honestly - and I don’t want you to feel like Wiktionary’s hostile, because I think you’re a very valuable member of the community. I find it frustrating when I don’t get much engagement on things I care about as well, and sometimes it’s difficult to know what to do because you can’t force people to respond.
By the way, I’ve fixed that issue with the Korean translit - not sure why it started happening as there haven’t been any changes that stand out as an obvious cause to me, but it was something specific to {{ko-l}} ({{l|ko}} was working as expected). Theknightwho (talk) 15:38, 17 July 2023 (UTC)[reply]
I am fine with the status quo, but I would prefer that he edit under a single account going forward. My guess is that this is a non-starter for him, and in general his accounts are blocked pretty quickly so I'm not sure what more could reasonably be done. Mass deleting everything he does feels counter-productive to me. - TheDaveRoss12:26, 17 July 2023 (UTC)[reply]
@TheDaveRoss If you look at User:AryamanA/Wonderfool, WF was mostly editing under individual accounts for long stretches until Koavf decided to start blocking him way more often about 2 months ago. That period has seen 120 blocks, and it's just a massive waste of everyone's time. Theknightwho (talk) 13:49, 17 July 2023 (UTC)[reply]
I guess that if Koavf wants to spend their time hunting WF that is fine with me. I don't bother blocking WF most of the time because it doesn't accomplish anything, and his edits are mostly benign. - TheDaveRoss14:44, 17 July 2023 (UTC)[reply]
@TheDaveRoss I'm less fine with it, if I'm totally honest - this new development has effectively been a unilateral decision by one admin (Koavf), who has decided to dogmatically carry on with it despite several other admins and users asking him not to, and who has been very dismissive of the concerns of others (which @Chuck Entz, @This, that and the other and others can attest to). Regardless of what policy says, it doesn't seem like Koavf's doing this for the right reasons. Even if I'm wrong, then it still makes me concerned about having someone so literal-minded and uncompromising in the sysop role. Theknightwho (talk) 22:28, 17 July 2023 (UTC)[reply]
My takeaway from the interaction with Koavf on his talk page was that he was prepared to slavishly follow policy (as he interprets it) and its attendant guidelines, unwilling to acknowledge that the policy is deliberately written in a vague way, there are other interpretations of it besides his, and the guidelines are not mandatorily enforceable and subject to exceptions, just as is the case on his home turf, Wikipedia. I don't get why Justin suddenly decided to adopt this as a personal crusade; hopefully he'll eventually realise that WF is unstoppable and he is wasting his time. This, that and the other (talk) 01:38, 18 July 2023 (UTC)[reply]
That's kind of the impression I got as well. I'm not sure why it's necessarily to be so hardline about the rules simply on principle. As I think Aristotle said, laws/rules are for general cases and cannot envision every individual circumstance. There may be legitimate, often pragmatic, reasons for exceptions and since Koavf's such a valuable editor himself, I'd rather he be doing other things... Andrew Sheedy (talk) 02:10, 18 July 2023 (UTC)[reply]
I should also point out that he’s been extremely quick to assume bad faith of any admins who’ve cautioned him against this, and has taken a battleground stance at the drop of a hat (even in the face of attempts at conciliation). It's been really disappointing. Theknightwho (talk) 02:30, 18 July 2023 (UTC)[reply]
In my view, the power to block editors must be used when an editor is detracting from building of the dictionary. This is a for-fun, volunteer project, and ultimately you want a person that is willing to do work to work on the project. To me, blocking should not be about a mere rule breaking: it's about stopping someone from detractinf from the dictionary-building mission, which is more important than any specific rule or rules. I believe that the user has interesting, worthwhile edits to make, though they have done some absolute crap edits and has an unusual personality, and hence there is room for toleration of the editor. I say this without having experienced the burdens of leadership, and I totally understand the admins who block this person. It makes sense, and I know it's hard on you all to maintain order. I thank that editor for that editor's work, and I thank the admins for their hard work as well. --Geographyinitiative (talk) 09:50, 17 July 2023 (UTC)[reply]
Thumbs up! I think WF has done immense good for this project. The vast majority of his edits are improvements to Wiktionary and I have seen him correct his own vandalism many times. It's not great that some of his edits are vandalistic or low quality, but rarely do they actually harm the project. Although I understand koavf's approach, I personally oppose it, simply because everyone used to know which accounts were WF. Now that he keeps resorting to creating new accounts, it's harder to spot and monitor his activity. I am in favour of only banning him when he does something harmful to the project and having a sockpuppet is not one of those things when he isn't trying to hide who he is. Andrew Sheedy (talk) 17:21, 17 July 2023 (UTC)[reply]
When koavf blocks WF, they redirect the blocked account to User:AryamanA/Wonderfool. I suppose it can be argued that it's harder to recognize WF because it's not possible anymore to have the names of their active sock puppets memorized, but since this information (that it's WF) is instantly revealed by simply clicking on the user's name, I think the "harderness" is rather trivial. I'm not sure what the point would be of only banning him when he does something harmful, since he'll create another account in a matter of minutes; the ban itself won't change his behavior and he'll be rewarded by editing/participating in discussions/creating votes/etc. without being reverted.
If WF stops making socks to abuse blocks, I think it may be worth having a vote to unblock his original Wonderfool page to see if there is consensus for it, and allowing WF to edit from another account (since they won't be able to access their original due to the global lock). Right now I'd abstain on such a vote because I don't think WF would stop vandalizing if we readmit them, but I won't oppose others who think it could work. Megathonic (talk) 19:11, 17 July 2023 (UTC)[reply]
I'm mainly referring to being able to identify currently active accounts easily. I prefer being able to tell at a glance that an edit was made by WF, because I recognize the account name, than for the name to be different every week. Andrew Sheedy (talk) 21:17, 17 July 2023 (UTC)[reply]
Reading through this thread, I find myself arriving at one simple question:
Is a given editor behaving in a way that is a net positive, or a net negative, for the overall Wiktionary project?
WF's earlier time here was riven with disruptive episodes, and that has earned him several bans. More recently, I have not heard of anything he has done that is actually all that harmful. There are mild annoyances, like the sloppy or incomplete editing mentioned above, but nothing like deleting the main page.
In practical terms, the simple fact is that anyone can create any number of accounts. This is part of the underlying intentional design of the various WikiMedia projects. It is therefore literally impossible for us to prevent anyone from participating on a permanent basis. The best we can do is to slow someone down.
Considering that WF's activity over the past couple years, as far as I'm aware, has been mostly good-faith editing, I do not understand why we would want to slow him down. ‑‑ Eiríkr Útlendi │Tala við mig18:51, 17 July 2023 (UTC)[reply]
Note that I am not the only admin who has blocked Wonderfool socks on sight, purely for having been sockpuppets within the past few months. I can only assume they do not agree that it's okay to let him keep on using sockpuppets and they may have a perspective to add on their blocks. @Fenakhay: (e.g. and also) @Equinox: (e.g.), TheDaveRoss (e.g. and also)Sorry for the spurious ping, Dave, you have already posted, etc. I resent the implication or assertion that I am somehow alone in blocking sockpuppets as a matter of principle, including for Wonderfool. (Obviously, other admins also block sockpuppets, e.g. and also.) —Justin (koavf)❤T☮C☺M☯07:16, 19 July 2023 (UTC)[reply]
Okay, let's look at some stats. Starting from May, the block tallies on WF accounts are as follows:
Until yesterday, I was unaware of @Koavf's history on Wikipedia. He is apparently a highly productive contributor there with a long history of being blocked for edit-warring, and was, for a while, indef-blocked. This history adds another dimension to this. Without too much imagination, one could interpret his behavior here as at least partly motivated by jealousy: "why is he being allowed to edit when I wasn't? I wasn't given special treatment because of being a prolific contributor, but he has been. And I honored my blocks, but he doesn't." Chuck Entz (talk) 13:10, 19 July 2023 (UTC)[reply]
lol, no. This is textbook bad faith psychologizing and you are wrong. "A long history of being blocked" with a decade of not being blocked? When I was blocked there, I had plenty to do elsewhere, so no, I was not jealous of anyone, nor am I. —Justin (koavf)❤T☮C☺M☯15:58, 19 July 2023 (UTC)[reply]
I can easily give the reasons to these questions arising to the intuitive mind, as to Koavf they may be more unwont than they should. Equal treatment is regularly meted to those who are in the right. We call it Keine Gleichheit im Unrecht. Additionally it is a target only within the domain of the same governing body: When one state does X, subjects have no standing by reason that another state does Y, else we would level out the purposes of having different community projects. When could a culprit defend himself with others having gone unpunished yet? It is thus objectively illogical to generalize equiform action upon conceived offence. This is all even discounting the peculiar personalities and hence individual faults people have. They are hence banned for different reasons even when formal reasons have been stated, and the ban reason “Wonderfool” used to mean “Wonderfool-specific reason” rather than the distrust and disrespect created by the use of multiple accounts, which is the usual reason why people get extra-banned when circumventing bans via multiple accounts. So Koavf now banned Wonderfool for reasons that other administrators don’t find banworthy in Wonderfool’s case and then wrongly pretends the reasons to be the like. Fay Freak (talk) 14:16, 19 July 2023 (UTC)[reply]
This is pertinent and fitting here, though I warn that we construct it differently in civil law.
In detail: We don’t generally aim at equity (~ Billigkeit, Gerechtigkeit) between parties in the first place, but the lawmaker and administerer of the law is bound by subjective rights of the individuals (against the state) to equality—in treatment, not outcome or result; this is the long technical form for the assertion of “equality before the law”, written so short within the fundamental right context of a constitution:
To assess infringement of this equality before law, interference in it is checked by whether there is a comparison group (Vergleichsgruppe) that is treated unequally, and then a possible justification is checked: according to the crude earliest theory (so-called Willkürformel) by the criterion whether it has been arbitrary or rather has a factual reason and according to the more refined theory (Neue Formel) by whether there are differences of such sort and such extent that they justify the unequal treatment; the choice between more crude and more refined theory, to get the job done, also depends on whether individuals are bound to their comparison group or are rather able to leave it.
The test can also be reversed with a demand that unequal affairs need to be treated unequally, but this is practically obscure, if only for epistemological difficulties to make sense of this postulate, as it becomes unintuitive to dissect. Fay Freak (talk) 15:17, 19 July 2023 (UTC)[reply]
I see no consensus (yet), but it does seem like the most likely path towards consensus is to allow WF to continue editing. However, it is undesirable for WF to continue editing in a gray zone, especially one where no clear consensus exists regarding how to handle their sockpuppets. Consequently, I have created a vote to formally unblock WF and readmit them into the community. I intend to abstain on this vote: The vote's intention is to create consensus and put this issue to rest. Megathonic (talk) 08:41, 20 July 2023 (UTC)[reply]
And I voted in it after seeing it on my watchlist. Perhaps I didnt see the "Starts Jul 27" part. If we can leave my vote in place, that'd be great, since I've been following this intently, despite not having said much, and I'm confident I won't change my mind in the next seven days. If not, I'll just restore what I wrote when the vote goes live. Thanks, —Soap—09:15, 20 July 2023 (UTC)[reply]
I've also posted some comments on the talk page of that vote. I would like to keep most of the conversation there rather than further expanding this already very long thread, but I think one part of what I wrote is important enough to repeat here where more people are likely paying attention. Namely ....
Would you be willing to re-word the first of your four proposals in line with what I wrote on the talk page? I think the proposal as it stands now is setting up an unnecessary conflict based on the assumption that, even if the vote to unblock Wonderfool passes, the stewards will refuse to acknowledge that consensus has changed and we will be forced to defy their authority. I don't see any reason to assume a resistive attitude on their part, as my impression is that the stewards' job is to enforce community consensus, not overturn it. Best regards, —Soap—10:08, 20 July 2023 (UTC)[reply]
Currently, the Hokkien-Taiwanese mainly uses POJ romanisation on entries. Vowels that occured in non-Amoy/prestige Taiwanese dialects such as /ə/ or /ɯ/ do not have common attested forms in the main body of POJ works. At the moment, the conventions from the Tai-lo system are adopted, /ə/ as <er>, /ɯ/ as <ir> and /ɛ/ as <ee>. Changing toward more POJ-ic conventions may be more reasonable here since sources which discuss these vowels in the context of POJ do not use the -r type solutions. See Douglas’ dictionary which has diacritic solutions such as ö ü and ɛ for each vowel respectively in a seemingly primitive form of POJ. Compare also with sister systems PUJ (Teochew), PFS (Hakka) and BUC (Hokchew & Hinghwa) which use bottom umlauts (bottumlauts, if you will) o̤ ṳ for similar values. Also a contemporary source in 台字田 promotes this kind of system in POJ for values of such kind (although they seem to use it more phonetically since they are mostly oriented to the prestige accent of Taiwanese which doesn’t require those distinctions in POJ). Penangites have a dictionary which offers e͘ to represent ɛ which patterns if I’m not mistaken with openness in vowel quality as in o͘. In summary, ir -> ṳ, er -> o̤ and ee -> e͘. This would help bring POJ on wikt closer in line to what users may actually choose to write when it comes to these vowel qualities. @justinrleung, Mar vin kaiser, The dog2, rcAlex36Eyteo (talk) 13:52, 17 July 2023 (UTC)[reply]
Hainamese BOT uses (or used) Æ, not E͘. But E͘ seems to make better sense in the Hokkien context, and it's coming into use around Penang. 釆 (talk) 08:26, 25 July 2023 (UTC)[reply]
Do territorial changes make new definitions of countries?
The borders of France have changed a lot, but at France we cover the country via a single sense, and leave it to encyclopedias to clarify the years during which it encompassed Algeria, Guyana (a current but not e.g. medieval thing), Alsace (currently but not always), Vietnam, etc. Likewise, Poland is one sense, though its borders have changed wildly and it's ceased to exist at times. In contrast, at United Kingdom, selected territorial changes are new definitions. At United States of America, we cover the country with one sense even though it's expanded in scope/states many times and at one point split in half, but at Germany many many selected changes are new definitions. (Likewise in other languages: most countries are one sense but a few are inconsistently multiple.) Other dictionaries consistently cover single countries with single senses AFAICT — Merriam-Webster, Dictionary.com and Collins cover Germany, France, United States of America and United Kingdom with one sense apiece — and I think we should too. (Obviously, where two entities are not the same, like ancient vs modern Macedonia, we, like other dictionaries, wouldn't combine those... but note how I made the modern-country sense cover the 2023, 1990s and 1970s incarnations, rather than making new senses for every time it changed name.) What do you think, do France and US need more senses or should UK consolidate a few? - -sche(discuss)17:49, 17 July 2023 (UTC)[reply]
To be clear though the senses at Germany and UK aren't based on territorial changes, they're talking about distinct sovereign states (or at least a change in the name of the state for the UK's sense 2). That is still handled inconsistently but it doesn't look like anyone thinks we need new senses to cover a province changing hands.
Given the vagaries of political history and the fact that countries are socially constructed I'm not sure there's a one-size-fits-all policy available for these things. In these specific cases: IMO senses 1 and 2 for the UK should be combined given that it's just a change of the official title, but Great Britain should be kept as a separate sense given that it was not officially the United Kingdom (the OED also lists the 18th-century Kingdom of Great Britain and the UK as distinct subsenses of "United Kingdom"). Germany is more complicated to deal with but my inclination would be to combine 2.2, 2.3, and 2.5 (empire 1871–1918, republic and Nazi Germany 1918–45, the current federal republic) and leave the medieval kingdom and East Germany separate. —Al-Muqanna المقنع (talk) 18:08, 17 July 2023 (UTC)[reply]
Given that we try/pretend to be a historical dictionary, among other things, the extreme now-ist, snapshot definitions of countries (eg, US) covered by a given headword seems grossly insufficient. Our definition of Metropolitan France implies that our now-ist definition of France is not sufficient even on now-ist terms. The definition of United Kingdom may fall short of the coverage of Macedonia, but it is far better than the now-ist definition of United States of America.
As a dictionary, our focus has to be on how various national toponyms are and have been used. We don't need to follow every border adjustment, conquest, land purchase, or change in jurisdiction (eg, territory vs. state in US), even if attestable, but our definitions should acknowledge change explicitly. DCDuring (talk) 18:52, 17 July 2023 (UTC)[reply]
Here's something I can share. I have encountered this situation with Kinmen and Lienchiang, of Quemoy and Matsu fame. The counties now include territory that was not part of those counties before 1949. I do try to do a special documentation of the use of Kinmen with respect to Wuqiu and Lienchiang with respect to Juguang on the Citations pages. However, I have not yet seen fit to create bona fide separate senses. I think there might be room for subsenses if I got really, really technical. This same kind of topic came up before on Shanwei/Swabue at Talk:汕尾 with justinrleung: "usage is restricted to referring to the historical town rather than the modern (prefecture-level) city". My answer to the problem: wait until someone else figures it out, because Wiktionary is probably still too primitive to deal with these issues. In the mean time, I'm building the core strength of Wiktionary to prepare for a more scholarly treatment of a future editor ("He it is, who coming after me is preferred before me, whose shoe's latchet I am not worthy to unloose."). Also, I did to a split of 'Taiwan' into five different administrative division possibilities. The island and country senses aside, there were five different administrative division concepts to document, which I saw as subsenses. --Geographyinitiative (talk) 19:00, 17 July 2023 (UTC)[reply]
What do other dictionaries do? Is the OED, for instance, equally inconsistent? I think that would be a good starting point for the discussion. There doesn't seem to be an obvious answer, to me. Andrew Sheedy (talk) 20:44, 17 July 2023 (UTC)[reply]
I mentioned above the OED does treat the Kingdom of Great Britain and the formal, post-1801 UK as separate subsenses of "United Kingdom". That seems reasonable to me. The OED generally has a fairly strict no-proper-noun policy though so e.g. "Germany" and "France" don't have entries at all. —Al-Muqanna المقنع (talk) 20:55, 17 July 2023 (UTC)[reply]
No indeed, it is a lot of work, even more so to make definitions intersubjectively defensible, as editors would have to keep in mind the multi-faceted way of thinking I have outlined, and to fit the categorization is an additional hurdle. Fay Freak (talk) 20:58, 17 July 2023 (UTC)[reply]
These terms refer to nations and civilizations and only by rhetorical figure, and simplification of the juridic lay, they are used for “countries” (nation states; the distinction between states and countries works in English and French), because nations and civilizations tend to unity in one statehood for efficiency: the opposite is elaborated upon for Germany because this has blatantly not happened till 1871, and Austria is still a different country (while it is customary to speak of their leading ethnicity as German according to official Austro-Hungarian usage), whereas again even the recent so-called Germany is only from 1990. We have a similar situation as Germany before 1990 with Cyprus, where I am still pissed about the template-based simplification of my short fact sheet in the entry, that was intended to make sense of a Phoenician translation of the name of the island (since a country did not exist but multiple principalities), though Fenakhay did no wrong there. Lacking clarity of concepts consistently causes editors to be unable to define historical civilations consistently: Say Akkad and Aksum. To say nothing about our categorization.
I tried to integrate typical understandings and perspectives. France and Germany and Poland and Russia and Ethiopia are the territories typically occupied by these cultures, in particular but not necessarily, depending on the context, by mediation of a government in which they have coalesced (again “government” is not a distinct concept from “state” either, this is English-language particularism).
You see, now going to Russia is a very different thing depending on the particular subject matter: effectively, the proposition is fulfilled when you enter Crimea, in which to abide won’t much appear different from any other territory administered by the Russian government, you are not a liar if you say you have been to Russia during the holidays but only were to Crimea, and this is also recognized by Civil Procedure Law of other countries, while at the same time you won’t recognize the territorial changes, if only, but not only, because Ukraine will win. So I answer your question: Territories of countries do not constitute the entities we describe as so-called country names. Complication and hedges are unavoidable in this matter—we just haven’t put in the effort to accurately define the most everyday words, a well-known observation about the common English words as presented by Wiktionary. The territories constitute the legal constructs designated by the so-called “official names of countries”, the states and countries, according to the w:de:Drei-Elemente-Lehre by which a state needs 1. territory 2. a population 3. state power. But this is not described by a term such as Russia. By the apparent likelihoods, Russia will exists independently of some civil war removing its state power and breaking its territory into independent parts, and notably international recognition, whether at all or with particular borders, is not a part of the definition of the country Russian Federation either, only the factual constitution determining whether it is a historical term, even less so of the civilization called Russia.
There are many ways I can formulate the same thoughts but it needs to suffice. There is a whole generation confused by Wikipedia knowledge on countries. Fay Freak (talk) 20:58, 17 July 2023 (UTC)[reply]
We're also bad at truncated senses, some of which may need a lot of work to find clear quotations for. France often (probably usually in speech) means just the European part (and I'm not sure of Corsica), Denmark is ambiguous as to whether it includes Greenland and the Faeroes, and I just got a Cornishman to confirm that he had 'crossed from Cornwall to England'. The latter, though, may just be a regionally or ethnically-restricted sense of England understood in context by most Englishmen. Our current quotations for 'France' the country can all be taken as meaning 'metropolitan France' by the term France. --RichardW57m (talk) 12:00, 18 July 2023 (UTC)[reply]
Rationale: Theknightwho has repeatedly engaged in personal attacks and harassment on my talk page after having been told to stop and after having been given a block for the same behavior. He has only recently gotten the tools and has already had one unsuccesful desysop and was blocked blocked by another admin within the past year. He has shown very poor judgement in how to talk to other users, has repeatedly harassed me, and has resorted to totally unacceptable language on my talk page when he's put out. I believe that he has shown poor judgement and repeated abusive behavior that is unbecoming of an admin with insults and invectives that are aimed at agitating someone else rather than any genuine conversation, making him unfit to be an admin. Blocking him once from my talk page did not stop him, so I am hopeful that this will express that his behavior is unacceptable.
Is this behavior the kind of way we want admins treating others here (including other admins on their talk pages)? Do we think that an admin coming to a user's talk page repeatedly to engage in personal attacks and provocation while an ongoing discussion is happening here is productive? See also this thread about how the community feels admins should act in regards to repeated attacks and badgering. See also our blocking policy with relevant language about "Causing our editors distress by directly insulting them or by being continually impolite towards them... Behavior which is counter to policy, productivity or community", and "to prevent edits that will, directly or indirectly, hinder or harm the progress of the English Wiktionary". —Justin (koavf)❤T☮C☺M☯06:46, 19 July 2023 (UTC)[reply]
@Koavf I'm referring to your demand that I seek permission before engaging in any use of the admin tools, which you said in response to me saying you don't have consensus to keep militantly blocking WF in the way that you are: you said "I am officially and formally asking you to not use any admin tools at all without prior approval and consensus for every action you take, even if it is literally what our policies state." I find it extremely difficult to believe that that was a genuine request and not some sarcastic, mean-spirited attempt to draw a false-equivalence, and I don't believe any reasonable person would have interpreted you as being sincere. It's one thing to act like that, and entirely another to pretend you weren't (which is a hallmark of something you brought up yesterday). Theknightwho (talk) 07:37, 19 July 2023 (UTC)[reply]
I don't know how to make this more clear: I think you have poor judgement, I think you act in a way that an admin should not, I think you shouldn't have the tools. Therefore, as I just wrote two comments above and you evidently didn't believe or read, I do not think you should act unilaterally with any of these tools as I do not have confidence that you will use them well, understand and implement policy, or reflect positively on how someone should act as an admin. I don't know how many times I have to tell you this. —Justin (koavf)❤T☮C☺M☯07:46, 19 July 2023 (UTC)[reply]
My use of admin tools is completely irrelevant here, though, because we were discussing your use of admin tools. I haven’t used them in relation to this, and it’s not clear how this is relevant at all, which means that you seem to be saying “no, no, I wasn’t being sarcastic - I was sincerely insulting you while ignoring your point”. Seems a bit hypocritical in the context of this thread, if I’m honest. Theknightwho (talk) 08:07, 19 July 2023 (UTC)[reply]
I was not insulting you: I was documenting my concern exactly the same as you came to my page to do. You seem to not trust my judgement to enforce policy and think I uniquely have to get permission to block sockpuppets. I entirely do not trust yours. We're not going to turn this thread into endless litigation of what I wrote. I stand behind it. If you stand behind what you wrote, that's something for the community to consider. No one's soliciting your feedback of who is and is not hypocritical. —Justin (koavf)❤T☮C☺M☯08:13, 19 July 2023 (UTC)[reply]
I’m actually pretty speechless at this kind of brazen dishonesty and wilful lack of understanding as to how community consensus works. Several people have asked you to stop, and you have been rude to everyone that posted on your page (not just me). Theknightwho (talk) 08:18, 19 July 2023 (UTC)[reply]
Here you go. You continually treated everyone else in that discussion like they were idiots, which is evidently why you said you cannot take any of you people seriously here ([13]) once it had moved to my talkpage. It's fine to disagree, but it's clear that you don't think anyone who disagrees with you has a legitimate view, and seem in genuine denial that you hold the minority view. Are comments like Surjection's really not enough of a wake-up call?! Grandiose statements about how we'll have to desysop you to get you to stop are all very well ([14]), but don't strike me as the attitude of someone who understands compromise or collaboration. No doubt you'll fire back with lots of attacks on my character, but none of that changes the fact that this started because of something you chose to do, in the face of a lot of opposition. Theknightwho (talk) 16:19, 19 July 2023 (UTC)[reply]
That is not the clear consensus of the linked discussion, as I read it. There does not appear to be a black-and-white answer to the question. Andrew Sheedy (talk) 17:30, 19 July 2023 (UTC)[reply]
Why are you ignoring this comment from a global sysop, which directly contradicts it? The steward in question also states that it wouldn't be fair if that user were causing disruption on other wikis, which doesn't seem to be relevant here, so your interpretation is disingenuous at best. Theknightwho (talk) 17:53, 19 July 2023 (UTC)[reply]
And again, you answered a question with a question and made a personal allegation about my purported character and motivations. I wish I could say I was surprised. Maybe you didn't realize this, but I was soliciting feedback from stewards on the notice board for stewards. Others can certainly chime in, but my question is directed to a certain group. I'm also ignoring your pleading there because it's irrelevant. Feel free to post it, but it's not answering the question I asked, just like how your comment above this one also doesn't answer the question I asked. It would sure be neat if you would do that in the future (when questions are actually addressed to you, which this one was not). —Justin (koavf)❤T☮C☺M☯18:14, 19 July 2023 (UTC)[reply]
You don’t get a free pass to wilfully mislead people on this thread by playing the victim when you get called out for it. You knew what that discussion said, and you knew your interpretation was not how most people would read it. Theknightwho (talk) 18:17, 19 July 2023 (UTC)[reply]
Good thing that I didn't willfully mislead anyone then! Yikes! Anyway, leave me alone. Your constant hounding of me and injecting yourself into every conversation and every comment in every conversation--even those explicitly aimed at other users--is just harassment. You're not adding value, you're just making distractions. If you persist, I'll seek another interaction ban against you, which I think is completely justified from your actions on my talk page, here notwithstanding. The question is for Andrew, not you, and your specious allegation that I was misleading while citing my source and tacitly soliciting others to read it is itself willfully disingenuous. Stop hounding me with your noise. —Justin (koavf)❤T☮C☺M☯18:28, 19 July 2023 (UTC)[reply]
I thought some participants in the discussion were stewards who are not. It looks like there's only one steward who's weighed in so far. So you were correct, at least about the steward's approach. Andrew Sheedy (talk) 18:23, 19 July 2023 (UTC)[reply]
Thanks. And to be clear, in case others hadn't read it or somehow were confused, there are other participants in that conversation who are global sysops and other users who lack elevated or global permissions. —Justin (koavf)❤T☮C☺M☯18:36, 19 July 2023 (UTC)[reply]
Here are all the claims that steward Amanda wrote. Please tell me which were contradicted by a global sysop:
If a users [sic] main/1st account is still locked, they are still subject to being locked until that original account has their lock appealed. (True.)
It is not fair to other communities for a project to "allow" a new account, which then continues to be disruptive under that username for other wikis. (True.)
I will lock anyone who remains evading. The process is that they need to appeal to the stewards. (True.)
Our appeals process is hell right now. (True.)
Mostly because unsavory characters continue to be such and their appeals are meant to be their own disruption. (True?)
While I am personally working on said administrative process... it's going to take time to clear the standing ~200-300 account appeals that we do have. (True?)
It is our process until more stewards wade in to the appeals process regularly or we elect more willing to deal with it. (True)
Eagerly awaiting your response, which I am sure is aimed at being factual and not just inserting yourself into a conversation to agitate. —Justin (koavf)❤T☮C☺M☯04:37, 20 July 2023 (UTC)[reply]
@Koavf I’ve already explained it to you: nothing suggests WF is disruptive elsewhere, which makes the point about other wikis irrelevant. It is not my problem that you seem incapable of understanding points which contradict or undermine your own. Given I can only conclude you are not a rational person at this point (as evidenced by the fact you seem to be in genuine denial that your view is not supported by the admin team here), I think the only way forward is going to be for us to systematically undo your blocks whenever we encounter them. Theknightwho (talk) 11:21, 20 July 2023 (UTC)[reply]
"Please tell me which were contradicted by a global sysop". Super easy request. Not my fault if you can't substantiate your claims. Do not wheelwar by unblocking sockpuppets: please read our blocking policy. —Justin (koavf)❤T☮C☺M☯13:50, 20 July 2023 (UTC)[reply]
@Koavf I’m not interested in your wikilawyering on technicalities, but what matters here is that the steward’s justification for their point was to stop cross-wiki abuse, which doesn’t seem to be relevant here. Given that it seems the global lock was actually implemented as a mistake, too, it seems that the whole point is moot anyway.
I have read our blocking policy. Please tell me where our policies say that admins are at liberty to override consensus if they feel like it. Theknightwho (talk) 14:10, 20 July 2023 (UTC)[reply]
You made the claim, not me. It's not my fault that you wrote that. What global lock was a mistake? This user has had dozens of global locks on his accounts made. Which one are you purporting was a mistake? Our policy does not say that. There is also no consensus that users are allowed to use sockpuppets to evade bans: in fact, it's the opposite. —Justin (koavf)❤T☮C☺M☯14:15, 20 July 2023 (UTC)[reply]
It’s not my fault that you were unable to understand what I was referring to when I said it was contradicted, which is made clear by the comment which states that it is up to the local community to decide, as I have already pointed out to you.
The issue of whether it was a mistake is something that needs to be determined at Meta, but very obviously if a global lock is put by mistake on account A, and then accounts B and C are locked with the justification that they’re the same user as account A, then that would also mean accounts B and C were locked by mistake. This is not difficult to understand.
Do you actually feel like acting this way is achieving anything good? Most of us don’t want admins to act like robots with no nuance when enforcing policy. Why do you keep acting like the only person opposed to you is me? It’s very surreal how you keep ignoring context when it’s convenient for you. Theknightwho (talk) 14:25, 20 July 2023 (UTC)[reply]
I already answered the question in the comment you just responded to: the global sysop who stated it was up to the local community. I already had this impression, but you are very clearly not reading my comments properly.
I gave a list of statements. I would like to know which statement was contradicted. And speaking of reading comprehension, "please list the five admins who told me to stop blocking these sockpuppets." "So you’re just enforcing policy like a robot then", no I am not, but the prospect of some kind of semi-automated tool that helps identify sockpuppets and assists in blocking them would be very much appreciated. Are you going to help make that tool? —Justin (koavf)❤T☮C☺M☯14:43, 20 July 2023 (UTC)[reply]
There seems to be enough behaviour that hinders productivity, community, and the progress of the English Wiktionary to go around. —Al-Muqanna المقنع (talk) 08:30, 19 July 2023 (UTC)[reply]
I am really tempted to just hat/collapse this entire section because, as Megathonic pointed out at the start, it's pretty clearly not going to accomplish anything but generating heat. (The second vote was even more lopsidedly against desysopping than the first vote, and was just two months ago.) - -sche(discuss)08:36, 19 July 2023 (UTC)[reply]
I support that motion, but as an involved party here, I'll defer to someone else to collapse and close this section. Thanks for the input. —Justin (koavf)❤T☮C☺M☯15:43, 20 July 2023 (UTC)[reply]
Koavf, the past month I've seen of you has not endeared me at all to your antics. I would much sooner be voting for you to be desysopped than TKW, and this is coming from someone who has expressed tacit support for removing TKW's rights in the past. — SURJECTION/ T / C / L /12:49, 19 July 2023 (UTC)[reply]
Agreed. Both of you are productive editors. Can you spend less time fighting and more time building Wiktionary? Koavf, if you want TKW to go away, stop responding to them. They'll eventually go away. I noticed you just blocked TKW. Was that for behaviour on your own talk page? If so, consider that you may not be the best judge of a conversation in which you yourself are involved and that it might be better in such cases to call upon another admin to intervene. Andrew Sheedy (talk) 13:54, 19 July 2023 (UTC)[reply]
Ah, it seems that TKW is only blocked from your user page. That is much more reasonable than the outright block that the block message itself implies. Andrew Sheedy (talk) 15:56, 19 July 2023 (UTC)[reply]
I wanted to follow the process by discussing here first. As per the above comment, this section is not going anywhere constructive, so the topic is dead/resolved/etc. for now. —Justin (koavf)❤T☮C☺M☯15:44, 20 July 2023 (UTC)[reply]
I think Koavf and TKW you should both take some time out for a while, perhaps from heated discussion, from Wiktionary, wikis, the internet, or whatever. Too much flinging of shit recently, not enough decent lexicography. — This unsigned comment was added by 90.166.56.201 (talk).
What admin provokes three independent calls for desysopping and multiple Beer Parlour threads in under a year? As much as TKW is wont to toss out accusations of "bullying," "manipulation," "deception" etc. as a default response to criticism, I'm not connected to Justin/Koavf, @Lingo Bingo Dingo (creator of the first desysop vote), @Huhu9001 (blocked in a clear act of retaliation by TKW, as I documented in the second desyop nom), @LlywelynII (also blocked inappropriately and threatened a second time). TKW keeps earning the umbrage of other users – including long-established ones like myself and Justin/Koavf – because TKW has shown a marked lack of judgment and restraint as an admin. TKW has a history of personalizing disputes and reaching for the big stick when it's totally uncalled-for and inappropriate. The leash that TKW has been granted is already far too long. At some point a user's problematic conduct becomes more detrimental to a collaborative project than the benefits brought by their good work or special expertise. I know the collective will seems to be to let cannons freely roam the deck but at some point one must consider the integrity of the ship. WordyAndNerdy (talk) 15:30, 19 July 2023 (UTC)[reply]
You cannot seriously be arguing that Koavf is being honest by intentionally misquoting me to make it sound much worse than it was. Come on - that's textbook manipulation. That's to say nothing of the surreal attempts to claim their sarcastic comments were actually totally sincere (even when that would make them bizarre non sequiturs instead). Theknightwho (talk) 15:53, 19 July 2023 (UTC)[reply]
The interaction ban between us is still technically in force. You shouldn't be directly addressing me, and I ask that you cease from this point. But if you intend to persist as if rules don't apply to you, one hopes that your future comments contain some measure of clarity and humility. Whatever the merits of Koavf's hardline stance on WF, he's provided "receipts" for outside parties to judge. You've elected to continue making sweeping accusations ("intentionally misquoting," "textbook manipulation") without sufficiently explaining them or supporting them with evidence. You even admit as much above ("Without going into the details here..."). WordyAndNerdy (talk) 16:27, 19 July 2023 (UTC)[reply]
If there is any "interaction ban" between you two, your initial comment was already out of line. In my mind there is no doubt you were all but straight up begging TKW to respond to you. — SURJECTION/ T / C / L /16:40, 19 July 2023 (UTC)[reply]
@WordyAndNerdy Oh come on - I've tolerated you making comment after comment about me. I've given specific examples, too; I just didn't feel the need to re-explain both of them after having explained them both already in this thread:
When Koavf threatened to start a vote against me, I said Do what you like, but crybullying is unlikely to get you anywhere. ([15]) Koavf quoted this at the top of the thread as "Do what you like... crybully. Whatever your feelings towards me, that is an obvious misrepresentation of what I said - both in content and tone.
Koavf responded to my comment that there was sizeable opposition to his actions by saying I am officially and formally asking you to not use any admin tools at all without prior approval and consensus for every action you take, even if it is literally what our policies state. ([16]), which is obviously a sarcastic false-equivalence, given he's intentionally ignoring that it would be appropriate for any admin to seek permission to do something if several others advised/asked them not to. In this very thread, he claims it was simply a sincere expression of his views on my decisionmaking capabilities, which would make it a total non sequitur given that the comment he was responding to (a) wasn't directed to him, and (b) wasn't about me. Either he just decided to throw a random formal request in there (honest guv!), or he's lying.
I'm not just throwing around accusations at random, here.
I'm allowed to comment in discussions of community importance, Surjection, even if they should happen to involve TKW. Otherwise, it's creating a situation where TKW is free register their views, but I am disallowed from contributing mine on technicality. I've pointed out before how TKW has shown a habit of gaming discussions to their advantage in this way. Namely, by shouting down critics with bad-faith accusations of "harassment" and "bullying" while knowingly and, I would hazard, intentionally disregarding other user's attempts at de-escalation and drawing boundaries.[17][18]. Anyway, the interaction ban has now been violated to the point that TKW has pinged me. I'm 100% behind Koavf's call for desysopping. This has been swept under the rug long enough. WordyAndNerdy (talk)
No, you're not going to make up your own rules about what interaction bans mean, provoke someone into a response and then pretend as if they broke it completely on their own. All you're managing to do is make anyone who has legitimate criticism of TKW's behavior look disingenuous like the way you are acting now. — SURJECTION/ T / C / L /17:12, 19 July 2023 (UTC)[reply]
Oh, and the interaction ban really does matter and this particular discussion spins out of control as it so often has before, you're the only one that will be blocked here unless TKW does something particularly egregious. — SURJECTION/ T / C / L /17:12, 19 July 2023 (UTC)[reply]
I just want to point out - for the record - that the example WAN gives of me apparently overstepping a boundary is one where (a) I reminded them that I didn't want them responding to me, not the other way around, and (b) they tried to ban me from responding to the desysop vote they started against me before it started, which is hardly the same thing. And this isn't even the first time they've talked about me since the interaction ban was put in place (see [19], where they bring me up out of the blue). Can I really be blamed for seeing this as manipulative? Theknightwho (talk) 17:22, 19 July 2023 (UTC)[reply]
@Surjection - I haven't received clarity on the exact terms of the interaction ban even after asking the admin who implemented it. I've been operating as if the line is directly addressing or responding to each other's comments in the absence of any clear guideposts. And that is the line I drew with TKW in the diffs above. I didn't directly respond to TKW in this discussion. I made a general comment to the room to highlight that this is a systemic behavioural issue rather than an isolated dispute. I only responded to TKW after TKW responded to me. I didn't contribute to Wiktionary for two months partially due to TKW's threatening a retaliatory block against me in April. Further threats of arbitrary and one-sided admin actions do little to restore my shaken faith in this project. WordyAndNerdy (talk) 17:28, 19 July 2023 (UTC)[reply]
I am simply pointing out your disingenuous behavior and stating that should the interaction ban actually be effective, you would indeed be the only one that would be blocked, because that is the only reasonable course of action. Your initial message was blatantly provocative, including pinging several other users who have had a history with TKW. Your "drawing a line ... here" is arbitrary, not my actions. I'm not in the slightest involved in your personal spat. What I am interested in is your clearly provocative and hypocritical behavior which I fully have the right to criticize you about. — SURJECTION/ T / C / L /17:33, 19 July 2023 (UTC)[reply]
"Disingenuous " - how? "Provocative" - how? I'll allow that the second comment (the one addressing TKW) was mildly sarcastic in places. But no more than the comment by TKW to which it was given as a reply. I'll also grant that TKW may have forgotten about the interaction ban. Which is why I made one reply pointing it out. I don't count that first reply to me as a violation of the interaction ban. Only the subsequent reply from TKW. To anyone keeping tally, I haven't replied to TKW since the reminder, but TKW has directly replied to me. This is similar to how TKW kept posting on Koavf's talk page after Koavf requested they cease. WordyAndNerdy (talk) 17:49, 19 July 2023 (UTC)[reply]
I also find it I ironic that I, a contributor of over a decade with a clean block record, am being threatened with a block for pointing out systemic admin misconduct on a community discussion page when the thing that ignited the powderkeg here is the question whether we should automatically ban sockpuppets of a globally-blocked account (something that, from a WMF policy standpoint, is entirely uncontroversial). WordyAndNerdy (talk) 18:07, 19 July 2023 (UTC)[reply]
This is textbook emotional manipulation. "Think of poor me, a long-time contributor, who is being threatened by an evil admin." No, you are simply being forced to face the consequences of your very actions. You are not being "threatened with a block for pointing out systemic admin misconduct". I am merely pointing out that if the interaction ban that you insist would be in place, you would be the only one to be blocked because of it, and it is all due to your openly provocative message. Nobody is going to believe you if you say that you were only joking, and I similarly have a very hard time believing you do not understand how your initial message was provocative. — SURJECTION/ T / C / L /18:25, 19 July 2023 (UTC)[reply]
You've now strayed into the territory of making unsubstantiated and personlised accusations ("textbook emotional manipulation", "your openly provocative message"). You've elided addressing most of the points I've raised about TKW's past or present conduct while heaping censure upon me. It's becoming increasingly difficult to read your responses to this thread as being impartial and made in good faith. WordyAndNerdy (talk) 18:37, 19 July 2023 (UTC)[reply]
You're in no position to be arguing about good faith when you have been acting in bad faith in this thread from the get-go. But I'm done arguing here, because I'd rather do something productive to this project and not argue here just to have the last word. Everyone else can make up their minds about who's in the right here. — SURJECTION/ T / C / L /18:45, 19 July 2023 (UTC)[reply]
How have I acted in bad faith? I won't deny that I don't have a favourable opinion of TKW, given TKW's threat of retaliatory block against me, and the lack of consequences for it. But I'm not the only user who's been on the receiving end of conduct unbecoming of an admin from TKW. If a user has repeatedly gotten into similar disputes with unconnected parties, the simplest explanation is that said user's conduct is in need of examination and remedy. Like I said at the outset: name another admin who's been subject to three independent calls for desysopping in the span of five months. WordyAndNerdy (talk) 19:00, 19 July 2023 (UTC)[reply]
For context it's worth pointing out that desysop votes seem to be exceedingly rare on Wiktionary. The last one seems to have been five years ago and was on the grounds of inactivity. That TKW has generated two desysop votes and a third call for one in just five months says a lot about how much their conduct has strained the system. WordyAndNerdy (talk) 19:17, 19 July 2023 (UTC)[reply]
I'm going to say this once and only once, because I think this needs clearing up:
You began by edit-warring by reverting 6 times on the same page.
When I asked you to stop, you decided to insult me by saying one of my entries "still looks like shit because of [my] tantrum" and tried to pull rank by saying you'd been editing for 11 years (as though that somehow excuses your behaviour) ([20]).
Given that you also directly referred to the first desysop vote in that comment, I interpreted that as intentional escalation to attempt to goad me into blocking you, so I told you to stop doing that and to take a bit of time out for a day or two ([21]).
You responded by calling it a "threat", and started a desysop vote against me, and have essentially decided to wage some kind of campaign against me whenever you sense the opportunity to get me desysopped.
Hand on heart, it is some of the most vindictive, self-entitled and nasty behaviour that I have ever had the misfortune to witness, and I am glad that someone else has had the courage to stand up to you and your dreadful attitude on this thread. You really do have a brass neck on you, that's for sure.
Please feel free to carry on with your trademark indignation and threats to leave, but I really am in disbelief at how you can possibly think your behaviour has been in any way acceptable. Theknightwho (talk) 19:50, 19 July 2023 (UTC)[reply]
Oppose I gave up counting the number of posts in some of these discussion threads. I would suggest using a rule that I have adopted for myself: absolute maximum of four separate posts in any given thread of comments, with emergency fifth comment in the limit (including explanation that no more comments will be made). The "was explicitly told that if he didn't stop posting harassing messages, I would call this vote" comment happened at approximately post FIFTY. Having a four comment limit makes every comment more meaningful, both to yourself as author and to the other readers. You may say "oh, it's necessary to have in depth discussions". Perhaps. But if there's any hint of confrontation, I would say just break it off and submit any lingering issues to the community at large, and don't participate in whatever outcome there is. Then just follow whatever the decision was. --Geographyinitiative (talk) 16:11, 19 July 2023 (UTC)[reply]
I am growing increasingly annoyed at these calls for desysopping active editors, made by people directly involved in emotionally charged spats.
Wiktionary lacks a formal dispute resolution process in contrast to most WMF projects. The community has also shown a general discomfiture toward reining in problematic user conduct and addressing systemic concerns. This was bound to come to a head eventually. That it seems to have come to a head all at once may reflect how long the matter has been deferred. "When it rains, it pours." WordyAndNerdy (talk) 19:27, 19 July 2023 (UTC)[reply]
Bro, I am literally problematic af. I literally cannot hold on to my ability to edit Wikipedia because of the system over there. But yet I can make massive contributions here, and a lot of entries I make, with the requisite cites, are at the friggin TOP of a Google search, beyond even Wikipedia's grasp my bros. Lack of dispute resolution mechanism is a strengh, not a weakness. Anarcho-wiki allows words that have been ignored by Wikipedia to have life here. Some of the words that I've seen here give me PTSD/flashbacks. --Geographyinitiative (talk) 19:42, 19 July 2023 (UTC)[reply]
I'll agree that the more permissive approach taken by Wiktionary does have its benefits. I'd never be able to document fandom slang to the extent I have here under Wikipedia's policy framework. It gives me warm fuzzies whenever I see people positively discuss Wiktionary's fandom coverage outside Wiktionary. But we can't expect a collaborative project this large to function over the long-term without any sort of formal mechanism for dispute resolution. If we refuse to get our own house in order, then it becomes increasingly likely this will come as a top-down imposition from the WMF, rather than as something generated through community discussion and consensus. Finding a way to deal with problematic admins and wider systemic issues on our own would be conducive to maintaining our semi-autonomous state. Let's deal with this while it's a two-alarm fire rather than a five-alarm one. WordyAndNerdy (talk) 20:16, 19 July 2023 (UTC)[reply]
Reading through all this stuff it doesn't seem like much has come to a head to be honest other than bombastic relitigation of a few specific personal disputes so not sure the portentous threats are particularly merited. —Al-Muqanna المقنع (talk) 22:57, 19 July 2023 (UTC)[reply]
I'm pondering how to link Northern Thai forms in its two scripts, Thai and Tai Tham. Multiple spellings and encodings are quite likely for the Tai Tham script. The current preference seems to be make a Tai Tham script lemma the main lemma (possibly simply because Tai Tham feels cooler). When linking from the Thai script form, should I indiscriminately use {{alternative form of}}, or should I indiscriminately use {{alternative spelling of}}?
Extremely multiscript languages, like Pali and Sanskrit, use an entirely different method.
Or should we abandon these templates, and put the alternative form in the headword line, bearing in mind that automatic conversion is quite capable of failing - the Tai Tham script spelling tends to be phonetically more conservative. For example, the modern New Testament has Northern Thai จะ ending in a glottal stop corresponding to Northern Thai ᨧᩢ᩠ᨠ ostensibly ending in a velar stop. --RichardW57m (talk) 11:41, 19 July 2023 (UTC)[reply]
In this case it looks as though I can bridge the gap that had me worried: Thai script with droppable glottal stop Northern Thai จะ(/caʔ/) is an alternative spelling of ᨧᩡ(/caʔ/) which is an alternative form of ᨧᩢ᩠ᨠ(/cak/). The biscript 2017 New Testament uses the first and last, which is a bit odd. Time to turn 'em blue. --RichardW57 (talk) 06:58, 20 July 2023 (UTC)[reply]
To me, it seems that our bureaucrats are unwilling reluctant or hesitant to enforce our Removal for inactivity policy, which defeats the purpose of the policy and renders it more or less useless. So, I propose the following modifications to make it completely unambiguous and more enforceable:
Per Wiktionary:Votes/pl-2017-03/Desysopping for inactivity, ifIf 1) the number of admins is greater than 20, and 2) a user who has admin rights has not used admin tools for at least 5 years as per Special:Log, the admin right can be removed from the user without further ado.
Procedure: To remove the admin right from an inactive admin, create a post in the Beer parlour requesting the removal of their admin right. A bureaucrat shall review the request and automatically desysop said user if the two aforementioned criteria are satisfied: No regard is allowed to be given to any mitigating circumstances, nor to any discussion or other consideration.
A user desysopped under this policy who later wishes to regain the admin right must undergo a new confirmation vote.
This will be a binding policy and thus removed from the Notes section and inserted into a new section called "Policy for removing inactive admins".
If this is possible (I think it's not?), it would be a better solution and a vote could be created to implement it. Ideally, it could also auto-post a message on a user's talk page giving them a heads up that they're scheduled to be desysopped in 1 month if there continues to be no use of admin tools. Megathonic (talk) 19:35, 19 July 2023 (UTC)[reply]
Ah, I see. I can't imagine the crats being willing to run a bot for it. I'll give this discussion a few more days, and if no opposition to my original proposal arises, I'll set the vote up. Megathonic (talk) 16:30, 20 July 2023 (UTC)[reply]
I would support a vote if it was made. I've also been a bit baffled by the reluctance to desysop. A vote really shouldn't be necessary though. Andrew Sheedy (talk) 14:33, 20 July 2023 (UTC)[reply]
I would too although I'd really really like to hear from the bureaucrats (hint, hint) as to why they are reluctant to enforce this policy. Something ulterior must be going on. Benwing2 (talk) 19:10, 20 July 2023 (UTC)[reply]
It's concerning that in the threads about admin here, the bureaucrats haven't taken action yet even when there's policy to do so already. AG202 (talk) 19:24, 20 July 2023 (UTC)[reply]
Given the fact, as observed above, that there hasn't been much justification advanced for ignoring the existing policy I'm sceptical that fiddling with the wording is actually going to change anything (though I also wouldn't oppose it). Perhaps there simply need to be one or two more bureaucrats. —Al-Muqanna المقنع (talk) 22:15, 20 July 2023 (UTC)[reply]
Speaking of bureaucrats, User:SemperBlotto is one, but has been inactive for almost a year now, and is unlikely to ever come back from what I've heard. It seems to me this poses an even greater security threat. PUC – 22:22, 20 July 2023 (UTC)[reply]
I originally debated whether I should call for strengthening the policy, or for us to come up with a new bureaucrat candidate who will implement our policies. Decided on the former since I wasn't sure if the latter would be too escalatory, but given that we aren't exactly getting much of any feedback as to why the policy as-written isn't being enforced, I think picking a new candidate would be a good move.
I would also open to voting to de-bureaucratize SemperBlotto due to inactivity. We should have tighter constraints on that in regards to bureaucrats. Megathonic (talk) 00:05, 21 July 2023 (UTC)[reply]
I share similar concerns on the existing situation of our bureaucrats. Special:ListUsers/bureaucrat lists seven of them, but in fact only two of them (Chuck Entz and Surjection) are performing what one would reasonably expect a bureaucrat would do. SemperBlotto has been inactive for a year now, others are even barely doing things that an admin would do, only the occasional (mostly self-)deletion, undeletion and moving pages without redirects; one of them is even out of touch of the template and policy changes, and I don't think anyone of those commented there was aware that they are a bureaucrat/admin. This effectively puts us at only two bureaucrats - while they are widely trusted by the community, I don't think having only two bureaucrats is ideal for a community of this size.
It looks like the main reason it failed was that people didn't think it should be based on use of the tools, which makes sense, because there are legitimate reasons why one might be active but not use the tools in several years (something as simple as one or two bureaucrats responding more quickly to requests than others or there not being many checkuser requests). I'm fairly confident that a vote based on inactivity would passed. Andrew Sheedy (talk) 13:54, 21 July 2023 (UTC)[reply]
That vote was also 6 years ago. Several of the participants in that vote are no longer active here, new users have joined, and consensus could have definitely changed. I think it's worth exploring again and agree it has a good chance of passing today if based on inactivity. The only question is, how long should that time limit be? Megathonic (talk) 16:17, 21 July 2023 (UTC)[reply]
2 years didn't pass last time, but maybe 3 years would. Perhaps the vote could be created with two options, one for the status quo (5 years) and another for 3 years. While we're add it, it wouldn't hurt to add something about inactive bureaucrats. Andrew Sheedy (talk) 22:46, 20 July 2023 (UTC)[reply]
Would it make more sense for the vote to have three options (3, 4, and 5 years), or would it be better to stick with 3 and 5?
For a bureaucrat, we could have a separate vote for that. Possible idea: After no account activity for 6 months or 1 year (shorter timeframe due to higher risk; how long should this be?), or less than, say, 50 edits in one year, anyone could post in the BP that so-and-so is eligible for de-bureaucratizing due to inactivity. Then a second person, either an admin or another bureaucrat, could confirm the eligibility, petition the stewards to remove their crat status with a link to our policy and vote results, and then confirm in the BP thread that they submitted the request. Megathonic (talk) 00:01, 21 July 2023 (UTC)[reply]
I completely agree we should remove the bureaucrat status from the 5 inactive bureaucrats; this is definitely a security risk. I also agree we need a new bureaucrat, but who would it be? I would suggest User:DCDuring but I'm not sure they would accept a nomination. Benwing2 (talk) 07:48, 21 July 2023 (UTC)[reply]
Instead of having multiple options where you have to pick one, could we instead be allowed to vote for or against any of the proposed time lengths? e.g. someone might vote "yes" to 3 or 4 years, but "no" to 2, and this way the option with the most votes is the one that we know the most people are okay with. It avoids a situation where a really unpopular option wins simply because the opposing votes were split between multiple alternatives that everyone's kinda okay with. Theknightwho (talk) 08:58, 21 July 2023 (UTC)[reply]
Yes, I was thinking of setting it up where everyone could cast a separate vote under each time length and making that explicitly clear in the voting instructions. Though, instead of using most votes to determine the winner, I was thinking of going with the one that has the highest percentage in favor, since that one would have the strongest consensus. If there's a tie on percentage, then the one with the highest vote total wins. Should the tiebreaker somehow also be tied, then the proposal with the higher time length wins, since the longer time length is the status quo. Megathonic (talk) 16:02, 21 July 2023 (UTC)[reply]
@Theknightwho, Megathonic: Could we write something to the effect of "A vote in favour of an option (say, desysopping/debureaucrating after three years of inactivity) will automatically be considered a vote in favour of all less stringent options (desysopping/debureaucrating after 4 or 5 years of inactivity); the first most stringent option to achieve consensus wins"? PUC – 20:06, 21 July 2023 (UTC)[reply]
That's a really good idea. Yes, we could do that. I'll give this thread some more time to gather additional input, and once it settles down, I'll create one vote for improving the desysop policy and a second vote for crat/checkuser rights. Megathonic (talk) 20:12, 21 July 2023 (UTC)[reply]
The only drawback is it would make the vote tabulation more complicated. Example: A vote for 4 years would automatically be a vote against 3 years, but it wouldn't show up under the "3 year" vote itself. But I think it's doable. Megathonic (talk) 21:03, 21 July 2023 (UTC)[reply]
Given that we're thinking of having a vote to make Benwing a bureaucrat, I think this vote will be unnecessary (and perhaps a bit inflammatory). So I would have a vote to add a bureaucrat and leave this specific issue be. Andrew Sheedy (talk) 23:44, 21 July 2023 (UTC)[reply]
User:DCDuring's plea for more civil discussions has fallen on deaf ears so I'll repeat it. Please reduce the heat; this is only a single online dictionary, it's not a major country on the brink of war or anything. Wiktionary isn't perfect but it isn't terrible either, and we seem to be losing track of the end goal, which is to improve the dictionary. If we all spent more time doing that and less time arguing, the site would be better for it. Benwing2 (talk) 19:08, 19 July 2023 (UTC)[reply]
Well...it is as simple as this: when there is a fever, there is a symptom, as well as an underlying cause. Of cause antipyretics are useful, but to what extent.
That aside, in this situation instead of "pleaing" or "calling for" something it is better to set rules for something. Similar problems will just keep popping up from time to time, especially given that some admins would even believe in things like "a good little 'f*ck off' is nothing to be frowned at" and are free to carry this out. -- Huhu9001 (talk) 01:06, 20 July 2023 (UTC)[reply]
Do you have evidence that announcing rules is more effective than encouraging good behavior, all things considered? Among the things to be considered are the cost of enforcement, both in terms of time spent and change in culture, eg, more wiki-lawyering. I, for one, can here because of WP's culture and stayed because of enwikt's. DCDuring (talk) 01:19, 20 July 2023 (UTC)[reply]
I tend to agree, all the wikilawyering at Wikipedia seems very time-consuming and a big sink in emotional energy, and dealing with the constant back-and-forth reverting happening in many articles must be exhausting. Granted, the reverting is often related to real-life political conflicts and there's inherently less of that in a dictionary, but even so I feel we do a better job of avoiding edit warring. Benwing2 (talk) 01:26, 20 July 2023 (UTC)[reply]
I do like the more anarchist governing style that we have here. At the same time though, there does need to be some order among the chaos. I don't think we need a new rule so much as admins more actively intervening when things devolve into derogatory, personal insults or a pointless slugfest between 2 people. If such intervention attempts don't work, then a 24-hour block could be used to force a time-out. For a pattern of this behavior over a period of time, the length of the block could be increased as needed. Megathonic (talk) 17:00, 20 July 2023 (UTC)[reply]
Desysopping only applies to sysops, so it doesn't address the problem, which may be caused by anyone. In addition, abuse of sysop tools is not part of the problem we have been having.
Time-limited blocks are better, possibly even page-specific blocks, eg. BP, TR, ID, User talk pages, etc. Perhaps even very-short-time blocks (one hour or a few hours) applied to those pages. This would be a modest punishment, but, more importantly, a cooling-off period. Determining whether one, both, or many sides of a discussion should be blocked in this way requires great wisdom, probably not to be found in a vote on a rule. DCDuring (talk) 17:19, 20 July 2023 (UTC)[reply]
Oh, gotcha. No, not a yellow card. Intervening in the thread to say behavior XYZ is getting out of line, and perhaps adding a suggestion for how to tone it down or resolve it. This suggestion would depend on the circumstances surrounding the thread, or maybe that wouldn't be appropriate based on said circumstances. If it continues, then link to our policy on disruptive behavior and that this will result in a block if it continues. Something along those lines, tailored to the situation at the admin's discretion of how to best handle it.
I thought a yellow card came with an explanation of the (apparent) violation. But you seem to be suggesting something less formal, with no explicit sanction for continued violation. That fits with a need for flexibility, but lacks teeth. DCDuring (talk) 21:41, 20 July 2023 (UTC)[reply]
I mean, I'm actually okay with that too, but you're right that it wasn't my suggestion. The main thing is I'd just like to see our admins more proactively intervene before it spirals out of control. Megathonic (talk) 21:46, 20 July 2023 (UTC)[reply]
Merely encouraging good behavior is indeed more pleasing and less tiresome if it works. But as you can notice and pointed out by BW, people began to turn deaf ears to it. Something that does not work is not effective in any sense. In this case rules at least will work. -- Huhu9001 (talk) 02:11, 20 July 2023 (UTC)[reply]
Although it's always been around, it does indeed seem like recently there's been an increase in emotionally-charged and blameful discussions, especially around discussions involving political or hot-button topics and involving users unfamiliar with Wiktionary's policies. We ought to extend some gratitude towards the mods and all other users who maintain civility when dealing with these discussions.
I think there should be some sort of generic "definition template" for particular terms, of course not for all, but for example in Slavic languages you have certain morphological derivations that create a regular, predictable lexical change, as can be seen with things like {{dim of}} or {{femeq}}. Polish and Russian, among others, use verbal prefixes to create things such as the semelfactive or the iterative, etc. We have specific templates for some of these, but not for all, and it would be a lot easier to have a single template that could be updated with potential parameters that could be used for various parts of speech and that could even categorize. This would also reduce the chance for errors in definitions. Of course such a template would not be applied to "root" words, and also not the same across languages, i.e. a "perfective of" might make sense in one language, but perhaps not the Slavic languages, etc.
@Surjection @Jack who built the house @Theknightwho who expressed interest on discord and @Benwing2, as I have heard that he also has interest in such a thing. As for "reducing complexity", it should be obvious that reducing the number of templates is generally a good thing. Vininn126 (talk) 07:05, 20 July 2023 (UTC)[reply]
@Vininn126 I think that {{infl of}} is what you're looking for, given the language-specific tag support I recently added to it. However, I'm not completely sure because I don't quite understand what you just wrote; can you give some examples? Benwing2 (talk) 07:11, 20 July 2023 (UTC)[reply]
Can you expand on your example? I'm still not quite getting what you're proposing. Maybe if you wrote up a specific proposal as to what you think the template would look like, it would be clearer to me. Benwing2 (talk) 07:21, 20 July 2023 (UTC)[reply]
@Benwing2 In essence you have certain word-formation that creates a regular and predictable definition, i.e. diminutives. As for the verbs, po- creates the definition (very often) to X for some time; to X one after the other, where X is the root verb. This is its own lemma, it's a lexical derivation, so {{infl of}} wouldn't work. na- creates "to X a lot of". Russian has similar prefixes, as do other languages.
In theory I could just create templates for these but at this point we have so many definition templates that point to another lemma, such as {{dim of}} that I think we should have one template where one of the parameters determines the definition given, so you can have "semelfactive of" or "diminutive of" etc. Vininn126 (talk) 07:31, 20 July 2023 (UTC)[reply]
@RichardW57m It's not an inflection, as I explained, it's a lemma. The difference is that inflection of would imply that the term is a non-lemma, i.e. a morphological ending added to all verbs, whereas these are not added to all verbs and the derived terms are lemmas. Vininn126 (talk) 12:36, 20 July 2023 (UTC)[reply]
Could the solution simply be that we eventually replace {{infl of}} with {{form of}}? The only issue here seems to be that people are reluctant to use {{infl of}} for non-inflections, but the name is essentially arbitrary and could be anything. Theknightwho (talk) 12:48, 20 July 2023 (UTC)[reply]
I suppose in theory, the idea is a general-purpose definition template that acts as what I call a "soft" redirect, i.e. pointing to a lemma with more information. Vininn126 (talk) 12:51, 20 July 2023 (UTC)[reply]
So why not create {{posid of}} as a hard redirect to {{inflection of}}? 'posid' means 'possibly opaque semantics-inheriting derivative'; just as with paper, we need to maximise the information per byte but we wouldn't want people to twig that meaning inhered in a prefix like 'na-'. --RichardW57m (talk) 13:11, 20 July 2023 (UTC)[reply]
The other issue with {{infl of}} is its parameter order, which makes sense for inflections where there can be multiple forms, but not for this case where it's enough to have one form per template/line, so the same parameter order as {{form of}} uses would be preferable. — SURJECTION/ T / C / L /13:23, 20 July 2023 (UTC)[reply]
I also think a big part of the confusion is that in some languages semelfactive is a conjugated form, as opposed to a derived, lexical semelfactive. Vininn126 (talk) 13:24, 20 July 2023 (UTC)[reply]
One part of this seems to hinge on the difference between derivational and inflectional morphology. In English, we wouldn't consider words with un- or anti- to be part of the paradigm for the unprefixed word, but we do for -er and -est. In some languages there are things like motion toward vs. away, personally witnessed vs. heard about, alienable vs. inalienable possession ("my hand" or "my father" vs. "my cup"), definite vs. indefinite, animacy, cause to do versus do, etc. that are a regular part of the morphology. There are word groups like food, feed and eat that would be all part of the same paradigm in other languages, with terms for each. The boundaries between derived, inflected and related words are different in different languages, so it's hard to come up with something truly language-agnostic. In some cases, one language might address this in the etymology, while in another it would be in the definition line. Chuck Entz (talk) 14:33, 20 July 2023 (UTC)[reply]
This is exactly right - in some languages some of these things would be considered inflectional morphology and in others derivational. Well put, Chuck. Vininn126 (talk) 14:37, 20 July 2023 (UTC)[reply]
@Vininn126 As others have pointed out, {{infl of}} has all the functionality for this already, and there's no specific reason it needs to be limited to non-lemma forms. If we want a different template, we could use {{form of}} and recognize inflection tags in the argument following the language code, and even allow multiple tags separated by a comma, colon, semicolon or whatever. Or we could create something like {{deriv of}} or {{der of}} for this purpose (derivational morphology vs. inflectional morphology). Benwing2 (talk) 19:08, 20 July 2023 (UTC)[reply]
Yes, using {{form of}} with its order with the tag(s) coming before the lemma should satisfy all these concerns. We could create a separate tag namespace for derivational morphology but I'm not sure it's necessary. Benwing2 (talk) 19:15, 20 July 2023 (UTC)[reply]
Don't you then hit at least the occasional problem with parsing a list of strings as input? Or should we expect to invoke it with a final blank parameter? With the current order, I have enough trouble remembering to separate the lemma and the tags with a blank parameter. A final blank parameter will be rather alien to our invocation habits. Or will alternative displays simply require an explicit |alt=? --RichardW57m (talk) 08:46, 25 July 2023 (UTC)[reply]
Basically, I don't see how you can have multiple tags short of a solution you might not like.
Not a real example, but to what would you convert what would currently be written {{inflection of|pi|pacati|pakati|causative}}?
Would it be to {{form of|pi|causative|pacati|pakati}}? If so, what would you convert {{inflection of|pi|pacati||causative}} to? To {{form of|pi|causative|pacati|}}? But won't that be interpreted by a module in the same way as {{form of|pi|causative|pacati}}?
So, if we regarded the 'double causative' as the 'passive of the causative' (or is it the other way round?), which is what it is often translated to, what would we convert causativepassive of pacati to? To {{form of|pi|causative|passive|pacati}}? But does that link to 'pacati' with two tags, or does it link to 'passive', display 'pacati' and have the single tag 'causative'.
A solution is to require the alternative text to be specified by a named parameter. Then the first unnamed parameter is the language, the last unnamed parameter is the link target, and the unnamed parameters in between are the tags. There's a good reason for lists of unnamed parameters to come at the end. --RichardW57 (talk) 19:17, 27 July 2023 (UTC)[reply]
┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ My intention was to use a comma-separated list of tags in the {{form of}} tag param, so you would write {{form of|pi|causative,passive|pacati|pakati}}. To avoid problems with embedded commas, commas are only treated as separators if not followed by a space or a less-than sign (which is generated when the {{,}} template is used). A different chracter could be used in place of a comma, such as a colon (:) or circumflex (^). Ampersand (&) might be logical but would cause issues with HTML char references. Benwing2 (talk) 22:22, 27 July 2023 (UTC)[reply]
In an attempt to resolve the issue of translinguality, I like to hammer out an agreement before we submit the individual pages to arbitration. My proposed definition is that the following should be true:
"To obtain information about term X in the context of language Y, consider what exists in the entries for term X in languages Y and 'translingual'. In the case of contradiction, what is said in the entry for language Y takes precedence over what is said under translingual. If information derived thus is demonstrably wrong, then Wiktionary is giving false information."
Ideally, the information for the translingual term X should indicate where it may be overridden by the information for specific languages.
I do notice that under this definition, 'translingual' does not imply widely used. Amongst the information that might be listed for a translingual letter is its origin; we might not wish to repeat that for every language that uses the letter. --RichardW57m (talk) 12:18, 20 July 2023 (UTC)[reply]
Agreement is needed for letters and the like (e.g. ligatures and at least some symbols); I had hoped to find a universally applicable definition. --RichardW57m (talk) 11:59, 21 July 2023 (UTC)[reply]
I do not use to think that way, whatever you were thinking. A major point here is that you basing your conceptualization of the translingual on letters, which are themselves special cases, even if we restrict or purview to one language only a kind of abstractions serving as building blocks for lexemes.
There is no rule that a general definition cannot be overwritten by specific information, this contravenes the rule lex specialis derogat legi generali: what I describe as particular may well not be the translingual situation. If such particularities are often in some context then I can hedge the translingual definition as only “regularly” or a “principle”, and likewise I can repeat on individual languages if I assess a need to ensure reader understanding. The case of contradiction you speak about I cannot comprehend. Fay Freak (talk) 01:00, 21 July 2023 (UTC)[reply]
@Fay Freak: I think the word 'thus' turned out to insufficiently clear. Are you happy if I replace 'thus' 'by this rule for resolving contradictions'. I don't think the legal rule you quoted is universally accepted, and I can certainly see it not being applied by readers, so it is worth restating.
This issue came up because of letters, so we should probably have a specific consensus for them, as well as for similar things (letter-diacritic combinations, digraphs, digits, symbols with Unicode code points, etc. Maybe best to leave emojis to their own guideline.)
With diacritics, there's a SoP issue. For example, r̥ is the international standard for transliteration of Sanskrit ऋ. That passes SoP. But then someone added a second definition of (IPA) voiceless alveolar trill. I deleted that because it failed SoP, but the editor complained, so I restored it. Do we want such things? We could add a hundred IPA letters with a voiceless ring, but personally I don't see the point. On the other hand, if such a combination exists already, as here, people might feel it's incomplete if it doesn't include SoP definitions. And then there are Unicode characters that fail SoP. If we exclude letter-diacritic combinations that fail SoP, should we make an exception if it's been assigned a Unicode character? Or is Unicode irrelevant to the content of a dictionary?
There's also the problem of adding language-specific information to the translingual section. For example, at ऋ the translingual pronunciation is given as "/ɻ̩/, /r/". This is simply wrong: there is no particular translingual pronunciation, and even if one could argue that there's some sort of translingual ideal, the ideal pronunciation of a Devanagari letter would not be English /ɻ̩/! The virgules also tell the reader that we've made some sort of translingual phonemic analysis, which is nonsense. (The other pronunciation, /r/, isn't even syllabic though we define the value of the letter as syllabic.) For another example, Richard has insisted in making Burmese the translingual standard for letters of the Mon-Burmese script. Why not choose Mon? (Richard says that the Burmese Army decides the issue.) I would argue that, while transliteration is good in such cases, we should not provide pronunciations under the translingual header unless they're actually translingual, such as letters of a phonetic alphabet, where the pronunciation is not language-dependent. IMO we should also define them as letters of a translingual script, and not of the alphabet of a specific language, or as having a particular sorting order within that script unless the sorting order is defined translingually. kwami (talk) 08:52, 26 July 2023 (UTC)[reply]
@Kwamikagami: Burmese letter forms have wider currency than Mon letter forms amongst languages using the script, though Mon forms nowadays mostly follow the Burmese forms, more so than the convergence of Shan forms with Burmese. Also, I made the point that the Burmese alphabet was more widely known than other Burmese script alphabets. This is because the Burmese army defeated the other armies of what is now Burma. (I'm not sure how significant their holding off the Chinese invasions was.)
Unicode has now made itself the underpinning of much. For example, a word can't have a page unless it can be expressed in Unicode. There is or was a policy along those lines, but I currently can't find it. Now ask yourself why we include letters at all.
Kindly take the discussion of translingual pronunciation to its appropriate topic on this subpage, where you will note that I suggested that narrow phonetic transcription had no place.
It's quite conceivable that you don't care about the utility of Wiktionary at all. If you do, consider that someone who knows that r̥ is a Sanskrit vowel might wonder what it was doing in the transcription of some South American language. It is not an easy SoP to look up part-by-part for the median user. Indeed, how would you look the combining mark up? The methods I use are somewhat arcane. It might make sense to have a policy that notations for phonemes, as opposed to phones, should be included. Of course, as Wiktionary is not paper, we might be able to accommodate phones as well. And perhaps it's already included in "Characters used in ideographic or phonetic writing such as 字 or ʃ." --RichardW57m (talk) 10:27, 26 July 2023 (UTC)[reply]
I must indeed not care about the utility of WT if I ask for clarification of our SoP policy.
I'm of two minds about things like this, as I expressed above, which is why I didn't argue about restoring the IPA usage, but it should be a matter for consensus.
The Devanagari example above was not a "narrow phonetic transcription", so it wouldn't matter if we excluded those. But this does play into the meaning of translingual.
Reg. your argument that Burmese takes priority, English takes priority with the Latin script, so all our Latin translingual sections should discuss English usage, and we can delete the English sections as redundant. Then where, say, the French or German entry contradicts English usage, that would override the translingual section per your proposal here. kwami (talk) 04:52, 27 July 2023 (UTC)[reply]
The notation /ɻ̩/ is indeed narrow when there is no contrast of rhotics to make. /r/ is what we should use for a broad transcription, unless there is a contrast that needs to be made. And Indian Indic writing systems have generally used Sanskrit as a reference point, unlike mainland SE Asia where spelling seems to have been generally conservative, so that words tend to continue to be written with the same letters.
The 'ISO basic Latin alphabet' used in defining the more widespread translingual Latin script letters is merely a glorified expression for the current (American) English alphabet, though to be fair I don't remember seeing anything different in English primary school classrooms. (Britons seem to care more about accents than Americans, though British newspapers can be pretty brutal with foreign accents.) This was not particularly achieved by the US Army, but rather by anglophone dominance in computing.
As for phonetic information, let us look at the usage notes of Translingual d(letter):
The letter d is used in the alphabets of many languages, and in several romanization systems of non-Latin scripts to represent the voiced alveolar or dental plosive (/d/). In some languages and transcription systems, d may also represent other sounds, such as /t/ or /ð/.
The difference from what I suggested is that Wiktionary then fails to say what the letter represents in any language lacking an entry for the letter. It merely gives one a best guess, even assuming there's a comma missing before 'to represent'.
Whether the English pronunciation of consonants should be the default is an interesting question. Counting languages equally, it might well be true, though it wouldn't work well for the European continent. I think standard practice has been to duck the issue of a default by saying nothing. For the case of Translingual သ(letter), while it is indeed part of the Burmese alphabet, its standard Burmese pronunciation appears to be an outlier, and the 'default' pronunciation should be /s/, not /θ/ (or /ð/). --RichardW57m (talk) 09:52, 27 July 2023 (UTC)[reply]
Agreed that they should be consistent and agreed that they should say gerund as well. Maybe include some option to not have gerund forms if there is somehow some word that is only a verb -ing... —Justin (koavf)❤T☮C☺M☯02:37, 21 July 2023 (UTC)[reply]
The way to have such an option is just to list the verb as a present participle rather than present participle + gerund; but I don't think any such verbs exist. In Old English, gerunds in -ing were clearly differentiated from present participles in -end, but by Late Middle English they had merged into either -ing or -in depending on dialect. Benwing2 (talk) 02:49, 21 July 2023 (UTC)[reply]
Speaking from a position of general ignorance, aren’t gerunds of verbs treated as nouns? If so it seems odd that we should say “and gerund of” under the verb heading. Also, what should one do if the gerund form has some sense other than “the act of doing [the verb]”—do we then continue using {{present participle of}} in a verb section with a separate noun section? — Sgconlaw (talk) 04:43, 21 July 2023 (UTC)[reply]
The gerund usage of verbs is like "His spewing of profanities was problematic" or "His speaking French would be useful". AFAIK all verbs can be used that way and the meaning is always the same. Benwing2 (talk) 04:49, 21 July 2023 (UTC)[reply]
Gerunds are sometimes countable (as in seeing where you have seeings), and I assume our verb templates wouldn't be able to handle that. It also happens fairly often that a gerund is attested but not the verb. Ioaxxere (talk) 14:17, 21 July 2023 (UTC)[reply]
"Seeings" is the plural of the noun "seeing", which would not be affected by the proposal since nouns of this type are treated as their own lemmas. The proposal as I understand it is only about how to describe the non-lemma form that appears under the heading "verb".--Urszag (talk) 14:49, 21 July 2023 (UTC)[reply]
The terminology is not particularly fixed or free of confusion, but what I learned was to reserve the term "gerund" for an inflected verb form ending in -ing that heads a phrase that is used in the same context as a noun phrase (e.g. as the subject or object of a verb or preposition) but that is differentiated from a noun in several respects, such as a) it does not take a determiner as a rule (although it can, exceptionally, take a genitive subject), b) it can take a direct object and c) it can take adverbial modifiers. That is, "spewing" is a gerund in "I object to his frequently spewing profanities", but not in "I object to his frequent spewing of profanities" or "I object to the frequent spewing of profanities". An -ing word used as a noun is not a gerund according to what I learned, just a type of derived noun that has the same form as one. Likewise, a gerund is not a noun; it is a verb (just like a participle is not an adjective). Any verb that has a present participle also has a gerund, and the Cambridge Grammar of the English Language considers the distinction misguided and refers to a unified category "gerund-participle". I don't think we should distinguish between "present participle of" and "present participle and gerund of", and I favor the latter wording.--Urszag (talk) 06:16, 21 July 2023 (UTC)[reply]
Support the "present participle and gerund of" form per Urszag, no reason to cling to the distinction when others have moved on. @Ioaxxere, Vininn126: The distinction between "gerunds" and "present participles" in English is basically arbitrary and has been discarded in recent grammars, as Urszag mentions (and Wikipedia also notes). —Al-Muqanna المقنع (talk) 13:59, 21 July 2023 (UTC)[reply]
Can you give some examples of affected entries? Translations should certainly not be deleted. I don't think anyone's advocating for that. What senses would need to be deleted? Andrew Sheedy (talk) 14:44, 21 July 2023 (UTC)[reply]
Verbal nouns are not the same thing as gerunds (e.g. to illustrate with modification by adverbs: "the pillaging of the city was atrocious" vs. "pillaging the city was atrocious", but "the savagely pillaging of the city was atrocious" ✗, "savagely pillaging the city was atrocious" ✓). —Al-Muqanna المقنع (talk) 15:08, 21 July 2023 (UTC)[reply]
@Al-Muqanna According to Wikipedia, gerund is sometimes but not always used as a synonym of a verbal noun. If we make it clear (in the glossary etc.) that we draw a distinction, then I don't have any opposition to the proposal. Ioaxxere (talk) 15:42, 21 July 2023 (UTC)[reply]
I have edited the glossary to add a link to Wikipedia and a description of how the verb form called the gerund differs in terms of grammatical behavior from a verbal noun in -ing.--Urszag (talk) 19:14, 22 July 2023 (UTC)[reply]
Should we similarly add some language that a noun in attributive or interjective use behaves differently than a noun as subject of a verb or object of a verb or preposition?
But seriously, I hope you didn't reinforce the antique notion that 'gerund' and 'present participle' are distinct PoSes (though identical in all appearance) rather than being labels for the different grammatical roles that -ing-forms take. DCDuring (talk) 21:03, 22 July 2023 (UTC)[reply]
Other OneLook dictionaries don't bother with any entry for, say, seeing as a verb; they do have seeing#Conjunction and seeing#Noun (astronomy). Some have another noun sense "sightedness" or an adjective sense "having vision" (though this might be viewed as attributive use of the 'sightedness' noun sense).
Personally I think chasing down the presence or absence of usage of an -ing-form of each English verb in a plural or uncountable sense would be a great waste of human contributor time. I see no value to a noun section for an -ing-form containing only a definition such as "the action of the verb {{...))". (We have 131 of these and another 151 with the words "the action of", all in English lemmas with titles containing [a-z]+ing. This would be an overcount.) As I see it, the main useful functions of our English verb form entries being primarily to direct language learners to the lemmas and secondarily to eliminate redlinks. Cases where there is some actual semantic unpredictability of course warrant noun, adjective, or other PoS sections. DCDuring (talk) 19:11, 21 July 2023 (UTC)[reply]
@DCDuring Yup, I agree that spending manual effort on each of the many English verbs like this is not worth it. My proposal is rather to use a bot to harmonize existing -ing-forms that are currently defined using either present participle of VERB or present participle and gerund of VERB to use consistent wording. Benwing2 (talk) 00:14, 22 July 2023 (UTC)[reply]
I would go further, albeit without the possibility of automation,. to eliminate the "action of the […]" definitions and the Noun PoS sections that contain only such definitions. The 280-odd instances of such definitions seem even worse to me than the non-standard, but templated wording, not that standardization seems likely to hurt.
I would opt for something that made explicit the fact that 'present participles' and 'gerunds' both always use the same form and that the -ing-form can function as an adjective and as a noun with all the various meanings of the verbs they are forms of. So we do not have to, essentially, reword the 45 definitions of set#Verb in adjective and noun sections, because the verb definition wordings transform predictably into the adjective/participle and noun/gerund wordings. All of this grammatical stuff should be available in Appendix:Glossary and on hovering over the words present participle and gerund. Both present participle and gerund are relics of grammar as formerly taught to and imperfectly learned by most normal English-speaking users of enwikt. DCDuring (talk) 01:45, 22 July 2023 (UTC)[reply]
It's standard for us to list derived words separately in English, regardless of how productive or predictable the derivation is. Compare derivatives in in un-, -y, -able, -ability. Since derived nouns ending in -ing are listed as their own words, not as forms of the base verb, Benwing2's proposal isn't relevant to our policy on words like seeing (n.) (plural seeings). If you think those nouns should be excluded, that would be a separate proposal, but I would oppose that (and it sounds like Ioaxxere would also).--Urszag (talk) 01:41, 22 July 2023 (UTC)[reply]
But all verbs have an inflected form ending in 'ing' that can serve as a noun modifier and as a noun, with perfectly predictable meanings assumed by the words in each case. We have only 'form of' definitions for plurals of nouns, comparatives and superlatives of adjectives, and the other inflected forms of verbs. Affixation is distinct from inflection, even in English.
That some -ing-forms bave derived homonyms with distinguishable meanings is indisputable. We should not let normal users be confused by separate vacuous definitions like "participial adjective of VERB", "verbal noun of VERB", or "action of the verb VERB" and still less by definitions based on some subset of the definitions of VERB. DCDuring (talk) 02:08, 22 July 2023 (UTC)[reply]
@DCDuring I'm a bit confused here. Are you suggesting that if an English word ending in -ing, originating as the present participle/gerund of a verb, acquires additional senses as a noun or adjective lemma, we should list the term *only* as a noun or adjective lemma and delete the separate POS entry stating that it's the present participle/gerund of a verb? That is contrary to standard practice around here and IMO would be confusing, because it would lose the connection between the word and the verb it originates from. Benwing2 (talk) 02:15, 22 July 2023 (UTC)[reply]
I concur with this. I have no idea why this particular instance would be confusing. To delare "We have only 'form of' definitions for plurals of nouns, comparatives and superlatives of adjectives, and the other inflected forms of verbs" only raises the question "and why not this inflection too?". Theknightwho (talk) 13:08, 22 July 2023 (UTC)[reply]
I was responding to @Urszag, who was arguing that -ing-forms are like affix-derived words rather than inflections.
I hope taking a word like seeing for example helps. In that entry I would delete the "action of see" sense and include "gerund" in the 'form of' definition that starts the entry. I would also make it a point to start all -ing-form entries with a Verb PoS section containing the 'form of VERB' definition, as seeing does. DCDuring (talk) 18:23, 22 July 2023 (UTC)[reply]
I doubt you'll get a consensus for removing those kinds of entries. No matter how predictable the meaning of such nouns is, they are nouns, not verb forms. For comparison, it might be difficult or even misguided to make entries for every noun in -ness that can be formed from an adjective and define them all as "the quality of being X", but that doesn't mean that nouns in -ness are just adjective forms. They're nouns derived by a highly productive suffix with a highly predictable meaning.--Urszag (talk) 19:03, 22 July 2023 (UTC)[reply]
@DCDuring I agree with putting the form-of present participle and gerund definition first; that is what I normally do with e.g. past participles and other inflected forms that have subsequently acquired separate lemma senses. I'm not sure whether it makes sense to delete the "action of see" definition; I get your point about it being predictable but I think once it's acquired a separate sense, it might be helpful to have the predictable definition as well to emphasize the possible usage. What I'm not so sure about is e.g. the separation of Etymology 1 and Etymology 2 under meeting, and having definition 1 of Etymology 1 just gloss the gerund. I would have a single Etymology section, where it can be noted that the word stems from the merger of two Middle Engish words, and underneath Etymology are two headers, a Verb form-of section and a Noun lemma section. `Benwing2 (talk) — Preceding undated comment added 20:01, 22 July 2023 (UTC)[reply]
┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ OK, this discussion seems to have reached its end and AFAICT there is no opposition to me changing English ing-entries to read present participle and gerund instead of just present participle. I am not proposing to add or delete any entries, just change the wording of existing entries. If no further objections I will do this in a day or two. Benwing2 (talk) 05:56, 27 July 2023 (UTC)[reply]
Benwing is my top pick and I really think they should be a bureaucrat, if they're willing to take that role on. The second person I thought of is -sche. I'm also open to considering other candidates. Megathonic (talk) 15:54, 21 July 2023 (UTC)[reply]
can non-admins vote? I would definitely vote for Ben, he helps out a lot with making templates and stuff for smaller languages on wiktionary. سَمِیر | sameer (talk) 18:53, 21 July 2023 (UTC)[reply]
Honestly, it seems like there's broad support for Benwing, so if he accepts the nomination (pinged him in an above thread), I'll go ahead and make the vote. AG202 (talk) 19:27, 21 July 2023 (UTC)[reply]
The skills required of a bureaucrat do not include technical ones. They seem to me to be knowledge of the rules under which we operate and of the mores of this community, interpersonal skills, sound judgment, including ability to know which way the wind is blowing. DCDuring (talk) 23:05, 21 July 2023 (UTC)[reply]
I didn't mean I wanted him as a bureaucrat solely because of his technical skills but because he's constantly been trying to help other editors on Wiktionary. As a novice I feel very comfortable tagging Ben and asking him for help with things (I've never spoken to the current bureaucrat's because I'm scared of them haha), just look at how many people ask him for technical help regularly on his talk page. Also in all the drama this month he mainly was trying to deescalate the tension. Yes he primarily works on the technical part of the site, but even if he mainly works in lua and not entries there's no denying that he's a good community member in all aspects and constantly trying to help other editors.
Now admittedly I'm not sure if helping people is part of the bureaucrats job but I definitely trust that he has the interests of the wiktionary in mind. No one is gonna spend hours writing code for something they don't care about. He definitely has community and interpersonal skills and as for judgment, I personally have no doubt that he would choose what's best for the wiktionary. سَمِیر | sameer (talk) 04:36, 22 July 2023 (UTC)[reply]
Benwing's interactions probably provide good evidence of relevant personal characteristics. I also suppose it is implicit that we expect a bureaucrat be familiar with the issues that unite or divide us. This make me wonder about whether bureaucrats should loose their bureaucratic powers when they have not been active for some time. I think active participation in community discussions, rather than mere (and unverifiable) lurking, within the last one, two, or three years should be a requirement. DCDuring (talk) 20:55, 22 July 2023 (UTC)[reply]
There are only two sets of rights which I think it is important to limit (in number) within the project, bureaucrat and checkuser. Each of these require a level of discretion and conservatism that is less important in other roles (though some may argue that there has been too little discretion among admins lately). We currently have two quite active bureaucrats, which is sufficient to get the job done. I also think it is OK, even preferable, that the bureaucrats use their tools sparingly and deliberately, erring on the side of acting slowly in non-emergency situations. While I think there are a number of admins who would make fine candidates, I don't think we need more at this time. - TheDaveRoss13:36, 21 July 2023 (UTC)[reply]
It is extremely rare that the things which get rev-deleted (which replaced oversight) are sensitive, doxxing or similar. Mostly they are spam or vulgarity, neither of which really need to be rev-deleted in my view. If we actually used the function primarily or frequently to remove what should be non-public data, I would add it to my list. (We also don't have any suppressors, if we had them that would likely be added as well.) - TheDaveRoss14:38, 21 July 2023 (UTC)[reply]
The fact that they're not implementing our explicit policy on desyopping admin, and then generally not responding when asked why, unfortunately. Even when pinged directly by people like Benwing. They've been asked to several times as well. AG202 (talk) 18:51, 21 July 2023 (UTC)[reply]
Correct. Surjection finally stepped in and did it today, but there seemed to be resistance to doing so at least on Chuck's part and no explanation for why action was not being taken ("I'm busy and I'll get around to it next week" would have worked fine, but all we got was silence and a seeming unwillingness to comply). I'm not at all convinced we need more bureaucrats, but I would support having more active ones in addition to Chuck (who is a great bureaucrat, this odd case aside). Andrew Sheedy (talk) 23:17, 21 July 2023 (UTC)[reply]
Notwithstanding all the discussion, AFAICT there are no problems that have resulted from inactive admins. Account hijacking is a potential problem, if not an actual one, for all accounts with powers sufficiently above those of registered users, not just admin accounts. IPs give us more heartburn than inactive sysops. This non-issue seems to be in use as a stalking horse for other issues that people have. DCDuring (talk) 23:47, 21 July 2023 (UTC)[reply]
Sure, but that's an argument against the initial vote on this. But the fact of the matter is that that vote passed and so the real non-issue should have been complying with the vote. It isn't so much that it's really, really necessary to get rid of inactive admins as it is that we as a community voted that this is what we were going to do and that bureaucrats should do it "without further ado". It's the fact that there was "ado" about it that was the issue, not specifically the existence of inactive admin accounts. Andrew Sheedy (talk) 23:55, 21 July 2023 (UTC)[reply]
I expect those with extra powers to use some discretion in exercising them. A responsible 'crat does have to confirm that the admins haven't in fact used their admin powers, which does take a few minutes per admin. The risk of actual harm in delay frankly seems much less than that. Trying to follow the heated BP and other discussions was probably more important than this matter. DCDuring (talk) 00:03, 22 July 2023 (UTC)[reply]
I'm finding it quite difficult to imagine what "stalking horses" would be in play for removing someone who's been gone for 5 years. That's a very generous time span and far less strict than Wikipedia's policy. Megathonic (talk) 23:58, 21 July 2023 (UTC)[reply]
I agree that the number of bureaucrats and checkusers should be limited, but I think four active bureaucrats would be better than two for a project of this size. In any case, we should have bureaucrats who are willing to enforce a policy to desysop for inactivity. Given that no bureaucrats have weighed in on the proposal to strengthen the policy and clarify why they aren't implementing the existing one (such as, is it not worded clearly enough, or something else?), I'm thinking that we have zero bureaucrats who—outside of an emergency like an account getting hacked—are willing to desysop someone unless we put it to a vote each and every time. That is grounds to add someone of a different mindset.
My ideal situation:
Pass a policy to de-bureaucratize automatically after 1 year of inactivity and use this to remove @SemperBlotto if they don't return by then.
Hold separate votes to remove @Paul G, @Hippietrail, @EncycloPetey, and @Ruakh because times have changed and we view them as no longer contributing actively enough to the bureaucrat role, despite still being active users. While we wouldn't (and in my view, shouldn't) remove admins for this reason, stricter criteria should apply for bureaucrats.
Huh? No. In a proposal to remove someone from a bureaucrat role, the wording would be first person plural (our consensus is blah blah) or third person singular (the community's consensus...). It's the same reason I wrote "We will petition..." in the vote to unblock WF. Megathonic (talk) 19:10, 21 July 2023 (UTC)[reply]
I genuinely don't understand your point. That "we" is referring to the collective we, which would be the case if this vote were created and we passed it based on the view that we (as a community) view them as no longer contributing actively enough to the bureaucrat role. Megathonic (talk) 19:33, 21 July 2023 (UTC)[reply]
It is easy to read all of your uses of we as pretending that a consensus exists, which pretense often serves to influence the outcome of votes. I know that there is something invigorating in making calls to action ("Call for a new bureaucrat"), but consider that you may be wrong. DCDuring (talk) 23:00, 21 July 2023 (UTC)[reply]
I don't think that's a very charitable interpretation, if I'm honest. I initially read it in the way Megathonic is saying they meant it. Theknightwho (talk) 13:03, 22 July 2023 (UTC)[reply]
I appreciate my name being thrown into the mix, but I am already a 'crat on Wikiquote and Wikisource. Adding another project would seem like an exercise in collecting hats. I think Benwing and -sche would be fine choices. bd2412T16:36, 21 July 2023 (UTC)[reply]
I am fine with remaining or leaving, as the community decides. If I am needed for the numbers, I will stay, but it is true that I am no longer especially active here. --EncycloPetey (talk) 17:56, 21 July 2023 (UTC)[reply]
Hi, before accepting I'd like to find out what the expected time commitment is for a bureaucrat. I know User:Chuck Entz does a lot of behind-the-scenes work looking for and blocking new user accounts that are trying to add spam to the site or otherwise be abusive, and User:Surjection puts in the time patrolling recent changes, maintaining abuse filters etc. There may be other things that the two of them do that I'm not aware of. I'm willing to take on the role of helping to arbitrate disputes but I have my hands pretty full with the technical side of this site and I'm concerned taking on the bureaucrat role would cut into this. Any comments from our current bureaucrats or from User:BD2412 or anyone else who is a bureaucrat for another site? Benwing2 (talk) 20:02, 21 July 2023 (UTC)[reply]
@Benwing2: Most of what I do behind the scenes has nothing to do with being a bureaucrat: the only thing specific to bureaucrats is setting and removing privileges. Some of the stuff I do is related to my checkuser role, and some of it is just something to do while I'm deciding what else to work on- not even really necessary half the time. The main consequence of being a bureaucrat comes from being the only person who can enforce sanctions on admins- a lot of conflict resolution and mediation. Chuck Entz (talk) 20:15, 21 July 2023 (UTC)[reply]
Let me add that a bureaucrat is mostly defined by what they don't do. Basically, the bureaucrat tool is our equivalent of the nuclear option: if a bureaucrat is doing much, then something is seriously wrong. Don't judge by the nastiness of the past couple of weeks. I can only think of a handful of times in the entire history of Wiktionary when a bureaucrat has stepped in to desysop someone without a vote telling them to. The main thing we do is working to defuse situations so we never have to use the tool. Diplomacy and restraint are critical. I probably should have done more in recent weeks, but my reluctance to tell other admins what to do has stood me in good stead the rest of the time. I was the fourth of five siblings growing up, so I learned a lot about dealing with people who don't have to listen to me. Chuck Entz (talk) 21:04, 21 July 2023 (UTC)[reply]
As someone who bears some of the responsibility for that, I don't want you to feel reluctant to step in where it's warranted. Regardless of the issues at hand and who's "right" (whatever that may mean), it's always better to have the perspective of someone who isn't personally invested, as they're more likely to have a level-headed perspective on things. Theknightwho (talk) 21:39, 21 July 2023 (UTC)[reply]
Don't forget that Chuck is also a checkuser; some of his work might be related to that hat more than to bureaucrat one.
Also, I don't see why you should feel obliged to assume more responsibilities than you're willing to endorse: you can volunteer for taking care of some aspects of the bureaucrat role and not the others. That's how I've always understood adminship, anyway: I'm using the tools to block vandals and delete vandalism, but I'm mostly staying away from other responsibilities that the role may theoretically entail.
Personally, I'm looking for someone who would be willing to curate the list of administrators at regular intervals. I don't really care about the rest. Others may disagree about this, though; it could be helpful for them to state what they would expect from a new bureaucrat. PUC – 20:24, 21 July 2023 (UTC)[reply]
Normally when a vote for a new admin or authorizing a bot passes, only two bureaucrats are pinged, presumably because they are the only responsive ones. It would also be nice to have the option of pinging a third person. The only extra work for that would be a few clicks to add/remove the right. I wouldn't expect anything more than responding to such requests to add/remove rights, and if it's denied (like in the inactive admins thread), to state why. Megathonic (talk) 20:35, 21 July 2023 (UTC)[reply]
It depends on the project, but on Wikisource all admins must be reconfirmed annually, and bureaucrats close those discussions at the end of each month. In the rare event that someone's rights are removed, we file a request at Meta to remove permissions, since we do not have a large enough community to have local desysop controls. bd2412T20:45, 21 July 2023 (UTC)[reply]
Now that this nomination is in process, this will bring the total number of active bureaucrats to 3 if (when) it passes. What are people's thoughts regarding whether we should leave that number at 3, or whether we should get an additional nomination underway to bring it up to 4? Megathonic (talk) 02:42, 23 July 2023 (UTC)[reply]
I say leave it at least until we've removed the bureaucrat status of the inactive bureaucrats. There's no rush at all and I rather suspect three will be enough anyway. Andrew Sheedy (talk) 03:22, 23 July 2023 (UTC)[reply]
Feel free to remove me, as long as we have enough bureaucrats who are active when needed. I don't actively follow votes or discussions anymore, and while I'm happy to implement sysoppings/etc. when pinged, in practice what happens is that someone pings a bunch of bureaucrats at once and I'm never the first to see it. The only recent-ish time that I remember using bureaucrat privileges is when the "interface administrator" group was introduced and I needed to add myself to it so I could edit some sitewide CSS (for the {{trans-mid}} deprecation); but there's no particular reason that I needed to be the one to do that. —RuakhTALK23:53, 22 July 2023 (UTC)[reply]
Temporarily Duplicate L2s
@Benwing2, JeffDoozan: As a result of @Kwamikagami's improper conversions of translingual sections to monolingual sections, we are being left with a lot of duplicate L2's, particularly Thai and Thai, but possibly also Burmese and Burmese. Please leave them until the moratorium on editing such entries ()see above) has been lifted and we can undo his damage (or conceivably tidy up the minor wounds from his surgery if we accept it). It's probably not malicious vandalism, but is an example of the damage that can be caused by misconsidered enthusiasm.
The damage the merger does is that the resulting language section then gets organised by etymology, with changes to the heading levels. When we undo that merger, it is highly likely that the section levels will not be restored correctly, which can then have bad results when AutoDooz (or whatever JeffDoozan's bot is called) starts sorting misordered sections. --RichardW57 (talk) 21:47, 21 July 2023 (UTC)[reply]
Yes, we still have issues to work out. I haven't seen User:Kwamikagami participating so much in these discussions lately. Also, Kwami, I do see some changes to single-letter entries in the last few days, including some like [22] where you are removing L2's without any explanation. Maybe this is warranted but you really need to include changelog messages when making changes that might be contested or controversial. Benwing2 (talk) 00:20, 22 July 2023 (UTC)[reply]
I'm afraid I starting dealing with orange links from ca(“and”) before I realised what I was doing; I've deleted the previously unentered one character forms, with the result that Wiktionary continues to look broken rather than incomplete in that respect to those not logged in. --RichardW57 (talk) 14:52, 22 July 2023 (UTC)[reply]
I think @RichardW57 has already reverted my cleanup of the affected duplicate L2s, but just to be sure, here's the list of duplicate L2s as of July 20: User:JeffDoozan/lists/duplicate_l2. You can check previous revisions of that page to see prior duplicates in case this issue started before July 1. As far as the bot goes, it can fix section levels for multiple etymologies [23] and for a single etymology [24]. If you've seen it make an edit that worries you, please let me know. Or, if you can give me a list of the pages you want it to exclude (or just "every single letter page") I'm happy to oblige. I haven't followed the discussion mentioned and I'm not confident I know what's going on so please be very specific about any action that may be needed on my part. JeffDoozan (talk) 13:08, 22 July 2023 (UTC)[reply]
That's odd; I thought I saw more. One was Translingual ำ/Thai ำ, which as you say I've dealt with. For ဝဒတိ(vadati), you guessed wrong - I hadn't finished the editors after copying and pasting. Likewise dealt with, but should now have no duplicated L2's. --RichardW57 (talk) 14:45, 22 July 2023 (UTC)[reply]
That looks more complicated than the simple downstep of Standard Japanese, and I don't see if he describes whether the pitch of joshi is phonetically or lexically determined, which is part of the downstep analysis of Standard. I would think we have a choice between marking all 'high' syllables with an acute accent, or using the IPA upstep and downstep symbols as equivalents to pitch-marking in the kana transcription. Both strike me as somewhat awkward. kwami (talk) 04:55, 23 July 2023 (UTC)[reply]
Kwami block
I blocked User:Kwamikagami for 3 days for continuing to change Translingual entries to language-specific entries and secondarily continuing to edit Translingual and single-character entries in general, after I specifically warned him several times not to do any of these things while this issue was being discussed and in contention. I notice also that Kwami has not been participating in the discussions that User:RichardW57 has been trying to initiate in the BP on these issues, after I have asked him several times to do so. Furthermore, in response to my reverts, Kwami accuses *ME* of introducing inaccurate information (basically, whatever was there prior to Kwami's changes, and whatever reappeared after the reversion), and generally adopts a hostile attitude towards me and pretty much everyone else who interacts with him. At this point, I do not see this user as being a productive contributor to Wiktionary, but rather someone who has strong opinions about this matter and will edit war and ignore warnings to get his way. He has been blocked before for this sort of behavior. I considered a week-long block but I felt that might be unwarranted given that the previous blocks were awhile ago. But if he returns and continues this behavior, IMO 1 week is a minimum for the next block. He will probably request an unblock, but I would recommend against it given his continued pattern of behavior over several weeks. Benwing2 (talk) 07:42, 24 July 2023 (UTC)[reply]
In my experience the user has a tendency to push what they want while not being willing to discuss things others they think they should change, essentially ignoring consensus and forging their own idea of how things should be. Vininn126 (talk) 08:00, 24 July 2023 (UTC)[reply]
There is also a multi-character user-perceived character that he could sensibly raise an RfM on as a pipe-cleaning exercise, but I'd rather get the evidence in place before the request is raised. --RichardW57m (talk) 09:36, 24 July 2023 (UTC)[reply]
Thank you, because I had told them to stop and look for consensus as well, and they only continued. Now we have to clean up their mess. Ex: this change which completely goes against what we do with Yoruba entries. AG202 (talk) 15:22, 24 July 2023 (UTC)[reply]
I have noticed that quite a lot of the changes feel rushed. While I generally agree that it's annoying to have low-value Translingual entries that are blatantly based on Unicode character names, I think we need to mass revert these deletions and have a proper discussion about them.
From my perspective, the Translingual entries are a useful place to put the English name of the letter (if there is one). This change means that ѵ no longer points to izhitsa, which is very silly, and also doesn't make it clear that it was part of both the Old and modern Cyrillic alphabets (even though it became obsolete in the first half of the 20th c.). Simply deleting the entry is just not the right approach. Theknightwho (talk) 15:31, 24 July 2023 (UTC)[reply]
I agree that we need to revert them. There was a discussion, but Kwami blatantly disregarded what folks said there. They should've been blocked earlier imho, it was clear to me that they didn't care what was said. Now we have to parse through months of changes. AG202 (talk) 15:50, 24 July 2023 (UTC)[reply]
I've just been made aware of entries like ẹ́, ọ́, ɛ́, & ɔ́ that were made without any consultation of Yoruba editors and blatantly go against WT:AYO at the very least. AG202 (talk) 15:54, 24 July 2023 (UTC)[reply]
@AG202 Yeah I should probably have blocked him earlier. I had hoped that by telling him unequivocally that he needed to participate in discussions rather than just make changes he would get the hint, but apparently not. I am also familiar with him from his work on languages in Wikipedia, which probably biased me towards a benefit-of-the-doubt approach. Benwing2 (talk) 21:04, 24 July 2023 (UTC)[reply]
As I said, my error. Actually, I initially thought that was a different entry, where the only definition in the lower-case was a Mandarin word, and the capital was a case variant of that word, something we do not do on WT. (I seem to remember now that it was defined as 'Mandarin' in upper case as well.) Anyway, when I discovered my error I should've reverted myself and I didn't. That's why I said I wouldn't contest the block. kwami (talk) 21:56, 24 July 2023 (UTC)[reply]
I'd like to request a longer block if possible, or at least until the entries have been cleaned up. I'm very very frustrated that this issue was brought up many times before, and it took this long to get a block. Now, it's infiltrated the entries that the Yoruba community is directly responsible for, our templates, and messing up how we do things. We're already a smaller editing community and now there's so much that I'll have to go through and revert/update/nominate for deletion. This is why I truly wish that we'd be more proactive with action when editors are clearly going around changing entries without consensus. A block is not a death sentence, and we should think about the exponential damage that can happen if we don't act faster. Some of these changes are also for smaller languages so who knows if they'll ever get fixed. This is how we end up with problem entries for years to come. AG202 (talk) 22:20, 24 July 2023 (UTC)[reply]
@AG202 I agree in general we need to be more proactive. The most problematic damage comes not so much from vandals (where the damage is obvious and usually caught quickly) but from people who think they know what they're doing and don't, but are persistent. The issue is that it's difficult for admins not directly involved in an area to track everything going on, and knowing whether and when to block requires both finesse and significant looking into the edits of the users involved in the area to see what happened (which in turn requires a time investment). Maybe a better way of dealing with this is to have involved editors who see something problematic happening to request a block from an admin. In this case I understand you weren't looking at the single-letter entries so didn't realize what had happened. How much damage is there and do you still want a longer block, or is it enough if Kwami makes a list of all affected entries, reverts his changes and tags the newly added pages with {{delete}}? Benwing2 (talk) 01:52, 25 July 2023 (UTC)[reply]
The reasons for the block have nothing to do with what I've done with Yoruba! The block was for deleting material, not adding it. If you were frustrated by my Yoruba edits, you could have have said something on my talk page and I could've taken care of it myself. Someone shouldn't be blocked because you can't be bothered to communicate with them. It's not like I even made many Yoruba entries. (I can even list them for you if you like.)
Personally I think the Yoruba guideline should be consistent with what you want for Yoruba articles: either you want separate entries for letter + tone or you don't. But if the consensus is that you don't, and that's not at odds with broader WT consensus, I have no problem reverting myself or (where necessary) nominating my creations for deletion. kwami (talk) 22:29, 24 July 2023 (UTC)[reply]
@Kwamikagami Yes and no; I blocked you most immediately for changing Translingual entries to non-Translingual entries but secondarily for continuing to make changes related to single-letter entries (which includes adding new ones) after I asked you (several times) not to. The second issue is actually the more problematic one because people will now have to sort through all your changes to see which ones need reverting (and may resort to mass-reverting, see below). Benwing2 (talk) 01:55, 25 July 2023 (UTC)[reply]
I hadn't realized there was any problem with single-language entries or adding information. That wasn't at all clear to me. kwami (talk) 05:49, 25 July 2023 (UTC)[reply]
I asked for a moratorium on changes to single-letter entries. I suppose it's a judgment call whether that applies to adding new single-letter entries but it definitely feels like a "spirit of the law" interpretation would include them. Adding new entries without discussion after being repeatedly told that you should not change entries without discussion and consensus feels like you're deliberately trying to make an end run around the process and establish "facts on the ground" to support your viewpoint. I think part of the problem is you're used to a "do what you can get away with" attitude at Wikipedia, which seems to work there but definitely doesn't work here. If you want to be an accepted and productive member of this community, it looks like you need to radically change your M.O., seek consensus whenever there is a dispute, and reach out to existing editor communities rather than expecting them to reach out to you. You still don't seem to "get it": Wiktionary has a much smaller editor base than Wikipedia and does not have a formal arbitration process (we simply don't have the resources for this), so we depend on consensus-building and collaboration. You've been running roughshod over these norms, and your responses to User:AG202 show (to me at least) that you haven't fundamentally changed your attitude even after being blocked. Benwing2 (talk) 06:06, 25 July 2023 (UTC)[reply]
My understanding was that you objected to turning, say, a 'translingual' entry on Burmese into a Burmese entry on Burmese because that deleted the translingual entry, not because it created a Burmese entry. The broad objections to my edits centered on deleting material, not creating it. kwami (talk) 10:00, 25 July 2023 (UTC)[reply]
@Benwing2, Kwamikagami: You suffer from a lack of scruples, as shown by that quotation from Harrington for Mon ဣ(adverb). (In that case, there is a way of showing that what is punctuated as a verbatim quote, which Chuck Entz eventually used on the quotation.) Remember, when I automatically started fixing orange links for the Pali word ca(“and”), I reverted my edits because I realised I was in breach of the moratorium. You also implicitly objected to my filling in missing definitions on the basis that I might make errors. --RichardW57 (talk) 22:57, 25 July 2023 (UTC)[reply]
It definitely is related, you don't reach out to editing communities when making major changes like that. And honestly, I hadn't seen the entries until today because I'm not constantly checking, but this conversation made me see the changes that have been made. I'm also almost certain that I've talked to you about translingual-like changes before anyways. To be frank as well, even if I did see them ahead of time, I'm unsure how much leaving a comment on your talk page would've changed things considering that others have told you to stop making changes multiple times, yet you continued it anyways. WT:AYO was internally consistent before your changes to it, as understood by the Yoruba editing community + my additional comments below. AG202 (talk) 22:36, 24 July 2023 (UTC)[reply]
Check the article history: I never made a change to AYO.
Not being sure if someone would respond to your comments is not reason for a block. And these are not "translingual-like" entries, they're Yoruba entries. You're bringing up an entirely new topic. kwami (talk) 22:51, 24 July 2023 (UTC)[reply]
Changes to the letter list which is listed on WT:AYO, sorry for being unclear there. But regardless, I brought up the changes to other languages as well with Tlapanec, which in turn is related to the Translingual changes. The Translingual changes aren’t in isolation, they’re one part of the recent wave of changes that you’ve been making without contacting the right people or looking for consensus. You may have been blocked for this specific issue but it’s hardly the only piece of this puzzle. AG202 (talk) 01:59, 25 July 2023 (UTC)[reply]
WT:AYO is contradicts itself on that point, not that I was aware of that guideline. AYO itself lists vowels with tone marking as "letters", and links to all of them. If they're listed atomically there, it would seem appropriate to create entries at those links. If I had been aware of the guideline, I would've concluded that it referred to words and not to letters. Some require a colon to link to, but many do not, and anyway that's not an uncommon unintended consequence on WT. If there should not be entries for Yoruba vowel or nasal letter + tone, then IMO they should be removed from the alphabet list, or at least delinked. There's also the SoP issue being discussed below that would affect this, though the current practice has been that it's appropriate to create entries for them. IMO non-African languages that have pre-composed letters in Unicode shouldn't be privileged in that regard -- that after all is a common complaint among Africans trying to write their languages in e-media. kwami (talk) 21:45, 24 July 2023 (UTC)[reply]
@Kwamikagami: I'm not sure who added the letters like "ê", considering that they aren't used in contemporary Yoruba (actually wait you added them 🤦), but if you actually click on Ẹ́, for example, you'll notice that it doesn't lead to Ẹ́, but instead Ẹ, per our guidelines on stripping tone diacritics from links. The same thing applies to the letter name template. Prior to your changes to it, it worked as intended, and I'll be reverting you now. The reason why characters like Ẹ́ are listed are because that's how the letter list template works for other languages as well. Diacritics (or letters) that are used but aren't a part of the letter list are put in parentheses. See: {{list:Latin_script_letters/es}} or {{list:Latin_script_letters/pl}}. Regardless, you should have absolutely pinged Yoruba editors before making these changes to see what the practice is. If WT:AYO truly was contradicting itself, it's on you to reach out to the community to get clarification. CC: @Benwing2 for more context. AG202 (talk) 22:10, 24 July 2023 (UTC)[reply]
Some of the letter + tone diacritics redirected. Others did not. In fact, a great many did not.
Yes, I added some missing letters to the alphabet list. Since they're no longer used in contemporary Yoruba, they should of course be taken out of the alphabet list, but they are found in Yoruba sources, so just like dated English usage they should be addressed somewhere.
BTW, I'm curious why you would remove all tone marking from the Beninese orthography but leave it in for Nigerian. Is Beninese Yoruba really written without tone marking? kwami (talk) 22:57, 24 July 2023 (UTC)[reply]
As of the current (reverted) revision, all the letters direct to where they should be. We ((Notifying Oníhùmọ̀, Oniwe, Egbingíga): have talked about whether or not to move Yoruba to having tone in the entry title, but have not yet reached a consensus due to the amount of work it'd take to move them all (or opting to redirect toned entries to toneless ones). Until that has been decided, Yoruba entries should not have tone in the entry title at all, per our guidelines at WT:AYO. This also applies to letter entries. As for the Benin side, that change should stay, and I'll add it back later, but honestly, it's not easy to find standard Yoruba text in the Benin orthography to begin with. AG202 (talk) 23:07, 24 July 2023 (UTC)[reply]
There are also speedied page deletions of mostly low content pages that should be restored (this is beyond my rights). If someone can nag me with short lists, say no more than 64 at a time, I'm happy to see what information I can pull out from the Unicode records for non-CJK characters, such as they are, to explain the characters. I don't feel up to rescuing 'turtles'. (The records are bad for early Unicode, but by Version 5.0 they're fairly good for additions to blocks. Whole scripts have, I think, remained poor.) I suggest we change {{delete}} to {{rfd}} if non-trivial content was taken out. We may need to extend the one month deadline on responding to {{rfd}} and {{rfv}} for what look like potentially accented single characters. --RichardW57m (talk) 17:09, 24 July 2023 (UTC)[reply]
@RichardW57 Do you have any idea of how many pages were speedied? I reverted the {{delete}} tag added to all the pages not deleted around July 12, but maybe some got deleted before or after then. Benwing2 (talk) 21:00, 24 July 2023 (UTC)[reply]
Speaking of mass-reverting, it may be time to write the revert script that I've been meaning to write for awhile (e.g. to revert translation entries added by Rajkiandris). The tricky thing about this is that you need to revert the changes made by a particular editor even if there are other changes on top of this, so essentially you need to pull out the diff(s) of the change(s) and apply them in reverse onto the latest version. Can someone help me identify the pages that need reverting? Either a list of pages, or maybe a time range for changes made by Kwami along with some filter to restrict only to the correct pages (e.g. those with one or two Unicode chars, or something)? Benwing2 (talk) 21:08, 24 July 2023 (UTC)[reply]
@RichardW57 Per Richard's suggestion, I changed User:Kwamikagami's block to only apply to content namespaces (Main, Appendix, Reconstruction) + Category, Template and Module. This should allow him to participate in discussions if he wants to. Benwing2 (talk) 21:15, 24 July 2023 (UTC)[reply]
@RichardW57 You can get a list of all deleted pages like this: [25] You can filter by date, which should help. (Unfortunately it doesn't look like you can restrict to the mainspace.) Benwing2 (talk) 02:11, 25 July 2023 (UTC)[reply]
Since the end of May, only about 60. I need to do some processing, as some are combining characters that are difficult to unstick from the previous characters. Some may be double counted as some were deleted, recreated, and deleted again. The italic partial differential operator seems to have been very popular, perhaps because the Unicode name captures its functions so well. --RichardW57 (talk) 03:24, 25 July 2023 (UTC)[reply]
19-23. ểỄễỆệ (Recreated, but are we in breach of copyright because of duplication of old content? History and thus attribution now starts at @Kwamkagami's edits.}}
Woah woah woah. ѵ still pointed to ижица (ižica) and had entries for both Old Cyrillic and Cyrillic, so I don't see how someone could possibly not know that was the case. I did check that. If someone wants a translingual section, fine, but it should be something other than blindly copying the Unicode name as the definition, when we know that may very well be wrong. Though admittedly this entry was better than most; it was really just a matter of that section being redundant. kwami (talk) 21:39, 24 July 2023 (UTC)[reply]
This really feels like a case for {{rfc}}. I think we need a better way of getting people to look at translingual request lists. Or is it just me whose monitoring is remiss? RichardW57m (talk) 09:02, 25 July 2023 (UTC)[reply]
@Benwing2 Kawami put requests for verification on multiple single letter translingual entries. Can I remove them, or do we still need to go through the verification process? Would we even need to verify the existence of letters encoded in Unicode?? سَمِیر | sameer (talk) 19:13, 25 July 2023 (UTC)[reply]
Many of those were changed from deletion requests, per Benwing's instruction, or added after he said to use rfv rather than rfd tags.
As for whether they need to be verified, yes they do, for several reasons:
Unicode names are sometimes simply wrong. Sometimes there's a Unicode "alias" to correct that (Unicode names cannot be changed), but not always. Even if not actually wrong, often they're poor summaries and not adequate as definitions. And of course most are descriptions and not definitions. If our definition is simply the Unicode name, that needs to at least be confirmed, and more likely needs to be reworded.
Some Unicode characters are not used in any language. They only exist for internal reasons, not for any actual need. Calling such letters "translingual" is misleading.
Some are only used in a single language and so are also not translingual.
Ugh. Kwami, talking to you is like banging my head against a brick wall. User:Sameerhameedy, which single letter entries got tagged with {{rfv}}? The proper process, which Kwami seems not to have followed, is to post at WT:RFVN listing all the challenged pages and giving reasons why they are challenged. When there are several of them they need to be in a single post. However, the reasons being given by Kwami are the same reasons he used to try to delete the pages in the first place. This discussion needs to happen at WT:BP and it needs to be a *DISCUSSION*, not Kwami simply repeating his assertions and refusing to engage in discussion. Kwami, I see no evidence that you're heeding my request for discussion, so I will extend the block come tomorrow unless you start doing so. Benwing2 (talk) 05:47, 26 July 2023 (UTC)[reply]
@Benwing2 Here are the entries I found that she marked as rfv
ᵣ (subscript r) , Ꞁ (turned L) , ꞁ (turned l) , ꝿ (turned insular g), Ꝿ (turned ins G), ƿ (runic wyn), Ʈ (T w retro).... Actually I can't list them all but while searching I found Category:Requests for verification in Translingual entries which has 90 entries. I don't think we can find all the changes he made here, since some might not be marked as translingual (unless TL is the default??) but a good portion are here. (not all of these are because of Kwami but a lot of them are).
On another note I found a couple entries that, after people had fixed them, Kwami had removed the definition and moved it to "description". An example can be seen here (ń), but there's a bunch of pages like this and I'd have to go through his contributions to find them all which will take a while. I don't have time to sit down and go though that many edits but I can sort though some tomorrow. سَمِیر | sameer (talk) 07:23, 26 July 2023 (UTC)[reply]
A description is not a definition. The definition of "apple" is not "a word spelled a-p-p-l-e". That's all that those spurious definitions were, so I moved them to 'description' and tagged them as needing a definition. Which they do. kwami (talk) 07:30, 26 July 2023 (UTC)[reply]
Your example is very disingenuous. Before you changed it the letters definition was "the letter n (link to the definition of n) with acute accent (link to definition of acute accent)". Yes it was a "discription" but it very much was ALSO a definition, as the individual parts of the letter are described elsewhere and linked. What would you prefer? To create a definition for every letter - diacritic combination? That would be redundant and almost impossible to maintain. Creating definitions for "base" forms and linking to those is far more practical. Especially since definitions would be repeated over and over otherwise. Also if you've ever actually looked at the definition of "apple" it is very much also a description. سَمِیر | sameer (talk) 07:48, 26 July 2023 (UTC)[reply]
I would prefer an actual definition: what the letter is, what it is used for, etc. It might be used for a certain value mathematically, or in some international standard. It might be a convention in transliteration. We could mention the languages that use it in their orthographies -- something of actual substance, rather than just parroting the Unicode name. kwami (talk) 07:56, 26 July 2023 (UTC)[reply]
If a reader looks up N with a grave accent in WT and discovers that it is "N with a grave accent", how is that of any value at all? kwami (talk) 07:58, 26 July 2023 (UTC)[reply]
Benwing, you said to change my rfd's to rfv's, or you would revert them, and you reverted most of them, but I did change some of them to rfv's. Someone even asked you if we should really have so many rfv's, and you said yes, that was the proper way to handle things. Now it's no longer the proper way to handle things. I wish you would tell me up front what you expect.
If I'd known you wanted a centralized discussion, that would have saved me a lot of time. The main reason I didn't convert most of my rfd to rfv before you reverted them was that other things were going on that had a claim on my time. With a centralized discussion, I could have finished before your deadline.
And what are you talking about re. not discussing things? I have been participating in the discussions. Is there a specific thread you have in mind? Because I am here at the Beer Parlour, as you say I should be. Again it would be helpful if you would tell me up front what you expect. Though just now you seem to be saying that I'm not supposed to present my reasons for challenging these entries; I'm confused how I'm supposed to argue my case without stating it. kwami (talk) 07:27, 26 July 2023 (UTC)[reply]
@Kwamikagami The larger issue here is you have been making unilateral changes based on certain opinions you have about how things should be that you assert as facts. What I'm looking for is this, in a nutshell: (1) Don't wikilawyer. (2) Participate in BP discussions so we can come to a resolution on under what circumstances single-char Unicode entries should be Translingual. Even better, open such a discussion, make a proposal with reasons, and be open and receptive when others challenge you. I see none of this, however. I don't see a single discussion you started, nor even very many discussions you've taken part in at all. For example, there's a discussion below here "NFC v. SOP" that Richard started that you haven't participated in, and another one "Meaning of Translingual" above that you haven't participated in, etc. If you believe you have been starting and participating in BP discussions on these issues, please point out which discussions those are, and please show me that you're actually seeking consensus rather than just re-asserting the same points over and over again. Although I get frustrated with Richard, too, I can say that he observed the moratorium that I asked for, which you didn't, and he has opened several discussions to try to resolve these issues, while it appears to me you've opened none. Benwing2 (talk) 07:54, 26 July 2023 (UTC)[reply]
BTW I have told you multiple times over several weeks to participate in BP discussions and seek consensus. I don't know where you're getting this idea that I never said such a thing until now. Benwing2 (talk) 07:56, 26 July 2023 (UTC)[reply]
I said no such thing. I said I don't know how I'm supposed to participate in the discussion if participating in the discussion counts as "refusing to engage in discussion". You even criticized me for not being at the Beer Parlour when we are at the Beer Parlour. It's like we're through the looking glass, where joining the discussion at the Beer Parlour can get my block extended for refusing to join the discussion and not being at the Beer Parlour. kwami (talk) 08:07, 26 July 2023 (UTC)[reply]
(1) not sure what you mean by Wikilawyering, since you don't provide an example. I've been arguing that WP entries should meet our criteria for dictionary entries, and many of them do not. I'm not sure if that's what you mean or not.
(2) I have been participating in the discussions to some extent, but my intent was to present my case and let people discuss it and see if a consensus emerged. I was trying not to add my 2 cents after every comment. Really, what's there to discuss? I present my argument, and people can judge for themselves. I might explain further if it's not clear what I mean. And indeed, if I repeat myself where people don't seem to get the point, you criticize me for "refusing to engage in discussion."
You said I should join in e.g. the threads "NFC v. SOP" and "Meaning of Translingual". But I did comment at "NFC v. SOP". I don't have a strong opinion on SoP for letters, so I restricted myself to clarifying Richard's point to someone who didn't seem to understand it. As for the other, I had no idea what Richard was trying to say, so I had nothing intelligent to add and therefor didn't say anything. Rather than trying to impose my view, I was waiting for people to come to consensus so I could follow that. kwami (talk) 08:27, 26 July 2023 (UTC)[reply]
I've gone ahead and added my opinion, though I don't know if it's on topic because I don't follow what the topic is supposed to be. I think Richard might be saying it's okay to present English or Burmese as 'translingual' and then let that be overridden by the entries for specific languages, but that's only a guess. kwami (talk) 09:03, 26 July 2023 (UTC)[reply]
No, I'm looking for a criterion to decide whether a lemma is translingual, and at what assertions can be placed there. Whether facts about an English or Burmese letter belong in a the article for the translingual letter depends on how close they are to being universal for the letter. I hesitated at reverting the changes from 'translingual' to 'Burmese' for voiced stops when I realised that the use as voiced stops might be limited to Burmese and Burmese pronunciations, and that many Burmese script alphabets may actually lack the letters. Mon needs (and has) different entries for these letters, and the Pali Burmese script entries for these letters would need a separate division for Mon, with complexity dependent on the truth of the Tai Burmese script variants for Pali, whose documentation sometimes gives a strong whiff of 'We haven't quite finished inventing our Pali alphabet yet, let alone tested it'. For example, when I find assuredly durably documented Burmese-script Pali used by Shans, it's written the Burmese way.
The fact that a lemma is translingual does not mean that its content is suitable. There are some means to address the content of entries. {{rfv}} and {{rfc}} usually cover bad senses, and {{rfv-quote}} covers fabricated quotation, but for disputed content there seems to be nothing but edit-warring for irrelevancies masquerading as quotations, and pronunciations seem no better. Moving away from translingual entries, we then find doctored quotations, which are quite common for Prakrit, but are better than merely citing dictionaries without quotations, which is the usual resource for Prakrit, and is better than some languages, which often have nothing to back their entries.
So, one could start with a translingual entry defined a letter as being the same in all respect as an English or Burmese letter, but if that is atypical it should be revised. For example, the English and Welsh 'f' are the same letter, but have little overlap in pronunciation. Initial content doesn't have to be good, but we progress by improving rather than making the best the enemy of the good. I've found some horribly empty Pali lemmas during my occasional clean-ups, but if I can find anything for the word, I fill in the gaps rather than request the word's deletion for lack of content. --RichardW57m (talk) 13:17, 27 July 2023 (UTC)[reply]
I'd like to check editors' feelings on this issue. For something used in words, in each language individually, if the something is only a sum of parts, e.g. a letter plus diacritic marks that convey their own meaning, would that disqualify it from inclusion as an element of the system for writing words, even though it be expressible as a single Unicode character? That does not preclude it being included because the combination is idiomatic (a separate issue) or because it is considered a letter of the language's alphabet (or set of modifiers, if distinct)?
Oppose Whether a decomposable letter has an atomic NFC form (e.g. Latin é but not Cyrillic и́) is of total irrelevance to Wiktionary, because it wasn't done for linguistic reasons. All it tells us is that someone managed to get it encoded in Unicode before they introduced the policy of not encoding them separately anymore, and all that means is that it was probably part of some older encoding standard that Unicode replaced. That doesn't mean letters which don't have NFC forms are any more or less valid or relevant. At its extreme it would entail having pages for thousands of Korean syllables, simply because they were encoded atomically.
Plus a consequence of using this as a benchmark would be to discriminate against languages which use letters that don't have NFC forms, and I simply can't see the sense in that. Theknightwho (talk) 18:57, 24 July 2023 (UTC)[reply]
So you hate users who inch their way through text. Fair point about Hangul syllables, I think we reached a decision to exclude them a few years ago. --RichardW57 (talk) 19:06, 24 July 2023 (UTC)[reply]
If you want to include letters with accents because people might search for them, then that would entail including letters which don't have atomic NFC forms as well, like и́. Whether they're composed in NFC or not isn't relevant. Theknightwho (talk) 19:09, 24 July 2023 (UTC)[reply]
That sounds sane, but I don't really want us to add just any chain of combining marks - and what is a combining mark? If we ignore Unicode encoding, subscript consonants could be considered combining marks because of the analogy of Tibetan, and then most Tai Tham script monosyllables would come under this class, which means most of the native vocabulary for the Tai languages. (A few Tai Tham subscript consonants do follow the Tibetan model.) And even this definition is not closed under change of script (to Thai, Lao or New Tai Lue). The ancestor of many Indic transliterators here called the vowel marks 'diacritics', so I'm not sure where to draw the line between diacritics and other combining marks. Oh, and of course, in Tai Tham, we have MAI SAT, which can be a vowel shortener, a vowel mark, a final consonant and arguably even a tone mark, though I suppose we could 'just' classify instances by primary function.
A useful operational definition of what to include (as something, not necessarily a letter) besides the letters in grammars is the minimal non-blank string left after deleting from either end of a word. For all major systems I am aware of, that boils down to a Unicode codepoint, most likely one allowed in NFC, but perhaps not all of them in all these systems. --RichardW57m (talk) 10:08, 25 July 2023 (UTC)[reply]
OpposeSupport E.g. n and ñ are separate letters of the Spanish alphabet: to have an entry for 26 of its letters and not the 27th, merely because it is made up of distinct and meaningful written elements does not seem sensible to me. Similarly the two Is in Turkish, slashed O and A with a ring in Northern Germanic languages, etc. —Justin (koavf)❤T☮C☺M☯18:59, 24 July 2023 (UTC)[reply]
@Koavf Do you mean "support", or am I misunderstanding what Richard is asking? As for myself, I Support treating such characters as letters in the sense of having an entry for them. Note that we straightforward compositions when it comes to words, like encontrarse, which we define as "infinitive ofencontrarcombined withse". That's just the word equivalent of what you're talking about. So we should have entries (especially since many people looking them up are likely to be copy-pasting them in their composed form), but then depending on the case, either have a full letter entry or a redirect, indicating that it is "letter + combining diacritic". Andrew Sheedy (talk) 19:09, 24 July 2023 (UTC)[reply]
But both are a single precomposed letter, so both would be listed. It's news to me, that tilde is individually meaningful in Spanish, but even if ñ were not a precomposed letter, it should have an entry because it counts as a letter in Spanish, just as Welsh ng(letter) has and would have an entry because it is a letter. --RichardW57 (talk) 19:13, 24 July 2023 (UTC)[reply]
The wording of what you wrote is not entirely clear or grammatical ("even though they only be sums of parts"?) so I'm struggling to make sure that my point comes across in response. I'm writing that enye should have an entry and be considered a letter or broadly an independent graphical element of written Spanish. I have no perspective on if every possible combination of radicals in Chinese written language should have this status or every Cyrillic letter with a cedilla at the bottom or combinations that do not exist in any language like a long s with a circumflex and a cedilla should, etc. Examples in your introductory text and some very light editing/proofreading would be helpful. —Justin (koavf)❤T☮C☺M☯19:18, 24 July 2023 (UTC)[reply]
What grammar issue do you see? Do you think the subjunctive is inappropriate?
As to an example, we currently have Vietnamese ọ(letter), which consists of the vowel 'o' plus the Vietnamese tone mark ̣(“nặng”) - a sum of parts. Likewise with Latin macrons, or, to go back to Antiquity, apices, though it seems a bit odd that we don't have Latin ā(letter). Perhaps it's because the combination is used by us to write Latin, but not by the classical Romans.
I'm not proposing that every combination of base and marks be included. Although Wiktionary is not paper, the number of combinations is infinite, which is beyond the WMF memory capacity.--RichardW57 (talk) 20:17, 24 July 2023 (UTC)[reply]
"even though they only seem to be sums of parts"? "even though they only are sums of parts"? "even though they only could be sums of parts"? Etc. Not sure what you mean here. There are finite numbers of combinations, unless you count adding an infinite number of acutes or diareses above a letter in a stack up to the sky, in which case, yes, you could have an infinite number then. I also agree that we should not have an endless amount of combinations of things that do not actually exist in real languages, but I am inclined to say that we should have combinations that actually exist. —Justin (koavf)❤T☮C☺M☯22:26, 24 July 2023 (UTC)[reply]
The subjunctive isn't universal in English, and it's used to different extents in different places, by different generations, by different people, etc. The phrase, "even though they only be sums of parts" is grammatical in my idiolect as well. Andrew Sheedy (talk) 22:29, 24 July 2023 (UTC)[reply]
Comment Adoption would mean deleting a number of articles for Unicode entries. For example, there is a long run of Unicode characters intended specifically for Vietnamese, though they may also be used for other languages of Vietnam. There is no attestation of those characters ever being anything but SoP, so they would all need to be deleted, or possibly changed to redirects to the base letter, assuming of course that we list the possible tone marks in that article.
Reg. the discussion leading up to RichardW57's last comment, he did specifically say at the beginning of this proposal that this would be for SoP. Any combination of letter + diacritic that could not be predicted from the language's own orthography, as in Spanish, German or Polish, regardless of whether they're considered letters of the alphabet, would still be acceptable as WT entries. But something like the Hawaiian vowel letters with macron would not be. kwami (talk) 22:20, 24 July 2023 (UTC)[reply]
This is what I’m opposed to, yes. Arguments about ñ being a letter in Spanish miss the point, because they have entries by virtue of not being the sum of their parts. What Richard proposes is that we have purely SOP combining forms purely because Unicode decided to encode them with their own codepoints.
This would lead to absurdities like having a Serbo-Croatian entry for í (stressed Latin i), as it’s encoded atomically as U+00ED, but not и́ (stressed Cyrillic и), because it can only be composed from its component codepoints. Given there’s a one-to-one correspondence between them in Serbo-Croatian, we should not have an exemption for one but not the other.
Note: I am not making an argument for or against the two examples above - I’m just saying that a policy which includes one and excludes the other is incoherent for our purposes. Theknightwho (talk) 01:34, 25 July 2023 (UTC)[reply]
If I understand this correctly, I would be in favour of having entries for both the precomposed and composed letter-diacritic combinations. Someone copy-pasting isn't necessarily going to be able to figure out how to get to the entry for just the diacritic, otherwise. Andrew Sheedy (talk) 01:52, 25 July 2023 (UTC)[reply]
Right, exactly. That’s a much more sensible rationale. I assume you’d draw the line at the 11,172 Hangul syllables which have been atomically encoded, as they’re all composable from their jamo components. We should obviously still have one-syllable Korean terms (not SOP), but it’s completely irrelevant that 쐆(sswap) happens to be encoded as U+C406, because it doesn’t mean anything beyond ᄊ(ss) (U+110A) + ᅪ(wa) (U+116A) + ᆹ(p) (U+11B9). Theknightwho (talk) 02:20, 25 July 2023 (UTC)[reply]
Being ignorant of how any Asian scripts work, I would be fine with leaving those kind of decisions to the editors of those languages. However, speaking for myself, if I wanted to look up one of those Korean terms, I would copy-paste the whole symbol 쐆 (you cannot, on a computer, highlight each component individually). How as a user would I figure out that it was (a) composite, and (b) what the components were? Andrew Sheedy (talk) 02:29, 25 July 2023 (UTC)[reply]
If you click on your red link, you will at least get a decomposition. You just have to know to click on it even though its red. --RichardW57 (talk) 02:37, 25 July 2023 (UTC)[reply]
Oh, good to know! However, if I were a typical user putting it into the search bar, it would simply tell me there was no page and I would walk away disappointed that Wiktionary didn't have an entry yet. Andrew Sheedy (talk) 02:51, 25 July 2023 (UTC)[reply]
On a completely unrelated note, I'm pretty sure we hit a record today for the biggest Beer Parlour page (and July isn't even over)... What a month. Andrew Sheedy (talk) 02:51, 25 July 2023 (UTC)[reply]
Also, just to add: the software automatically merges the composed and precomposed forms, so we don’t need to worry about that. No redirects or whatever are necessary. That’s in keeping with the Unicode standard that the composed and precomposed forms should usually be treated as the same thing, because ultimately it’s of no relevance to the user how many codepoints there are - all they care about is whether they’re dealing with the right characters or not. Theknightwho (talk) 02:26, 25 July 2023 (UTC)[reply]
There is a fair bit of software that saves the text that was typed scalar value by scalar value, without changing it to something that, if certain people had their way, would be different. However, there's also a lot of software that forces text to NFC. What's lacking is the non-programming ability to convert the text to NFD, delete the last character, and convert back to NFC. Unicode gave up asking for regular expression engines that worked on strings modulo canonical equivalence. --RichardW57 (talk) 02:35, 25 July 2023 (UTC)[reply]
Oppose per Theknightwho. Pretty sure I would also oppose entries for the said Vietnamese combinations if I were into Vietnamese. If there are Vietnamese sections of the base letters and the combinations only ever were used for Vietnamese then they could be hard redirects, but even this is not useful, and many of such combinations are deprecated encodings also. We even exclude precomposed positional forms of Arabic letters in page titles as spam, the situation is comparable: A letter changes form in environment and the changed forms, subforms of the “base” form, have been separately encoded—only that in the original post the separately encoded forms stand for two letters while in this Arabic case such a separately encoded form stands for one letter nigh another letter or two other letters instead of white-space. The Unicode encoding of aught is not as a rule lexicographically suggestive, as also ISO language codes are not universally accepted by Wiktionary to lead to senseful concepts. Fay Freak (talk) 00:59, 25 July 2023 (UTC)[reply]
@Fay Freak: The best way to foul up rendering of Tai Tham text is to normalise it, so I don't know what you mean by 'deprecated'. Are you perchance referring to composition exclusions? Hard redirects are often a bad idea, and soft redirects would be the way to handle scalar values with multiple interpretations as base letter plus diacritic(s). --RichardW57 (talk) 03:32, 25 July 2023 (UTC)[reply]
Support. Given the characters we will see, and the poor support of the concept of a trace of characters, I think we should support some comprehension of a flow of characters a user is likely to encounter. --RichardW57 (talk) 03:39, 25 July 2023 (UTC)[reply]
Mu? IMO, anything which is felt by speakers of a language or by reference works to be "a letter" (or "a symbol", "a character", etc) in a language(s) should be seriously considered for inclusion (if it meets attestation requirements, etc), but whether Unicode has given it a single codepoint, or whether NFC or NFD normalization does anything to it, is irrelevant. As TKW put it, a metric that treats í and и́ differently just because one has an atomic codepoint and the other doesn't makes no sense. - -sche(discuss)15:20, 25 July 2023 (UTC)[reply]
@-sche: But they usually behave differently under backward deletion. The Roman script form disappears completely, but the Cyrillic form disappears completely. More extremely, while the Tai Tham Northern Thai ᩃᩪᩢ(“sprog”) is a base consonant plus two marks - one grapheme cluster, while the Thai script spelling of Northern Thai ลูก has two base consonants and one mark - two grapheme clusters.
At this point, the split is 3 support, 2 oppose, 1 roughly abstain, so on this scanty evidence it doesn't look worth taking to a formal vote, and so the status quo should stand. But what is it? Well, WT:CFI says that we include 'Characters used in ideographic or phonetic writing such as 字 or ʃ. Graphic Unicode characters are characters, are they not? Accordingly, precomposed characters qualify for inclusion if they are attestably used in words, even if they aren't letters. (Unused Hangeul syllables fail this test.) This is also the clause that allows ligatures such as त्र, and, to a renderer, a precomposed character may often be considered a ligature. --RichardW57m (talk) 15:18, 27 July 2023 (UTC)[reply]
But that also applies to и́, and I'm not convinced that one of the previous support votes actually understood the question in the first place. Theknightwho (talk) 19:05, 28 July 2023 (UTC)[reply]
Whether и́ is a character is debatable, and for now I'm inclined to postpone the debate to another day. What the straw poll probably demonstrated was the lack of an overwhelming majority one way or the other - 3:2:1 or 2:3:1 makes no real difference. Except for unusual fonts, a renderer will treat и́ as two glyphs, and will not apply hinting to them as a unit. --RichardW57 (talk) 02:05, 29 July 2023 (UTC)[reply]
As already pointed out, we already have a character info box on pages for canonically composed (NFC) code points that shows the code points in the canonically decomposed (NFD) equivalent, so someone can click the links in the character info box and find out the meanings of the component code points, if someone has created an entry that says, for instance, that this diacritic marks stress or a given tone in this language. THe box is shown on nonexistent pages too via MediaWiki:Newarticletext and MediaWiki:Noarticletext.
Some readers would benefit from entries on minimal parts of text that can be selected with a mouse, or with the shift key and an arrow key, but just adding entries for canonically decomposable code points wouldn't achieve that. You would have to have entries on all grapheme clusters that are likely to be encountered, even if they contain multiple code points. At least my browser (Firefox) only allows selecting whole grapheme clusters. I can only select all of की in text on a page or an unmodified text box (and in the edit summary box), but can separately select क and ी in the edit box when the "Syntax highlighting" button (marker icon) is turned on. (I can't individually select the lowercase letter a plus combining grave accent if I preview this and copy it into the text box: à. Not sure how the writers of the syntax highlighting plugin achieved that.) Using the backspace does delete ी and then क, so it behaves differently from selection. Using the delete key at the beginning of की deletes the whole grapheme cluster when syntax highlighting is turned off, so if you want to find out what ी is, you apparently can't isolate it and go to the entry using the mouse or keys. So apparently in a browser you can only isolate whole grapheme clusters by selection, or the prefixes of a grapheme cluster using backspace.
Adding entries for all grapheme clusters would be a lot of work. For instance, entries for all combinations of an Arabic consonant letter with or without shadda and a vowel. This is not Richard's proposal, but it would more completely satisfy the purpose of his proposal, so I'm pointing it out. Another way of partially achieving it would be to make an extended character info box to show the code points in a grapheme cluster and put that on nonexistent pages. But right now there isn't a Lua module to identify grapheme clusters or show multiple code points in a table. — Eru·tuon07:44, 29 July 2023 (UTC)[reply]
I'd like us to have a tool that displays the canonical decomposition of a string that works the same way the display of the decomposition of a precomposed character works. However, there are a number of problems that seem major for me:
Advertising it
The interface. Saying "type this into WT:Sandbox" is not exactly user-friendly.
Hi. I notice some people are (ab)using the transliteration param of {{ux}}, {{quote}}, etc. to specify a "normalized" version of text already in Latin script, typically in dead languages. I think someone (User:Vininn126?) requested a normalization param to use for this purpose, which I think is a great idea. Any thoughts about this? The practical things, if we decide to add this:
What should it be named? I'm thinking either |norm= or |n=.
How should it be formatted? There will need to be at least two, maybe up to four formats: One for multiline usexes/quotes, one for inline usexes/quotes, potentially one for link templates if we choose to implement this (see next point), and potentially one for headwords if we choose to implement this.
Which templates should get it? Only {{ux}}, {{quote}} and variants, or more broadly (which would include {{syn}}/{{ant}}/etc., {{col}}/{{derN}}/etc., {{alt}}, maybe {{head}}, and lots of others)? Adding it across the board would be a lot of work (it would be great if the handling of link params were centralized, so you could add a new one in just one or a handful of places, but it isn't). Before adding it broadly, I'd like to see some examples of use cases involving these other templates, so we don't end up doing a lot of work that isn't actually used.
@Benwing2 Yes, please, I would desperately need such a parameter, particularly for historical variations of Polish. (On a similar note, if we add this, we're gonna have to update all the Old Polish and Polish quotation templates to change {{{tr}}} to the new one). I feel like this should go in just ux, quote. Perhaps cl, but not the others. Vininn126 (talk) 21:47, 24 July 2023 (UTC)[reply]
I also requested it for Okinawan entries. I'm fine with both parameter options. I don't think we'll need it for anything past ux & quote as syn, ant, col, and such should just link to the proper lemma anyways. AG202 (talk) 22:12, 24 July 2023 (UTC)[reply]
@AG202, Vininn126 See User:Benwing2/test-usex. The normalization is specified using |norm= and goes in parens, before the transliteration if both are present. An Okinawan example that uses both normalization and transliteration would be great. Note that I haven't added any CSS styling yet but will probably make the normalization be italicized. Benwing2 (talk) 04:10, 25 July 2023 (UTC)[reply]
@Benwing2 Looks fine to me too. I wish that the inline versions looked less crowded, but I guess that there's nothing that can be done about that. AG202 (talk) 13:22, 25 July 2023 (UTC)[reply]
|norm= would be very useful – I even went and created a separate template for Chinese Pidgin English ({{cpi-q}}) for this exact purpose; previously |lit=was abused for the normalised text instead, because |tr= was already in use.
Though I think [square brackets] might be a better option here, because it denotes that the text was changed. – Wpi (talk) 13:46, 25 July 2023 (UTC)[reply]
@Benwing2, Vininn126: Am I adjusting for Pali typos, oddities and ignorance wrongly? I generally output things for non-Roman quotations in the order: raw text, transliteration, transliteration with fixes, translation. This is on the principle that the reader may want to know what was actually written, but can't read the script - or hasn't got the font. It also works well for the Lao script, where in real life the writing system used is often ambiguous, and the normalisation has to be applied at the transliteration level. (I suppose I could fix the Lao script by switching to a larger, unambiguous alphabet in the normalisation.) --RichardW57m (talk) 16:18, 27 July 2023 (UTC)[reply]
@Vininn126: Yes. For typical Lao-script Pali, the fixing, to comply with the current new scheme would need the order:
Pali as seen.
Pali in extended alphabet, with corrections (|norm=)
Transliteration of the above.
Translation.
At present, the usual order is:
Pali as seen.
Transliterations:
Smart transliteration of the Pali as seen.
Transliteration with corrections* and remaining ambiguities resolved using the meaning.
Translation.
* Sometimes in the smart transliteration, especially if the original would defeat transliteration.
There is an example of the latter at the second quotation at ສັງຄະ(saṅga). There are no mistakes in that text. I don't know how widely available fonts that could handle the extended alphabet are; the extra characters were only added a few years ago. The Pali in the extended alphabet would be (ຕະຕິຍັມປີ ສັງຆັງ ສະຣະນັງ ຄັຈຉາມິ ຯ), which won't render properly so widely. And I didn't even choose this example to show a problem! I expect this little problem will eventually go away without us doing anything about it. --RichardW57m (talk) 17:30, 27 July 2023 (UTC)[reply]
Is there any recommended way of showing what was changed? I saw from the examples that I can't quickly see the real changes when comparing plain and italic Cyrillic. I've been adding asterisks to the Pali text to show where the quoted text is considered wrong. --RichardW57m (talk) 16:18, 27 July 2023 (UTC)[reply]
@RichardW57 If it's the italicized Cyrillic that's an issue, we can change it (using CSS) so that Cyrillic isn't italicized. If that's not the issue, maybe we can use a colored background or something. This is similar to how boldfaced text is converted to a pink background in Arabic-script text; see أراق (or any entry in Category:Arabic terms with quotations) for an example. The change from boldface to pink-backgrounded text is done in CSS, and we can do something similar with Pali scripts if needed. Other possibilities are underlining, using brackets, etc. Benwing2 (talk) 19:56, 27 July 2023 (UTC)[reply]
That's different to what I had in mind, but bracketing the whole normalised version would work. That wasn't what I had in mind, but it works. What I had in mind was highlighting the changed words in some fashion, which would be an editor's responsibility.
I think the Roman script is the only Pali script that will italicise - at least that's the only one which italicises for {{m}}. That's slightly odd, as I know that Thai italicises in Thai books. --RichardW57 (talk) 22:32, 27 July 2023 (UTC)[reply]
@RichardW57 You've misunderstood what I said. I indeed meant highlighting only the changed words or characters, and my intention was that this has to be marked up manually to get this. The markup could be boldface using triple quotes (which, depending on the script, might display with a colored background, as happens in Arabic text), underlining (again, this could be changed using CSS to do something else), bracketing (around just the changed word), etc. Benwing2 (talk) 22:42, 27 July 2023 (UTC)[reply]
Also, we've had this discussion before about italicizing non-Latin scripts. User:Erutuon believes this shouldn't be done, and implemented it this way, so that italicizing is ignored for such scripts; I don't remember the reasons but I disagree. After the last discussion, we changed it (I think) so that Greek does italicize, but IMO this should be done for Pali scripts as well if they support italicization. Benwing2 (talk) 22:45, 27 July 2023 (UTC)[reply]
I can't remember any real reasoning. It might have just been that Wikipedia doesn't italicize non-Latin script in English text because they are already visually distinct, but I don't know what's actually done outside of Wikipedia and maybe it's an odd tradition that doesn't have to be continued. — Eru·tuon07:51, 29 July 2023 (UTC)[reply]
I would strongly object to italicising scripts that are not generally italicised or never were historically. We shouldn't ever be italicising Tangut, for instance. Theknightwho (talk) 14:07, 29 July 2023 (UTC)[reply]
@Benwing2 Perhaps I should have used Chinese as an example instead. In fairness, though, I don't think italicised text is necessary for non-Latin mentions, because there's no necessity to distinguish the term from the surrounding text. It possibly makes sense to italicise the transliteration, which I think is what we do at the moment. Theknightwho (talk) 18:53, 29 July 2023 (UTC)[reply]
@Theknightwho Your thinking is constrained; there are a zillion cases that aren't mentions where italicization could be useful, but is arbitrarily turned off. User:Sarri.greek outlined a bunch of them for Greek, which is why we enabled it, but the same should apply to all scripts that regularly support italicization. Benwing2 (talk) 18:58, 29 July 2023 (UTC)[reply]
@Theknightwho I may be misremembering; the discussion is here Wiktionary:Grease pit/2022/August#Italics as alt in Greek and mostly seems to concern italicized taxonomic names in Greek and boldfacing to indicate a quoted section of text. Richard also wants italicizing or some other such means to indicate the difference between non-normalized and normalized versions of a script. I think there was at least one other discussion about this but I'll need to find it. Benwing2 (talk) 19:50, 29 July 2023 (UTC)[reply]
@Benwing2 Thanks - I'm curious what other use-cases there are, because I've no problem with it if it's either (a) standard within a language (e.g. taxonomic names), or (b) is being used to distinguish other text in the same script. Though in that second case, we'd need to have a backup for scripts which don't use italics (underlining, maybe?). Theknightwho (talk) 19:56, 29 July 2023 (UTC)[reply]
@Theknightwho Arabic uses a pink background to indicate text that would be boldfaced, see أراق for an example (boldface is barely visible in standard fonts used for Arabic). The same should be done for Persian, etc. We could use a different-colored background to replace italics (although keeping in mind colorblindness, which is most commonly between red and green). Benwing2 (talk) 21:49, 29 July 2023 (UTC)[reply]
@Theknightwho: Yes; I think it's possible to choose colors that aren't problematic. (BTW I notice when you ping me it's always substed; do you manually subst it or is this somehow part of the Reply gadget?) Benwing2 (talk) 22:47, 29 July 2023 (UTC)[reply]
@Benwing2 If you type "@" in the source version it should show you a drop-down list of people who've replied to the thread so far. It'll also work if you tag someone new, as well, but you have to type their username manually. Theknightwho (talk) 23:10, 29 July 2023 (UTC)[reply]
@RichardW57, Vininn126, AG202 BTW I'm thinking that if there's normalized text supplied in a non-Latin script, the transliteration should go off of the normalized text if possible rather than the original. That would allow, e.g., for Urdu or Farsi text to be presented in its original form as well as a normalized form that contains vocalization, which then gets transliterated automatically. Benwing2 (talk) 19:59, 27 July 2023 (UTC)[reply]
It makes sense for each line to be the precursor to the next. However, I'm seeing problems with the whole scheme for Pali. For example, the process of editing a Burmese script Pali text to give a Roman script Pali text starts with, to take the stem to maximise blueness, သံဃ(saṃgha) and ends with saṅgha. If one normalised at the Burmese script to get the end result, I would get သင်္ဃ(saṅgha), which failed RfV. At the moment I'm just assuming that the interested reader knows to convert niggahita before stops to the homorganic stop before using a dictionary. There may be the converse problem in the Tai Tham script with Northern Thai Pali ᨾᩘᩈ(maṅsa), though on the basis of a monk's Pali name cited in Tai Khuen text, I think the obvious normalisation ᨾᩴᩈ(maṃsa) may occur in Tai Khuen Pali, and I've no idea which the spelling is in Lao Tai Tham script Pali. --RichardW57 (talk) 23:21, 27 July 2023 (UTC)[reply]
I suppose we could add a separate param for a normalized translit, but I'm not sure how much demand there is for it cross-linguistically. Benwing2 (talk) 23:36, 27 July 2023 (UTC)[reply]
It might be restricted to languages with a script that we actually transliterate that also use the Roman script in an orthography very similar to the transliteration. Perhaps the problem is that it seems reasonable to also present Pali quotations in Roman orthography, rather than say leaving it in an odd-looking and sometimes ambiguity-preserving transliteration. I can't help feeling that if we present a quotation in two forms of non-Roman script, we need transliterations for both of them. We can probably get away with it if they are very similar. People can understand the principles of a writing system without understanding its glyphs.
I note that two levels of transliteration are being used for Akkadian, e.g. gammalu. First there is the reversible transliteration, full of Sumerograms and determiners. Secondly, there is the resolved, hopefully phonemically spelt form. This is similar to what I do for the Lao-repertoire (Lao-script) Pali I mentioned above. --RichardW57m (talk) 09:50, 28 July 2023 (UTC)[reply]
I think it would be pushing the semantics. In the systematic case of of Lao-repertoire Pali, the transliteration of the text often corresponds to Lao phonology, whereas the refinement corresponds to the original Pali phonology. I suppose it could be argued that the Lao script is an inadequate representation of Pali phonology. However, {{quote-book}} doesn't have a documented |ts=, so I could be scolded if I used it and it was then withdrawn. (Notifying @Sgconlaw.)
I suppose one could use it for converting to the orthography of edited texts in Roman script. Using |ts= for corrections would be an abuse. Perhaps we don't need to tell the reader what is going on in parts assessed as wrong, and corrections could then be done in script. So, in a complex case, we could have:
|text= the observed text in non-Roman script
|norm= the amended text in non-Roman script, consonant with the orthography for the script, region and date of production
@Benwing2: There's a potential issue with driving the transliteration from the normalised text - the mark-up will also get transliterated. That may or may not be a surprise, and may or may not be an issue. It may also constrain the extent of the highlighted text. For Pali, replacing a whole number of aksharas is OK, but one may have to work with syllables for the Burmese language - the transliteration unit is larger than the rendering unit. I'm also hitting punctuation problems with automatic transliteration, though these are nothing new. Prose and verse seem to need different rules! Perhaps I need to be pickier; Pali also has the question of word division. --RichardW57m (talk) 13:11, 31 July 2023 (UTC)[reply]
@RichardW57 I think it's OK to see the yellow highlight in both normalization and translit; probably the right thing, in fact. I think User:Theknightwho's changes to translit mean that the transliteration module won't actually see the HTML markup but it will be passed through unchanged. Note also the substitution mechanism using |subst=; this applies to the normalization before transliteration, and you can use it to do things like remove the markup or otherwise massage the normalization to get the right translit output. Benwing2 (talk) 18:34, 31 July 2023 (UTC)[reply]
Ah, that testing is something I've not got geared up for. I've now done a test of 'normalisation' for Pali using Pali ယက္ခ(yakkha). (The only template etc. of mine using the new parameter {{|norm}} is {{RQ:pi:Shan Paritta}}, and only when fed from Module:RQ:pi:Shan Paritta.) While the italic text works in itself, the font size for the Burmese script needs to be enlarged to get an x-height similar to the Roman script. Is the normalised text intentionally smaller than the basic text? --RichardW57 (talk) 04:08, 30 July 2023 (UTC)[reply]
@RichardW57 It isn't intentionally smaller. I suspect the issue is something to do with CSS classes; the smaller font size is what you get by default and normally the font size is increased by something in MediaWiki:Common.css. Either the normalization isn't properly script-tagged or the CSS tag of the non-italic font is increased by MediaWiki:Common.css. Let me investigate further. Benwing2 (talk) 04:14, 30 July 2023 (UTC)[reply]
@Benwing2, Fenakhay: Testing now hidden by Fenakhay. Unfortunately, doing testing of my usual mechanism outside main space needs three unusually named files - data module for quotations themselves to exercise the new capabilities, quotation picker template 'RQ:LANG:basename quote' and simple quotation template, rather than easily feeding off production work, or simply using a new quotation template. I suppose I should be thankful the quote-* templates don't check for misspelled parameters, which I had been going to suggest! Perhaps I can slightly complicate matters to avoid the need for a separate data module by letting the quotation picker templates override the choice of simple quotation template. And I'm already behind with the documentation! --RichardW57m (talk) 09:35, 31 July 2023 (UTC)[reply]
@Benwing2: A natural case of the full works is in the 1984 quotation at သံဃ(saṃgha): |norm= to correct a typo in original, |ts= to switch from niggahita to homorganic consonant, and |lit= to remove a word which is excessive to English tastes. The first two are cumulatively highlit. --RichardW57 (talk) 00:29, 8 August 2023 (UTC)[reply]
@RichardW57: Looks good, although I'm not sure you need the |ts= here; the switch to a homorganic nasal consonant is such a natural thing that it probably doesn't need mentioning. Benwing2 (talk) 01:12, 8 August 2023 (UTC)[reply]
@Benwing2: The |ts= is used to show the normalised Roman script form. IAST, which we have adopted as the transliteration scheme, distinguishes niggahita (ṃ) and homorganic nasal (ṅ/ñ/ṇ/n/m). There's a widespread formal preference for the homorganic nasal (except before fricatives) in Pali, and dictionaries tend not to show the alternative form with niggahita before stop consonants, and it is sometimes taught as wrong. If the reader is to look the other words in the quotation up from the Romanised form, he would in general not find the words spelt with niggahita before stop consonants. --RichardW57 (talk) 02:38, 8 August 2023 (UTC)[reply]
┌────────────────────────────────────────────────────────────────────────────────────────────────────┘|ts= won't be withdrawn but should be documented. Note that in my addition of non-Latin script handling of titles, authors, chapters, etc., the |ts= param variant is supported. I can add support for |norm= and |lit= variants if it will help. Benwing2 (talk) 23:08, 28 July 2023 (UTC)[reply]
I don't understand what you mean by "|norm= and |lit= variants". You're adding |norm= and |lit= already exists for the {{quote-*}} family. You've finally persuaded me that the new scheme can be used to improve the presentation of Pali quotations. --RichardW57 (talk) 02:17, 29 July 2023 (UTC)[reply]
@RichardW57 Good to hear. What I mean by "|norm= and |lit= variants" is actually <norm:...> and <lit:...> inline modifiers for author, chapter, title, etc. (now that I'm planning on implementing the subproperties of these params using inline modifiers). For each param such as |author=, |title=, |chapter=, etc., there will be various inline modifiers supported, including <t:...> (the translation/gloss), <tr:...> (the translit), <ts:...> (the transcription) and <sc:...> (manually specified script code). Do we need <norm:...> and/or <lit:...> in addition? Benwing2 (talk) 02:26, 29 July 2023 (UTC)[reply]
@Benwing2: I take it that if you support for one parameter, it will be actually be easier to support for all, barring any special issues because of formatting, as in numbers.
So one might have |location=Rangoon<norm:Yangon> or |location=Calcutta, Bengal (<norm:Kolkata, India>). I can't help thinking that the example I gave would be a waste of screen space, better served by giving links for (partially) superseded place names and territories.
In so far as links are to be precise versions of documents, such details are unnecessary. However, if the text was available in various versions, one might expect to find substantially the same text in most of them, so one might want a more generic reference. So, a reference to one of the four books of Kings, as in the cited work, might need a pointer to the more modern scheme of having two books of Samuel and two of Kings. <lit:> may come into its own where a foreign-language work has multiple English translations of its title, some more literal than others. I've seen a lot of minor variations in the names of parts of the Tipitaka; one might similarly want to emend now discarded forms (or outright typos) in a particular edition to more current names.
I don't think we need them, and where they seem useful, the problems are generally larger - a matter of alternative forms. For example, is a certain Chinese linguist "Fang Kuei Li", as on the cover of an important book he wrote, or "Li Fang-Kuei"? We're generally not citing the metadata as evidence of words, though I have on occasion cited titles as evidence for words. --RichardW57 (talk) 10:45, 29 July 2023 (UTC)[reply]
@Benwing2: I could see this also being helpful for particularly baffling examples of early modern English spelling (I spent more time than I should have puzzling over Stanyhurst in the gyreful citations). Would that fall under the intended uses? —Al-Muqanna المقنع (talk) 12:47, 31 July 2023 (UTC)[reply]
@Al-Muqanna I don't see why not. A lot of early modern English spelling, including Shakespeare's manuscripts, was non-normalized and a "normalization" param sounds like what we need. BTW I pushed the code to production for {{ux}} and {{quote}}. {{quote-*}} is next. Benwing2 (talk) 18:23, 31 July 2023 (UTC)[reply]
@Benwing2 When the normalization for Latin is ready, can we bot convert all Old Polish RQ templates to stop using {{{tr}}} but instead {{{norm}}} and also update any docus using the word "transliteration" to "normalization"? Vininn126 (talk) 08:22, 1 August 2023 (UTC)[reply]
@Vininn126 Yes, I'll change the wikicode of the templates so that any reference to |tr= in a call to {{quote-*}} will be changed to |norm=. This should avoid any need to change the callers of the templates in question, presumably. Benwing2 (talk) 08:41, 1 August 2023 (UTC)[reply]
@Vininn126 I am doing a bot run to change |tr= to |norm= in Old Polish reference templates. Note that some templates don't include a |norm= parameter in them (50 of them out of 338):
I also notice that some templates include a |pages= param, which is no longer needed; |page= and |pages= are now synonyms, and the module automatically figures out whether to show page or pages (if the module guesses wrong, there's also |page_plain=, which lets you explicitly specify the word "page(s)", but that's probably not needed for your templates). The templates that include |pages= are almost the same as the ones that don't include |norm= (53 of them):
I think it's a little silly for us to make up our own usage examples when we also have citations from formally published sources that do a good job of showing how the word is used. At tunnel vision, for example, the sense "A restricted field of vision as the result of looking through some type of object or a medical condition." has a usage example "It took some time for him to get used to the tunnel vision he experienced while looking through a periscope." and a quotation from a psychology textbook "The most frequent visual distortion was 'tunnel vision,' in which the officer became so focused on one object that everything else at the scene went unnoticed.…" In my opinion, the usage example doesn't add anything. —Kodiologist (t) 12:42, 25 July 2023 (UTC)[reply]
In my opinion, the two fulfil slightly different roles - quotes can sometimes get rather lengthy and are used to prove a definition's existence, whereas a usex is meant to be quick, short, and particularly illustrative. Both should be illustrative, but for me the usex can be almost cliché at times, if that makes sense. If a page has both, they should be somehow markedly different. Vininn126 (talk) 12:45, 25 July 2023 (UTC)[reply]
To add to Vininn's point, it's worth bearing in mind that usexes are available without specifically clicking for them, so they're useful to illustrate distinctions at a glance. Citations ideally also demonstrate a sense's use over time, which is more involved than the purpose of a usex. —Al-Muqanna المقنع (talk) 14:19, 25 July 2023 (UTC)[reply]
Agreed. Some entries are packed with quotes, which can be overwhelming to a user (and mean that the best examples get lost in the mix). Usexes serve the useful purpose of demonstrating the most typical usage and doing so in a simple way that is perhaps more accessible to readers than many quotes. However, bad usexes should be deleted or replaced if they do not serve that purpose. Andrew Sheedy (talk) 14:53, 25 July 2023 (UTC)[reply]
This. Quotes may be too complicated for learners or at least too long to read and comprehend. For the most basic terms usage examples need to be made up to get the point across; my exemplary entry for this point is German stellen. What’s the point anyway to quote everyday-life questions like Kann ich dir eine Frage stellen? or Die Beklagte stellte den Antrag, die Klage abzuweisen. occurring before every civil court’s reasoning? Sometimes I leave out quotes because they are contextually too obscure in the context of reading a dictionary, sometimes the usage examples are actually stolen snippets from the internet simplified—I remember Russian vulgarities, disagreeable also with occurrence in edited text. Fay Freak (talk) 18:19, 25 July 2023 (UTC)[reply]
I tend to remove a usex if there is a citation with very similar usage (e.g. same collocation), otherwise not. I'm more hesitant to remove them before we have the CFI-mandated three citations for the sense, as usexes may be helpful in knowing what to search for. Equinox◑14:58, 25 July 2023 (UTC)[reply]
The quote in the tunnel vision example is a terrible usage example, it isn't being used. Usage examples are meant to demonstrate the word or term being used in running text, the quote is a mention and description. The usage example my not be good, but the quote is worse in my view. Generally speaking a bespoke usage example can benefit from being tailored to demonstrate usage clearly and economically, so they are my preference if well done. I favor moving most/all cites/quotes out of entries if there are good usage examples, as they add a lot of bulk to the entry while often adding nothing of value for the reader. Cites are primarily meant to be for the benefit of the editor. - TheDaveRoss17:18, 25 July 2023 (UTC)[reply]
Yes, it has quote marks but the text talks about the concept, not the word, describing one specific case of tunnel vision, as experienced by that officer, it's clearly a usage. I don't agree that cites are just meant for editors, they make this project interesting. Usage examples can be well done, but more often than not they are unsourced quotes in disguise, or used for soapboxing or other personal fetishes. They are also more prone to errors, when "made up" by overly-confident non-native speakers. Jberkel18:24, 25 July 2023 (UTC)[reply]
I've definitely noticed a problem with usexes being used for soapboxing, IMO stuff like that should be changed or simply removed accordingly. The guidelines at WT:USEX are pretty solid I think. —Al-Muqanna المقنع (talk) 18:34, 25 July 2023 (UTC)[reply]
I don't understand the hate for usage examples that I'm seeing here. Of course they should be deleted if they aren't actually good examples of usage. But can anyone disagree that replacing the usage examples on take with (possibly long, complex, or archaic) quotes would be a terrible idea? Does anyone think the usage examples on unseat are redundant to the quotes? As for tunnel vision, I've added a usage example to illustrate use in a medical context which I think is more common. Ioaxxere (talk) 19:43, 25 July 2023 (UTC)[reply]
I broadly agree with Vininn. If the usexes are bad, soapboxing, etc, improve them (or remove them, if that's really best). But they can often be useful, especially if quotes are long, old, or not using the word typically (but are needed for ATTEST). - -sche(discuss)22:22, 25 July 2023 (UTC)[reply]
Should we go further and move citations that are not good usage examples, but are needed for attestation, to the citations page? DCDuring (talk) 23:14, 25 July 2023 (UTC)[reply]
I support this, quotes are important for attestation and for teasing out different senses, but I think a well written usage example is more helpful to a reader. JeffDoozan (talk) 23:22, 25 July 2023 (UTC)[reply]
Yes, IMO. If I add a quote that's needed for ATTEST but is otherwise quite bad, I usually think to put it on the cites page, and if I see a quote someone else added that's really egregiously unsuited-to-mainspace, I'll move it to the cites page. But I wouldn't move just any and all old/long/not-so-great quotes to the cites page, especially for more polysemous entries, because it takes work to maintain synchronization between the senses in the entry and the indications on the cites page of what sense each cite goes with. - -sche(discuss)04:58, 26 July 2023 (UTC)[reply]
I wrote up my thoughts on this issue at User:Soap#usex_vs_quote a few months ago. It seemed odd to me that quotes are collapsed by default and usexes are not, even though quotes are more important for establishing the existence of a word. But it makes sense that usexes are more accessible to readers, take up less space on the screen, and might encourage readers to contribute helpful sentences of their own. —Soap—12:21, 28 July 2023 (UTC)[reply]
From what I can see, we have no rules or even recommendations about whether quotes or use-examples should come first for a word that has both. I always put quotes on top, because to me they look better that way, but I can change this if most of us are doing it the other way. Thanks, —Soap—12:16, 28 July 2023 (UTC)[reply]
@Soap: WT:EL states usexes should "be placed immediately after the applicable numbered definition, and before any quotations associated with that specific definition", which is how I've seen it the vast majority of the time. Up to now I've changed it accordingly in the handful of entries I've seen where they were the other way round. —Al-Muqanna المقنع (talk) 12:31, 28 July 2023 (UTC)[reply]
Okay, thanks, I didnt see that sentence. Im surprised, though, because the inline formatting always displays as [quotes ▲] [usage examples ▲] even when the usexes are actually placed first. But it seems I've been using a gadget that allows the hiding of usexes, which most people probably don't have ... I had forgotten about that and don't even know where the code I added is. —Soap—13:06, 28 July 2023 (UTC)[reply]
I have further questions about 'collocations'. I had always thought of them as short usage examples, but they usually illustrate no novel syntactic information, and possibly showing only semantic linkage. (Are they just cliches?) Do they belong before or after usage examples that illustrate aspects of grammar? I think they belong before and probably deserve to be displayed two or more to a line. But is there even a maintainable bright-line distinction between collocations and usage examples? Collocations are mentioned once in the 'ordering of bylines' discussion below, but were not discussed at all. DCDuring (talk) 13:40, 28 July 2023 (UTC)[reply]
That much I knew. They seemed a special case of usage examples, but more to display semantic aspects of usage that syntactic ones. They are not much used, at least in English and Translingual var. taxonomy, so I don't have much basis for inferring more about them in practice. I have seen nouns that have NPs under definitions that display some cliched adjective use. @Ioaxxere expressed a definite opinion about their ordering and should be able to address this. DCDuring (talk) 14:09, 28 July 2023 (UTC)[reply]
This sounds about right to me; synonyms before usexes mostly, and within nyms, it's synonyms, antonyms and then everything else. Benwing2 (talk) 21:12, 25 July 2023 (UTC)[reply]
@Ioaxxere, Benwing2: Yes. When quotations of similar date are used to show how the word is used in alternative constructions, it makes sense to order them by similarity. Also, if the language has changed significantly during its history, it might be better to put the newer forms at the top. Would we really want archaic diction at the top? --RichardW57 (talk) 23:31, 25 July 2023 (UTC)[reply]
Not really a requirement - it says 'should'. I gave a reason not to do it. Besides, the document is not a formal policy. What are you going to do next - insist that modules and templates not be chiefly documented within the source code? --RichardW57m (talk) 11:10, 26 July 2023 (UTC)[reply]
Actually, most of the material I work with doesn't have a proper date, and I'm not really sure what event it should refer to. Does the date for a Pali word in Roman script go back to when the text was originally composed, or the date, millennia later, when it was published in Roman script? Ditto for Pali words in Thai or Lao script, with the problem that much of what I use doesn't actually have a meaningful date. On the other hand, the republication date is relevant for any development in spelling or text normalisation. RichardW57m (talk) 11:10, 26 July 2023 (UTC)[reply]
Hey, and what about web articles with no date? I generally order them with the other cites that have dates based on the date of the Internet Archive date. --Geographyinitiative (talk) 11:11, 26 July 2023 (UTC)[reply]
@Geographyinitiative: You can use a., c., and p. with an estimated date as with other published material, although I would question what undated web-only sources are eligible or useful for citations on Wiktionary. —Al-Muqanna المقنع (talk) 11:20, 26 July 2023 (UTC)[reply]
Thanks for your reply. Under my reading, undated web citations can be both eligible and useful for Wiktionary, although they probably don't add to the WT:ATTEST requirements. Undated government websites sometimes include uses of words that demonstrate the range of possible uses for a word. "Show the variety of contexts that a term is used in;" Wiktionary:Quotations#Choosing_quotations. Sometimes those undated materials are more valuable to show contexts of usage than the actually OCLC-able cites. --Geographyinitiative (talk) 11:25, 26 July 2023 (UTC)[reply]
@Al-Muqanna, Ioaxxere Here is an example of something that I recently added that has no date (as far as I see), but yet is valuable to me because it demonstrates local government official usage: diff. Take a look at this and I'd love to hear any comment you have. --Geographyinitiative (talk) 16:34, 26 July 2023 (UTC)[reply]
A quick search into the code reveals a "last updated" section, which says it was last updated on the 23rd of July of 2019. (For the record, I got this by Ctrl+Shift+I, and looking up "date".) cf (talk) 00:34, 27 July 2023 (UTC)[reply]
There are also second-hand sources, such as ill-attributed portions of manuscripts republished in other documents. The best one could do is date the manuscript by how long it could have lasted in a good state, which is of limited use if it's merely a copy. --RichardW57m (talk) 13:30, 26 July 2023 (UTC)[reply]
@Ioaxxere: So under the earlier to latest, should a quotation of a modern dictionary of canonical Pali come after a quotation of a commentary, even though canonical Pali is at least 500 years older? I noticed something close to this issue at jāṇu(“knee”). --11:42, 26 July 2023 (UTC) RichardW57m (talk) 11:42, 26 July 2023 (UTC)[reply]
I prefer having quotations above usage examples in some cases, because that way the user won't have to scroll to find them; I haven't been consistent with that though, but it is a valid opinion. Thadh (talk) 10:38, 26 July 2023 (UTC)[reply]
I generally prefer that, but mostly because most of the 'examples' I encounter are generally inferior materials or quotations that have come adrift from their context. --RichardW57m (talk) 11:34, 26 July 2023 (UTC)[reply]
I think I also have a tendency to put inferior quotations at the end more generally. Sometimes it is unclear whether the first elements of bahuvrihis, tatpurushas and the like are to be treated as words or as sums of parts, and combining forms hide information about inflection. I've looked through the Roman script Pali quotations, and found I rarely have as many as two quotations for the same script, so the effect on ordering will be small. --RichardW57m (talk) 14:46, 26 July 2023 (UTC)[reply]
I apologize for completely overlooking this thread and asking almost the same question up above. I prefer putting quotations on top, though as I've written at User:Soap#usex_vs_quote I can see merits in doing it the other way around. I don't want to be a pest, so I'll try to remember to put the usexes on top from now on, but I'd hope the bot will be able to automate this so I (or anyone else) don't have to go through hundreds of pages looking to see which of them have the order reversed. —Soap—14:11, 28 July 2023 (UTC)[reply]
One issue with putting use-examples above quotes is that the display will group them with synonyms and anything else using the #: formatting, so unless those things are also placed above the quotes, the formatting looks a bit awkward (to my tastes, anyway). See for example this noun section, where if it goes usex ---> quote ---> synonym from top to bottom, the formatting goes in, out, and back in again. Perhaps this is no big deal ... to me it's an aesthetic preference. But if a bot is going to run through our entries and have them put usexes above quotes unconditionally, it will lead to layouts like this. (And sorry, I should have read more closely ... I see now that this is indeed what we're proposing .... I guess I'll leave this up though just to show that it would be better to adopt the order shown up above as a whole, rather than uncoupling usex>quote from the synonyms and collocations.) —Soap—10:10, 1 August 2023 (UTC)[reply]
The ordering of quotations is not an issue for the imminent sorting - @JeffDoozan has just stated on the page with the bot request that quotations will not be sorted relative to one another. --RichardW57m (talk) 15:04, 26 July 2023 (UTC)[reply]
I’ve seen them listed in translation sections, and some of the language infrastructure is already set up (e.g. it knows which diacritics should and shouldn’t be in page names). Theknightwho (talk) 01:52, 27 July 2023 (UTC)[reply]
@Rodrigo5260, @Benwing2, @Theknightwho: I have no strong objection but no strong enthusiasm either. I added alt forms to Belarusian entries and translations myself and fixed some misspellings I noticed. Some enthusiasts of łacinka out there have little knowledge of the orthography rules, so just being cautious not to introduce rubbish. BTW, some sites I used to see Belarusian in both Cyrillic and Latin letters no longer display łacinka the way they did, so it's even harder to verify the correct spellings. Anatoli T.(обсудить/вклад)02:04, 27 July 2023 (UTC)[reply]
Also, there are variants of łacinka, such as the ones using cz for ч or others using č instead, and forms using w for /v/ and others using v. Rodrigo5260 (talk) 17:30, 29 July 2023 (UTC)[reply]
@Rodrigo5260: The differences between the relevant variants are minor and mostly cosmetic. The older books from the 19th century used to have sz, cz, w. The newer books from the 20th century replaced them with š, č, v, but these changes are evolutionary and not ambiguous. I would suggest to ignore the latest modernization attempt, which tried to introduce ĺ instead of ł, because to the best of my knowledge this variant wasn't ever used by any actual Belarusian paper books, it was designed to be ambiguous (the old łacinka word "stol" becomes стол instead of столь) and it had been already abandoned by its inventors even for its original street signs purpose. Ssvb (talk)
The most authoritative source for łacinka spelling is probably Biełaruskaja hramatyka dla škoł (1918).pdf written by Branisłaŭ Taraškievič himself (and available at be.wikisource.org). Together with a bunch of Belarusian paper books in łacinka, published somewhere between 1918-1933. There are older łacinka books and they are very interesting too, but the Belarusian grammar and łacinka wasn't formally normalized before that and some minor spelling differences could be spotted here and there. The "Naša Niva" newspaper also used łacinka in the beginning of the 20th century. Ssvb (talk) 10:35, 6 August 2023 (UTC)[reply]
I originally asked this question on the Discord and was directed here. So, I noticed we have Appendix:Capital letter but not an Appendix:Lowercase letter; is there a reason for this or could I create the latter? I presumed this is because lowercase writing is the default, but I have collected a few entries in my sandbox for such an appendix that IMO would justify having it. I am new to editing and so I want to make sure that no one would oppose to the creation of such an appendix or that there is a specific rule/reason it has not yet been created (i.e. not being necessary). Apologies in advance in case this is a silly question. Beat regards, LunaEatsTuna (talk) 02:29, 27 July 2023 (UTC)[reply]
The appendix namespace is very lax in what it allows. Wiktionary does not provide indiscriminate free hosting, but since what you're proposing has some linguistic content, it's legit in principle. And your question is not at all silly: it's completely appropriate. —Justin (koavf)❤T☮C☺M☯02:59, 27 July 2023 (UTC)[reply]
I suspect no-one bothered to create such an appendix because lowercase was felt to be the default used whenever the conditions set out for uppercase don't apply, but I think an appendix seems reasonable; there are interesting things that could be said about when lowercase is used. - -sche(discuss)05:38, 27 July 2023 (UTC)[reply]
Example: Should words like "disembarkment" and "disembarkation" go in "Category:English terms prefixed with dis-"? Should words like unlikable go Category:English terms suffixed with -able? Sometimes there is uncertainty about the order of derivation (as with disembarkation) but I think the majority of the time, or at least often enough, it is fairly clear which order makes the most sense, as in the case of unlikable, which (in its ordinary sense--the only one that we list) is derived by negating likable. While editing Latin, I've seen a few examples come up recently where I would have omitted the word from the prefixed category (e.g. inventiuncula) or suffixed category (e.g. inarticulatus). I'm not sure if we have a written policy on this somewhere. Looking at what current practice is, it looks like many un- words ending in -able are in both categories, but not all of them (e.g. unfindable is not (yet) in Category:English terms suffixed with -able). Urszag (talk) 18:35, 28 July 2023 (UTC)[reply]
Words like "impenetrable" bring up another issue: the question of how to treat formations that are relatively transparent but were not originally formed within the language. Its parts were already put together in Latin impenetrabilis, so diachronically, neither the prefix nor the suffix was added in English. In cases where the synchronic decomposition is still obvious and unproblematic, it makes sense I think to include it alongside the diachronic derivation. But there are cases that I'm not sure about like audacity and veracity; you could in principle treat them as derived nouns in -ity built on the adjectives audacious and veracious, but aside from not being diachronically accurate, the form of these words raises many tricky questions about how a derivation within the linguistic system of English would actually would work (e.g. should audac- and verac-, which exist nowhere as independent forms, nevertheless be described as real morphemes of English? What is their underlying phonological representation?). Regarding the concept of circumfixation: I think there is a tendency for negatives in un- to be more productively applied to adjectives that have one of the derivational endings -ed, -able, -ing than to simple adjectives, so I guess I can see the argument for seeing words like these as having a circumfix, but I feel like "circumfix" is a more complicated analysis so Occam's razor favors not using it unless it's clearly necessary to do so.--Urszag (talk) 22:41, 28 July 2023 (UTC)[reply]
We have {{surf}} = "surface analysis" for synchronic analyses of this sort. (I've always thought that it would be more accurate to say "synchronic" instead of "surface", but maybe this is linguistic jargon that most people won't recognize.) As for circumfixes, I suspect a lot of adjectives of the sort "unfooable" were formed before the corresponding "fooable" adjectives were formed, so diachronically the circumfix analysis is correct but I see your point that it is simpler to describe them as un- + fooable. My concern here is what happens if "fooable" is extremely rare or archaic (as often is the case)? Does this analysis still make sense? Benwing2 (talk) 23:04, 28 July 2023 (UTC)[reply]
@Benwing2: IMO it's still fine in principle to provide a surface analysis linking to a rare or archaic term, since I don't think a surface analysis or "equivalent to" has to be salient, but you'd have to consider case-by-case how useful it is to mention. —Al-Muqanna المقنع (talk) 23:10, 28 July 2023 (UTC)[reply]
@Al-Muqanna, Benwing2: I'm uneasy. I see 'surface analysis' as serving two rôles:
The interpretation of the form proceeds via that analysis.
The form may be recreated according to that mode of word formation.
The second may wither as the formation becomes more opaque.
Neither of those can be applicable if people don't know the base form or some other derivative, as in the literate audacious > audacity. RichardW57 (talk) 02:36, 29 July 2023 (UTC)[reply]
@Urszag: I don't think there's anything wrong or unusual about saying that the surface analysis of audacity is audacious + -ity. I did the same at negotiosity. Some entries (e.g. perspicacity, keeping with -ity) have chosen to mark off a stem, as perspicac(ious), but that seems unnecessary to me. Of course giving a surface analysis or "equivalent to" only makes sense where there is actually an obvious English "root"; for something like velocity it should be excluded (despite the weird veloce + -ity there atm). —Al-Muqanna المقنع (talk) 23:07, 28 July 2023 (UTC)[reply]
Shouldn't a surface analysis actually involve a correct morphological derivation? I don't see how one could get "audacity" from "audacious" without invoking a rule contrived specially for the purpose ("/-eɪʃəs/ changes to /-æs/ when suffixed with -ity"). Nicodene (talk) 03:40, 29 July 2023 (UTC)[reply]
@Nicodene I think User:Al-Muqanna's intention is that the ending -ious can be assumed to drop before -ity (I prefer making this explicit by writing perspicac(ious) + -ity). The rest follows by the normal English rules of derivational morphology (e.g. trisyllabic laxing). Benwing2 (talk) 03:55, 29 July 2023 (UTC)[reply]
The alternation of /s/ and /ʃ/ would remain to be accounted for. One could say that the ending -ious turns preceding /s/ to /ʃ/, but that would require truly reifying stems like */ˌpɜːspɪˈkeɪs-/. Perhaps we really could. Nicodene (talk) 04:38, 29 July 2023 (UTC)[reply]
I don't see that as so problematic; there are alternations like grace -> gracious that shows that -ious does indeed turn /s/ to /ʃ/. Benwing2 (talk) 04:52, 29 July 2023 (UTC)[reply]
Indeed not a problem if our surface-analysis equally derives 'perspicacity' and 'perspicacious' from a cranberry morpheme ending in /s/, rather than deriving the noun from the adjective. Nicodene (talk) 04:58, 29 July 2023 (UTC)[reply]
It's not a cranberry morpheme; if it's a morpheme in the strict sense it's a 'bound' morpheme, though it's actually a compound of two morphemes, and I think you'll find the written form is primary. Another derivative is perspex. --RichardW57 (talk) 12:52, 29 July 2023 (UTC)[reply]
I suppose one could still derive the noun from the adjective if one posits that the latter underlyingly consists of the aforementioned cranberry stem ending in /s/ + -ious, in which case removal of the suffix would prevent the change to /ʃ/.
The fact people coin nonce words like bodacity (which seems to meet CFI) from bodacious, or even the online-only hellacity from hellacious, suggests to me that there is a generally understood rule. —Al-Muqanna المقنع (talk) 08:29, 29 July 2023 (UTC)[reply]
The main problem I have with this is loss of internal structure. It's like breaking something into its smallest pieces and throwing them into a bucket to be fished out and inventoried later. While you can eventually figure out indefatigability from the entries for in- + de- + fatigue + -ability, it's much easier if you look at the entry for defatigable, which doesn't mean what you might think based on the common meanings of its component morphemes. Or compare with the approach at indefatigable. Chuck Entz (talk) 03:22, 29 July 2023 (UTC)[reply]
@Chuck Entz Yeah, I agree with you here, but I don't think User:Urszag's intention was to ignore the internal structure, rather to point out that there may be competing analyses of terms of the form prefix + root + suffix. Benwing2 (talk) 03:57, 29 July 2023 (UTC)[reply]
Is there any reason why reference templates shouldn't have language codes?
I made a couple of new templates to add into Category:Spanish reference templates. As you can see there, many of the templates start "R:es:" and some are "R:la:" or "R:mul:", but many are just "R:[foo]" (and even an "R:xaa" !). Is there any good reason why these "R:[foo]" ones shouldn't be moved to "R:[language code]:[foo]"? —Justin (koavf)❤T☮C☺M☯07:11, 29 July 2023 (UTC)[reply]
I think the only reason for not having a langcode is if there are multiple possible codes, e.g. it's a multilingual dictionary covering several different langs. Benwing2 (talk) 07:24, 29 July 2023 (UTC)[reply]
@Koavf: But it's an extra four characters to type! Someone's just gone to the effort to shorten script codes (and make them more difficult to remember). --RichardW57 (talk) 11:06, 29 July 2023 (UTC)[reply]
It is very useful to have these categorized by language, but why does that require that each use of the template bear the language code? I have not seem any rationale for requiring the extra keystrokes. There may be the exceptional cases such as when a two reference templates have the same name. But that is something that can be addressed by changing the name of one or both of the templates. DCDuring (talk) 13:48, 29 July 2023 (UTC)[reply]
It can make it easier to search for in the toolbar - typing just T:R: pulls up everything, typing up T:R:foo brings up just a few, often the one you need. Vininn126 (talk) 13:56, 29 July 2023 (UTC)[reply]
This is exactly it. Also the fact that many of the reference names are extremely obscure, because they've been shortened to very generic names. We should deal with the possibility of name conflicts proactively by separating them like this, and not reactively by doing it in a mish-mash way. Theknightwho (talk) 14:03, 29 July 2023 (UTC)[reply]
Human brains can handle mish-mash as long as they are not dragged into too many distinct unaccustomed realms.
It's not really a matter of data, though - it's a matter of accommodating needs. If the objection is because it's 3 or 4 more characters to type, I just don't see how that outweighs the obvious confusion caused by a mess of initials and acronyms. Theknightwho (talk) 18:55, 29 July 2023 (UTC)[reply]
I agree esp. when it comes to templates consisting of acronyms, although I'm not convinced we need to put 'mul' when there's no clear language code. Benwing2 (talk) 18:59, 29 July 2023 (UTC)[reply]
I think it would make sense to use family codes (where appropriate), which would probably cover a large number of these. If a source is truly multilingual (I can't think of many, but maybe something like Glottolog), then those could probably have mul.
The problem happens when a reference covers multiple languages from different families. I think if it is a bilingual dictionary between a common and uncommon language (e.g. Spanish-Mapudungun), the code should just indicate the uncommon language since that's the language whose entries will use the dictionary. But there are still cases like Template:R:Lindberg-Gerd which is Russian, Ukrainian, Belarusian but also Hungarian, Romanian and Armenian. Benwing2 (talk) 19:42, 29 July 2023 (UTC)[reply]
In most cases, this will be an improvement- but not always. If there are any cases where there is no obvious language code and no obvious family code, using "mul" strikes me as making stuff up so we can shoehorn a name into the standard format. Using "mul" for something that's not really translingual will make it harder to find things, not easier. Chuck Entz (talk) 16:30, 29 July 2023 (UTC)[reply]
I even commonly use some of those and forget what they mean. Sorting by language would also make it obvious from things like Special:AllPages or the dynamic search in the search bar or search results, etc. —Justin (koavf)❤T☮C☺M☯19:47, 29 July 2023 (UTC)[reply]
I fully support the use of language-by-language namespacing, but not the use of R:mul: for multilingual references. I'd prefer if the R:mul: names were used for references pertaining to areas that we consider translingual, like ISO standard symbols and taxonomic names. The "base" R: namespace would be left over for multilingual or non-lexicographic references. This, that and the other (talk) 04:53, 30 July 2023 (UTC)[reply]
The BP is difficult of access for newer users because of the redlinked template. I offer up voter suppression as a label because it is an often-used inflammatory meme in the US. DCDuring (talk) 20:33, 30 July 2023 (UTC)[reply]
I guess "all single-language reference templates should be changed to 'R:[code]:[foo]'" and then a discussion of multi-language ones can be separate. If it fails, the status quo remains, which is a hodge-podge. —Justin (koavf)❤T☮C☺M☯18:31, 30 July 2023 (UTC)[reply]
BTW to me there seems to be a consensus (bar DCDuring) that tagging single-language reference templates is OK, but multi-language ones need more thought and should maybe be left (as noted by This, that and the other). Benwing2 (talk) 18:48, 30 July 2023 (UTC)[reply]
The point would be memorializing that this is best practice, encouraging standardization now and in the future, and maybe documenting it on a guideline or policy somewhere, so that going forward, we don't have a mish-mash of names, but something standard and guessable without context. Maybe even semi-automated category sorting and generation due to the language codes. —Justin (koavf)❤T☮C☺M☯18:50, 30 July 2023 (UTC)[reply]
Not every policy needs a vote. If there’s ample consensus and it’s not changing something like CFI (and it’s something already in major practice), then there’s no explicit need to have a vote which imho can just complicate things a lot. If the vote failed what would even change? It just doesn’t seem worth it here. AG202 (talk) 23:52, 30 July 2023 (UTC)[reply]
A consensus of a narrow group is not a consensus of Wiktionary. After the recent unpleasantnesses on this page, fewer folks than ever are participating in discussions. If we are going to call something a policy it needs a vote. If we don't have a vote on something, than not acting in accord with that is not a violation of policy and can have no consequences such as blocking etc, removal of powers etc. DCDuring (talk) 02:53, 31 July 2023 (UTC)[reply]
It should be clear to everybody that the templates are being grouped in a particular way without writing down a formal policy. After all, the R: convention is hardly documented either, yet people see the pattern, and learn from it.
The technical cognescenti can force a change of this kind, as they have pushed for many changes without demonstrated consensus. DCDuring (talk) 02:56, 31 July 2023 (UTC)[reply]
It makes sense because of neutral point of view, i.e. if you don’t associate a particular language with a template, that’s the “reason” sought in the topic title. The important things else have already been pointed out: That it is insufferably uniconic to have template names like T:R:RBE, “R:DCF, R:GGC, R:XYZ, R:PDQ etc.”, typing laziness should also have its limits, and mul stays associated with a separate concept of a Translingual lexicon rather than multiple languages. Overall the naming scheme seems largely iconic. The result of my scheme is already known since I have created like a fourth of all existing reference templates, so I have not answered for DCDuring’s demonstration in this thread before, while the reasonings are appeals to intuition that people are unwont to formulate, and PUC already hinted that Dan Polansky’s bad intuition inflicted contrary practice and thus made reason overall less transparent. I don’t see a consistent solution indeed that now reference templates should consequentially follow a scheme like R:langcode:—it seems to contravene intuition but nonetheless appeals to intuition in other respects, as outlined. Different results appeal to intuition depending on the context in which it is applied. Fay Freak (talk) 17:43, 2 August 2023 (UTC)[reply]
Hello, on 呼#Japanese we have a minor problem with this kanji, in that the {{ja-kanjitab}} is placed on its own line, separated by two others, like this:
... but this causes extra newlines to form in the output, presumably because the template doesn't have any inline content and its placement has the same effect as multiple newlines in a row. In order to avoid this, one could do it like this:
The only problem is this puts the kanjitab (on my screen) quite a bit above where it actually refers to the concrete reading こ. Also, thinking about it, this associates the kanji tab with the entire kanji, whereas it only actually applies to the actual Japanese word. Is that all still okay...? Kiril kovachev (talk) 19:48, 29 July 2023 (UTC)[reply]
@Kiril kovachev Does the table refer only to the POS labeled Affix? In that case it should go directly after the ==Affix== header. In general, right-aligned templates such as Wikipedia boxes go either directly after an ==Etymology N== header or a POS header, before any content (or at least that's where I put them). Benwing2 (talk) 19:53, 29 July 2023 (UTC)[reply]
@Benwing2 Yes, that's right. That seems more suitable, and I went and moved it under "Affix" now. This case is an outlier for me because there's usually an etymology header for it to attach to, but in this case there's not. Thanks for your help, Kiril kovachev (talk) 19:58, 29 July 2023 (UTC)[reply]
Strictly speaking, we could just add an etymology header to this entry, couldn't we? Either providing whatever is known about the etymology or saying the etymology is unknown. Is the issue that the logical place (AFAICT) for the Etymology header in this case would be below the L2 language header, and you want the kanjitab box to be lower than that? - -sche(discuss)23:09, 29 July 2023 (UTC)[reply]
I agree with @-sche here -- the standard MO for {{ja-kanjitab}} on single-kanji entries is that the template should go just under the ===Etymology=== heading. If there is no such heading, we add one.
Also, most single-kanji entries will have multiple ===Etymology X=== headings, as most single kanji in Japanese have multiple readings. The {{ja-kanjitab}} details are tied to reading, so each template call should go in its appropriate etym section. ‑‑ Eiríkr Útlendi │Tala við mig19:36, 31 July 2023 (UTC)[reply]
Etymology sections including debunking false etymologies
Following from our sister project q:, where we include misattributed and spurious quotations because they are commonly reproduced and just lacking these "quotations" could be seen as a lack of information, I wanted to ask everyone here about false information and how to debunk it entries. We don't want to spread false information, but including these false ones and flagging them as false helps increase genuine knowledge. I included two references to false folk etymologies in marmalade and pumpernickel just now for the same reason. I believe that these are consistent per Wiktionary:Etymology#Folk_etymologies. The reason I'm soliciting feedback here is because on en.wq, we offset these false quotations as a way of making it clear that this is misinformation, but it seems like we've chosen a preference here for really downplaying false information so that it doesn't overwhelm reading accurate info. Do we have any thoughts on best practices for how to include folk/false etymologies? If we don't want to offset them somehow with sidebars or background colors, how do we keep them from visually blending in with legit etymologies? Just creating a new paragraph? Simply relying on readers to know the difference from actually reading it and context? I'm asking to make sure that I'm not inserting any information that would inadvertently spread more misinformation. —Justin (koavf)❤T☮C☺M☯04:49, 30 July 2023 (UTC)[reply]
I think what you did on those pages is reasonable. I definitely think we should include false etymologies if they're commonly found in Internet sources (otherwise people will tend to add them as real etymologies), but of course no need to go out of our way to include them if they're not common. Benwing2 (talk) 05:02, 30 July 2023 (UTC)[reply]
I agree, if a folk etymology is very commonly mentioned, it makes sense to explicitly reject it; we also do that on e.g. fuck (for the backronyms). I think we can just use normal text, as you did (whether set off in its own paragraph or not); I don't think we need a different background colour (which seems like it would actually draw attention). Obviously another reason to mention folk etymologies is if they influenced the form or meaning of the word, e.g. buttonhole, belladonna, shamefaced. - -sche(discuss)18:41, 30 July 2023 (UTC)[reply]
Yes. Ideally we wouldn't need to mention them, but it is helpful from the editing perspective to stop IPs re-adding them. (The fake acronyms are the worst! Oh sure "dog" comes from "dirty old growler".) Equinox◑18:43, 30 July 2023 (UTC)[reply]
What about a template and tracking category? Standardize the language with something like {{false etyl|en|[text]}} resulting in something like "A common false or folk etymology for "[term]" is [text]" and a tracking category like "Category:en:Terms with false etymologies"? Or is that too much? —Justin (koavf)❤T☮C☺M☯18:47, 30 July 2023 (UTC)[reply]
@Koavf Dunno. Seems maybe low-return, esp. since the number of potential false etymologies is not really limited in any fashion; anyone can make up a false etymology. If we are to include it maybe it should be renamed Category:English terms with common false etymologies (NOTE, this should be a POS category not a topic category) but then you run into the issue of how to define "common". (OTOH we have Category:English terms with unknown etymologies, which I think includes those with "uncertain" etymologies, which is also potentially problematic by definition.) Benwing2 (talk) 19:41, 1 August 2023 (UTC)[reply]
We could make some kind of rubric like "has to have at least two citations from mainstream publications" or "at least one citation explicitly calling it a common folk etymology", but as you point out, I don't think it's going to be that common of an issue of someone adding wildly obscure false etymologies. —Justin (koavf)❤T☮C☺M☯19:45, 1 August 2023 (UTC)[reply]
Systematically remove the label "simile" from definition lines
Our current practice is (most of the time) to put {{lb|xx|simile}} at the beginning of definition lines.
While it is definitely useful to gather similes in a category (Category:English similes, etc.), I don't see the point of displaying "(simile)" prominently like that, and I've come to think it looks a bit silly.
The only use case I can see is for people who barely know anything about the language of the entry that they're looking at, but catering to their needs should not be done at the cost of making things look silly for everyone else.
Therefore, I think this label should be systematically removed and replaced by {{cln|xx|similes}} at the bottom.
However, since it is convenient to just type {{lb|xx|simile}}, I'm not proposing to ban its use, but to introduce periodic bot sweeps to replace all new instances of it. PUC – 10:23, 30 July 2023 (UTC)[reply]
What do you mean "silly"? What's silly about labelling similes? Also, is there a difference between a phrase of "as" or "like"; and a simile? Our definition of simile is "A figure of speech in which one thing is explicitly compared to another, using e.g. like or as.", where a "figure of speech" is "A word or phrase that departs from straightforward, literal language." Compared to instances where there is no figurative use of the comparison being made, there is arguably a distinction between a comparison used as a simile—a figure of speech or literary device—and the same phrase conveying a plain idea. This would be like those sections in which there's both an idiomatic definition and a sum-of-parts one that is given its own SoP definition to distinguish it from its idiomatic use; I propose the same thing for "similes" if there is a need.
Additionally, labelling can be a teaching mechanism for people learning using Wiktionary. I don't agree that anything is made silly, I think it's helpful information, like "transitive", "uncountable", or whatever else. Kiril kovachev (talk) 11:31, 30 July 2023 (UTC)[reply]
The difference is that it is not possible to determine whether a verb is transitive or a noun is countable just from looking at the word, but similes are by definition immediately obvious as similes. I'm inclined to agree with PUC. As far as "figure of speech" goes, I think we are currently missing a sense there comparable to turn of phrase, which is currently listed as a synonym: similes are "figures of speech" in the sense of being distinctive expressions, not usually in the sense that they depart from the meaning of the words involved. —Al-Muqanna المقنع (talk) 11:47, 30 July 2023 (UTC)[reply]
"What's silly about labelling similes?": I've just explained that above: it's stating the obvious. It is obvious that swear like a trooper is a simile; transitivity, on the other hand, isn't obvious.
"Compared to instances where there is no figurative use of the comparison being made, there is arguably a distinction between a comparison used as a simile—a figure of speech or literary device—and the same phrase conveying a plain idea. This would be like those sections in which there's both an idiomatic definition and a sum-of-parts one that is given its own SoP definition to distinguish it from its idiomatic use; I propose the same thing for "similes" if there is a need." I don't understand what you mean by that. Please give some examples of what you have in mind.
"labelling can be a teaching mechanism for people learning using Wiktionary": please elaborate. Also, as I said, I'm not proposing to ban its further use, just to systematically replace them. PUC – 11:50, 30 July 2023 (UTC)[reply]
"teaching machanism" Well, more a learning mechanism. Labels are visible to someone not admitted to the cult of Wiktionary cognescenti.
'stating the obvious' IMHO we do that a lot. Why single this kind of obviosity out?
BTW, what are the obvious signs that something is a simile? Are they obvious to an English learner? Are they infallible? DCDuring (talk) 21:19, 30 July 2023 (UTC)[reply]
@PUC For your second bullet point: I'm referring to entries like $100 hamburger, in which there is both a figurative and literal sense. I may be misunderstanding something, but I see a big difference between entirely non-figurative comparisons (black as coal, cut like a knife) and cold as charity, drunk as a wheelbarrow, which are definitely figurative and idiomatic. Whether this difference actually affects whether it's a "simile", who knows, I may be stretching. In the same way that we mark the second sense of $100 hamburger, however, we could do the same for many similes whose meaning is obvious and sum-of-parts. That is what I mean.
For your third point, I mean when you see "simile" associated with a term, that tells you that that is the format of similes in that language. Each language has its own manner of creating similes; thus with gros comme une maison, only someone with sufficient French knowledge could discern that it's a simile by looking at it. (Bear in mind that the English translation, as in this case, may not be a simile itself.) By stating the obvious, it makes it easier to spot the pattern for those who are learning a language using Wiktionary, and in the above case you would know that similes in French can be identified with the word "comme". This is what I mean by "teaching".
I have no serious grievance whatever the outcome of this proposal; even if we were to wipe the label altogether I wouldn't hate it. I'm just arguing because I think there's worth in the label as it is, and we could make something good out of it by using it judiciously. Kiril kovachev (talk) 08:57, 31 July 2023 (UTC)[reply]
Support. One could add all kinds of superfluous labels repetitive due to explanation in the actual sense line or etymology, but in general we have the sound principle of avoiding duplication, we would have a specific reason to shirk, as readers can be numbed by information overload. It just seems a variation of our old “(idiomatic)” figleaf label to signify to editors that it according to the entry editor it is “not SoP”, this may be the idea of it being a teaching mechanism. Fay Freak (talk) 12:41, 30 July 2023 (UTC)[reply]
Yes, it would be kept, and populated through adding "[[Category:English similes]]" or "{{cln|en|similes}}" at the bottom of the entry. Then again, I'm not proposing that we ban the "{{lb|en|simile}}" mark-up completely; users who want to use it because it saves them some typing would be allowed to. It's just that it would be in a state of permanent deprecation, so to speak: all new uses would be replaced by bot, and there would be no replacement in the other direction (from "[[Category:English similes]]" to "{{lb|en|simile}}", that is). PUC – 18:40, 30 July 2023 (UTC)[reply]
@Sgconlaw, DCDuring Pinging a couple of interested parties. Currently if there is more than one author, they are formatted like this:
2025, John Doe, Title, etc.
2025, John Doe; Richard Roe, Title, etc.
2025, John Doe; Richard Roe; Sally Schmoe, Title, etc.
2025, John Doe; Richard Roe; Sally Schmoe; Natalie Noe, Title, etc.
I'm thinking of maybe changing this as follows:
2025, John Doe, Title, etc.
2025, John Doe and Richard Roe, Title, etc.
2025, John Doe; Richard Roe; and Sally Schmoe, Title, etc.
2025, John Doe; Richard Roe; Sally Schmoe; and Natalie Noe, Title, etc.
Separately, I'm also planning on handling multiple authors (semicolon-separated or comma-separated) in the |author= param, to avoid the need to specify |author=, |author2=, |author3=, etc. (This is already frequently the case, and this would just bless this usage while formatting it nicely.) This needs a bit of smartness to recognize cases where someone has written |author=Doe, John, but I think this is doable. Benwing2 (talk) 21:23, 30 July 2023 (UTC)[reply]
(e/c) BTW I also implemented smart handling of numeric params such as |page=, |volume=, |line=, |column=, |issue=, etc. so that if you write e.g. |page=7–10 you get pages 7–10 instead of page 7–10, as well as supporting, for each numeric param, e.g. |issue=, |issues=, |issue_plain= and |issueurl=. (Currently some params support some of these but with no consistency.) Benwing2 (talk) 21:28, 30 July 2023 (UTC)[reply]
i.e. I'm fine with this, I just need to know which scheme and how to format it (and whether you think we should completely change what we have to match APA or whatever, or just change the author handling for now). Benwing2 (talk) 21:29, 30 July 2023 (UTC)[reply]
This just seems like 5 extra character-widths. The and has the very modest advantage of indicating when the author list is about to come to an end, though a change of typeface to italics also does so. Isn't a list of authors always followed by some typographic marking of a change, either italics or quotation marks? DCDuring (talk) 21:38, 30 July 2023 (UTC)[reply]
@DCDuring The answer is actually no. I found several cases where the author list is not set off by italics or quotation marks, e.g.:
a.2003, John Ahier, John Beck, Rob Moore, quoting Harriet (a Cambridge University student), Graduate Citizens?: Issues of Citizenship and Higher Education[26], Routledge, published 2003, →ISBN, page 114:
I mean a lot of the money that obviously goes into universities and their libraries and their facilities and their academics and stuff but I mean I haven’t had a very quality degree to be honest. I think the quality of my education has been crap . . .
2004, Vance M. Thompson, MD, edited by J. Kevin Belville and Ronald J. Smith, LASIK Techniques: Pearls and Pitfalls[27], SLACK Incorporated, →ISBN, page 187:
For one I wanted to have what I considered a very quality tracking device.
In both of these, the authors are set off only by commas, leading to potential ambiguity. In the first one, the three authors (BTW all stuffed into a single comma-separated entry in |author=, which I will parse and split) are followed by a quotee. The second is even worse: One author (specified using |author=Vance M. Thompson, MD) and two editors (specified using |editors=J. Kevin Belville and Ronald J. Smith), and it's near impossible to tell who's the author and who the editors. Parsing and as a separator is tricky because it leads to a lot of false positives; I'm thinking I will do an offline parse run for cases where |author= (and maybe |editors=) uses and or &, and convert them (with manual filtering to remove the false positives) to use semicolons, which can then be parsed and split online, when displaying the {{quote-*}} template. BTW once the online author-splitting code is ready, it will be able to split on both semicolons and commas, but it's better to specify multiple authors using semicolons because there are various checks to weed out false positives when comma separators are given (e.g. the common 'LAST, FIRST' format and weird cases like |author=United States House Committee on Science, Space, and Technology), which may result in your multi-author param not getting properly split. Benwing2 (talk) 07:04, 1 August 2023 (UTC)[reply]
Semicolons would help for most and are probably good enough for us. I wish my eyesight hadn't degraded to the point where I often can't distinguish a period from a comma or a colon from a semicolon. Sometimes I miss punctuation altogether. But, my eyesight is better in the morning: Why are there two commas(?) after "2003" in the first example above? DCDuring (talk) 13:15, 1 August 2023 (UTC)[reply]
Oh, good point. {{ante|2003}} automatically adds bolding and a comma. This is probably to reinforce its intended place of usage (in quotes), but I find that even when manually using it in typing up an untemplated quote, people (even me) have for years added their own commas after it (and now the quote templates do), not expecting that the template itself generates a comma, leading to two commas. Personally I would be tempted to make the ante template stop adding a comma after a bot run to add commas after any existing uses that aren't already followed by commas, but I suppose someone might prefer to figure out how to make the quote templates strip / not add an extra comma if comma-adding {{ante}} is used, and fix existing (manual, etc) uses that are followed by extra commas, instead. - -sche(discuss)13:26, 1 August 2023 (UTC)[reply]
@-sche Oops I didn't notice the extra comma. IMO definitely the templates {{ante}}, {{circa}} and {{post}} should not be generating an extra comma. I'm going to change this and do the bot run you mentioned. There are about 350 uses of {{ante}}, 2000 of {{circa}} and only 40 or so of {{post}}; not too many to fix up. Benwing2 (talk) 21:04, 1 August 2023 (UTC)[reply]
Academic referencing systems generally start the reference with either the first author's surname or a sequential number. We need to start with the year, so even if we were to adopt a formal academic referencing system (noting that these are typically specified in copyrighted, paid manuals), we still would need to adapt it for local use. This, that and the other (talk) 03:41, 31 July 2023 (UTC)[reply]
Regarding |page=7–10, you may have thought of this already, but some books restart page numbers per section (I have seen textbooks and car manuals do this), so "page 7-10" is a single page, and after "page 11-9" comes "page 12-1", etc. Perhaps you are accounting for this by only the n-dash version "7–10" rendering as "pages": will the hyphen version display "page 7-10"? To be clear, whatever the answer to that is, I do think it's reasonable to make |page=7–10 display "pages", I just think anything we do has downsides. If the en dash has different behavior from the more commonly typable hyphen, that could be unintuitive to or unnoticed by people who will probably keep writing "page=7-10" (meaning "pages") however often they presently do ... but doing nothing (leaving either symbol rendering as "page 7-10") is probably wrong in most cases, as people probably mostly mean "pages" ... but rare cases where a single page "7-10" was meant also exist ... but then, any such cases are probably already unclear: someone looking at "page 7-10" will not know, without consulting the book itself, whether it was meant to be a range or a single page. I suppose it is our fate as a dictionary with citations to hundreds of thousands of books in hundreds of languages, to run into every edge case. - -sche(discuss)02:06, 31 July 2023 (UTC)[reply]
@-sche Hmm, I didn't realize there are such things. We can handle something like "page 11-9" by noticing that the numbers are decreasing, but for "page 7-10" that won't work, of course. I'm thinking we should probably make |page=7-10 go ahead and display pages 7-10 (or maybe pages 7–10 with an en-dash); there's now (or rather, there will be, once the code goes to production) support for |page_plain=, and so you could always write |page_plain=page 7-10 if you really wanted that. Benwing2 (talk) 02:17, 31 July 2023 (UTC)[reply]
Yeah, I agree with you. I suspect (although I suppose someone could try to quantify this) that in almost all instances of people typing page=7-8, they do mean a range, so making the template display it as a range ("pages 7-8", whether with a hyphen or a dash) is probably the best course of action, as long as there's a way for people to force "page 7-8" in the rare instance where that's the actual number of the singular page. (I would almost be tempted to suggest putting some kind of explanatory footnote in the style of {{SIC}} after such page numbers, precisely because people are otherwise liable to think it's meant to be a range.) Here is a tractor manual which uses that scheme: in the table of contents you see that the first page of each chapter is 1-1, 2-1, 3-1, etc, and inside the manual, page "4-2" is followed by "4-3". Here is another vehicle manual page 7-2, and here is a ?textbook? page 3-1. (I seem to recall Richard or someone else mentioning encountering a Thai(?) book that did something like that, too.) - -sche(discuss)03:41, 31 July 2023 (UTC)[reply]
@Benwing2 I think the handling of page/pages needs more refining as it converts to the plural in a bunch of quotes where it shouldn't, e.g., when {{gbooks}} contains a hyphen (slo-mo, zhing-zhong, peer pressure), when the page numbering contains a hyphen (bowhuntress, playgroundful, shoulder verb sense 6) or in other cases such as in essere (verb senses 4, 5) or bowswoman (1879, 1889 quotes). I would say writing pages instead of page is more misleading than the other way around so I'm not sure if it is worth keeping this functionality unless it is possible to eliminate and prevent all these false positives. (Maybe it would be safer going through quotes that potentially need fixing in a semi-automated manner and changing page= to pages= when appropriate.) Einstein2 (talk) 11:17, 2 August 2023 (UTC)[reply]
@Einstein2 Most of these are easily fixable by tightening the criteria for displaying pages. In particular, I need to change it to extract the displayed text from links and to only display pages if there are numbers on both sides of the hyphen (and also beware of HTML entities like [). The only one of the examples you give that isn't easily fixable is the one in shoulder verb sense 6, where the page number is 12-81. We actually discussed this particular case up above and decided it wasn't worth throwing away the auto-handling given the existence of |page_plain=, given how many people don't bother to correctly use |page= and |pages= (in both directions). I would rather not go the "fix the existing quotes in a semi-automated fashion" if possible, both because of the large number of them (238,000+) and the fact that it won't help going forward. It is possible to make the pages algorithm sensitive only to en-dashes and em-dashes rather than hyphens, as a workaround for a page number like 12-81, but I suspect that will result in a very large number of cases incorrectly saying page, because few people can be bothered to correctly enter en-dashes instead of hyphens. An alternative that can also be done is to provide a special syntax to indicate that a page number is compound, e.g. something like |page=!12-81; this reduces the typing compared with |page_plain=page 12-81. I could then imagine a semi-automated bot run to convert existing cases of compound page numbers to use this syntax; but this would require an ability to automatically find most of the relevant cases in a semi-automated fashion, so that I could avoid having to manually scan thousands of potential cases (suggestions for how to do this?). Benwing2 (talk) 21:21, 2 August 2023 (UTC)[reply]
@Einstein2 I implemented a better algorithm. I also wrote a script to check for |page= values of the form ###-### with a hyphen. Among all the quote templates I downloaded a few days ago, there are 861 such cases. The following are the first 20:
Page 97 trade: Saw possible compound page '391-2': {{quote-journal|1=en|year=1950|author={{w|W. H. Auden}}|title=A Playboy of the Western World: St. Oscar, The Homintern Martyr|journal={{w|Partisan Review}}|url=http://archives.bu.edu/collections/partisan-review/search/detail?id=284002|page=391-2|text=In a homosexual of this kind—corresponding to the test of eccentric behavior in the drawing-room—one usually finds a preference for "'''trade'''," i.e., sexually normal males, because, if another homosexual yields to him, he is only one of a class, but if he can believe that an exception is being made in his case, it seems a proof that he is being accepted for himself alone.}}
Page 113 head: Saw possible compound page '156-64': {{quote-journal|en|year=1968|author=Fred Davis|coauthors=Laura Munoz|title=Heads and freaks: patterns and meanings of drug use among hippies|journal=Journal of Health and Social Behavior|volume=9|issue=2|page=156-64|passage=The term, "'''head'''," is, of course, not new with hippies. It has a long history among drug users generally, for whom it signified a regular, experienced user of any illegal drug—e.g., pot "head," meth "head," smack (heroin) "head."}}
Page 261 above: Saw possible compound page '206-7': {{quote-journal|en|year=2013|month=May-June|author=[http://www.americanscientist.org/authors/detail/william-e-conner-1 William E. Conner]\n|title=[http://www.americanscientist.org/issues/feature/2013/3/an-acoustic-arms-race An Acoustic Arms Race]\n|volume=101|issue=3|page=206-7|magazine={{w|American Scientist}}\n|passage=Earless ghost swift moths become “invisible” to echolocating bats by forming mating clusters close (less than half a meter) '''above''' vegetation and effectively blending into the clutter of echoes that the bat receives from the leaves and stems around them.}}
Page 263 abovedeck: Saw possible compound page '5-44': {{quote-book|en|year=1997| title=Gunner's Mate |author=Jim Bomar | page=5-44| passage=These heating elements allow the gun mount to continue to operate when the '''abovedeck''' temperature is as low as 40° C.}}
Page 333 absolution: Saw possible compound page '60-1': {{quote-journal|en| month=August| year=1919| journal={{w|The Smart Set}}| page=60-1| author={{w|H. L. Mencken}}| passage=The true aim of medicine is not to make men virtuous; it is to safeguard and rescue them from the consequences of their vices. The physician does not preach repentance; he offers '''absolution'''.}}
Page 404 accipient: Saw possible compound page '459-460': {{quote-journal|en|year=1905|month=October|work=The Lutheran Quarterly|author=J. W. Richard|title=The Old Lutheran Doctrine of Free-Will|volume=35|page=459-460|text=“{{...}}God is the author of salvation, Free will only is susceptible. None but God is able to give it; none but Free will is able to lay hold of it. Therefore it is given by God alone, and given to Free will alone. On the one hand it cannot exist without the consent of the '''accipient''', and on the other hand not without the grace of the giver.{{...}}”}}
Page 645 NATO: Saw possible compound page '305-306': {{quote-book\n|en\n|year=1971\n|last=Johnson\n|first=Lyndon\n|authorlink=Lyndon Johnson\n|chapter=Strengthening the Atlantic Community\n|title={{w|The Vantage Point}}\n|url=https://archive.org/details/vantagepointpers00john/\n|publisher={{w|Holt, Reinhart & Winston}}\n|ISBN=0-03-084492-4\n|LCCN=74-102146\n|OCLC=1067880747\n|page=305-306\n|pageurl=https://archive.org/details/vantagepointpers00john/page/305/\n|text=What concerned me most about De Gaulle's decision was that it threatened the unity of '''NATO''', which had been so carefully developed over two decades. '''NATO''' was essential to the security of Europe and the United States. I was convinced that the stronger and more unified we were, the more incentive the Soviets and their Warsaw Pact allies would have to work with us in solving outstanding problems and differences.}}
Page 720 food: Saw possible compound page '72-3': {{quote-journal|en|date=2013-06-29|volume=407|issue=8842|page=72-3|magazine={{w|The Economist}}| title=[http://www.economist.com/news/international/science-and-technology/21580133-how-microbes-promote-liver-cancer-overweight-punch-gut A punch in the gut]| passage=Mostly, the microbiome is beneficial. It helps with digestion and enables people to extract a lot more calories from their '''food''' than would otherwise be possible. Research over the past few years, however, has implicated it in diseases from atherosclerosis to asthma to autism.}}
Page 726 world: Saw possible compound page '206-7': {{quote-journal|en|year=2013|month=May-June|author=[http://www.americanscientist.org/authors/detail/william-e-conner-1 William E. Conner]\n|title=[http://www.americanscientist.org/issues/feature/2013/3/an-acoustic-arms-race An Acoustic Arms Race]\n|volume=101|issue=3|page=206-7|magazine={{w|American Scientist}}\n|passage=Earless ghost swift moths become “invisible” to echolocating bats by forming mating clusters close{{...}}above vegetation and effectively blending into the clutter of echoes that the bat receives from the leaves and stems around them. Many insects probably use this strategy, which is a close analogy to crypsis in the visible '''world'''—camouflage and other methods for blending into one’s visual background.}}
Page 740 add: Saw possible compound page '72-3': {{quote-journal|en|date=2013-06-29|volume=407|issue=8842|page=72-3|magazine={{w|The Economist}}\n|title=[http://www.economist.com/news/international/science-and-technology/21580133-how-microbes-promote-liver-cancer-overweight-punch-gut A punch in the gut]\n|passage=Mostly, the microbiome is beneficial.{{...}}Research over the past few years, however, has implicated it in diseases from atherosclerosis to asthma to autism. Dr Yoshimoto and his colleagues would like to '''add''' liver cancer to that list.}}
Page 938 CDC: Saw possible compound page '64-65': {{quote-book|en|title=The Aging Brain|author=Timothy R. Jennings|year=2018|ISBN=9780801075223|page=64-65\n|passage=A 2009 multistate '''CDC''' survey off over seventy thousand adults found that more than 35 percent of people slept fewer than seven hours per night.}}
Page 1162 perpendicular: Saw possible compound page '112-3': {{quote-journal|en|date=2012-03|author={{w|Henry Petroski}}|title=Opening Doors|volume=100|issue=2|page=112-3|magazine=[[w:American Scientist|American Scientist]]|url=http://www.americanscientist.org/issues/pub/opening-doors|passage=A doorknob of whatever roundish shape is effectively a continuum of levers, with the axis of the latching mechanism—known as the spindle—being the fulcrum about which the turning takes place. Applying a force tangential to the knob is essentially equivalent to applying one '''perpendicular''' to a radial line defining the lever.}}
Page 1193 zeal: Saw possible compound page '143-144': {{quote-book|en|1779|{{w|David Hume}}|{{w|Dialogues Concerning Natural Religion}}|part=12|page=143-144|url=http://name.umdl.umich.edu/004895521.0001.000|passage={{...}} the highest '''zeal''' in religion and the deepest hypocrisy, so far from being inconsistent, are often or commonly united in the same individual character.}}
Page 1201 skew: Saw possible compound page '6-70': {{quote-book|en|author=Uday A. Bakshi|author2=Mayuresh V. Bakshi|chapter=Three Phase Induction Motors|title=Electrical Machines|location=Pune, Maharashtra|publisher=Technical Publications Pune|year=2009|page=6-70|pageurl=https://books.google.com/books?id=-3l0ikD15x4C&pg=RA2-SA6-PA70|isbn=978-81-843-1696-4|passage=Thus asynchronous torques cannot be avoided but can be reduced by proper choice of coil span and by '''skewing''' the stator or rotor slots.}}
Page 1308 stem: Saw possible compound page '206-7': {{quote-journal|en|year=2013|month=May-June|author=[http://www.americanscientist.org/authors/detail/william-e-conner-1 William E. Conner]| title=[http://www.americanscientist.org/issues/feature/2013/3/an-acoustic-arms-race An Acoustic Arms Race]| volume=101| issue=3| page=206-7| magazine={{w|American Scientist}}| passage=Earless ghost swift moths become “invisible” to echolocating bats by forming mating clusters close (less than half a meter) above vegetation and effectively blending into the clutter of echoes that the bat receives from the leaves and '''stems''' around them.}}
Page 1382 moon: Saw possible compound page '308-309': {{quote-book\n|en\n|year=1737\n|author=John Brickell\n|title=The natural history of North-Carolina\n|page=308-309\n|passage=They number their age by '''Moons''' or Winters, and say a Woman or a Man is so many '''Moons''' old, and so they do with all memorable Actions in life, accounting it to be so many '''Moons''' or Winters since such or such a thing happened.}}
Page 1577 however: Saw possible compound page '72-3': {{quote-journal|en|date=2013-06-29|volume=407|issue=8842|page=72-3|magazine={{w|The Economist}}\n|title=[http://www.economist.com/news/international/science-and-technology/21580133-how-microbes-promote-liver-cancer-overweight-punch-gut A punch in the gut]\n|passage=Mostly, the microbiome is beneficial. It helps with digestion and enables people to extract a lot more calories from their food than would otherwise be possible. Research over the past few years, '''however''', has implicated it in diseases from atherosclerosis to asthma to autism.}}
Page 1739 ON: Saw possible compound page '4-5': {{quote-journal|1=se|year=2019|author=Astrid Helander|title=-Divvun ja Giellatekno reaidu livčče horbmat ávkkálaš min olbmuide maid|journal=Ávvir|url=http://urn.nb.no/URN:NBN:no-nb_digavis_avvir_null_null_20190429_12_81_1?page=3|page=4-5|text=Mannan gaskavahkku lágidii Norgga beale Sámediggi ovttasráđiid UiT «Divvun ja Giellatekno» ossodat diehtojuohkinčoahkkima buot eamiálbmogiidda '''ON''' Eamiálbmogiid áššiid Bistevaš Forum 18. čoahkkimis.|t=Last Wednesday, the Sámi Parliament of Norway organized an information meet for all indigenous peoples in collaboration with the University of Tromsø's "Divvun and Giellatekno" section at the 18th session of the '''UN''' Permanent Forum on Indigenous Issues.}}
Page 1807 from: Saw possible compound page '72-3': {{quote-journal|en|date=2013-06-29|volume=407|issue=8842|page=72-3|magazine={{w|The Economist}}\n|title=[http://www.economist.com/news/international/science-and-technology/21580133-how-microbes-promote-liver-cancer-overweight-punch-gut A punch in the gut]\n|passage=Mostly, the microbiome is beneficial. It helps with digestion and enables people to extract a lot more calories '''from''' their food than would otherwise be possible. Research over the past few years, however, has implicated it in diseases from atherosclerosis to asthma to autism.}}
Page 1880 slug: Saw possible compound page '16-115': {{quote-book|en|title=Lee's loss prevention in the process industries| page=16-115| author=Sam Mannan| coauthors=Frank P. Lees| year=2005| passage=Another phenomenon investigated was a '''slug''' of water falling through the cloud.}}
It looks like about 80-90% of them are really multiple pages while the rest are single compound pages. Sometimes it's clear which is which, e.g. '308-309' is likely to be multiple pages, same with '72-3'. '156-64' is probably multiple pages, while '16-115' is probably a compound page since the topic is industrial processes (industrial books, handbooks, manuals and the like seem to favor compound pages). 861 is few enough that I might be able to manually go through them and mark up the compound pages. Benwing2 (talk) 22:24, 2 August 2023 (UTC)[reply]
┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ Hiya, @Benwing2, I’m a little confused by the behaviour of the updated quotation templates. You did a bot run which replaced the punctuation in a list of editors from commas to semicolons. However, the output of such a lost uses commas. In that case shouldn’t editors type commas rather than semicolons in quotation templates? Also, where there is one editor the quotation templates display “edited by” followed by the editor’s name, while if there is more than one editor “, editors” is displayed at the end of the list. Why don’t we standardize on “edited by” in both cases? — Sgconlaw (talk) 05:29, 20 August 2023 (UTC)[reply]
@Sgconlaw I can explain my thinking. The discussion has been split in various places but there was a general preference for (a) commas instead of semicolons to separate authors, and (b) "edited by" and "translated by" to show editors and translators following the author(s), to make it clearer where the separation occurs. I am using semicolons to separate authors in the template input because commas are ambiguous due to the existence of embedded commas (a) in names such as "Sammy Davis, Jr." and "Doogie Howzer, M.D." and numerous other such suffixes; (b) institutional authors like "Department of X, Y and Z"; (c) names with professional titles such as "David Jeremiah Barron, Circuit Judge, United States Court of Appeals for the First Circuit" (a real example); (d) inverted names such as "Froude, James Anthony"; and (e) misc identifiers such as "Arthur Kornberg, witness". The bot run deals with this when splitting on commas by having a long list of suffixes that it recognizes as joining with the preceding word, as well as having various deny-listed (i.e. blacklisted) words, and also rejecting one-word names except in certain cases (which deals with inverted names), and also rejecting names with lowercase words like "of" and "witness", except for certain allow-listed (i.e. whitelisted) words such as "von" and "de". This logic is too complex and fragile to do on the fly. None of these problems exist with semicolons. The only trickiness with semicolons comes in cases like cases like Oliver Optic [pseudonym; William Taylor Adams] as well as in HTML entities like Peter Christen Asbjørnsen. The code deals with this by parsing balanced brackets and parens and ignoring semicolons within them, and also recognizing HTML entities and not treating the semicolon at the end of them as a delimiter.
As for "edited by Mary Bloggs" vs. "Mary Bloggs, editor", this is not triggered by whether there's one or two names but by whether there's a preceding author. So if the book has two authors James Anthony Froude and Sammy Davis, Jr. and two editors Mary Bloggs and Roger Taney, it will show as
2025, James Anthony Froude, Sammy Davis, Jr., edited by Mary Bloggs and Roger Taney, Crimes and Misdemeanors, New York: Holt, Rinehart and Winston
but if it has no authors and only editors, it will show up as
2025, Mary Bloggs, Roger Taney, editors, Crimes and Misdemeanors, New York: Holt, Rinehart and Winston.
The logic behind this is that we need the "edited by" to precede the editors when there are preceding authors to ensure that the division between authors and editors is clear, but without the authors, it looks strange (to me at least) if we begin by saying
2025, edited by Mary Bloggs and Roger Taney, Crimes and Misdemeanors, New York: Holt, Rinehart and Winston.
It gets even weirder if we have a chapter preceding the book, have only editors and use the "edited by" format because then we get
2025, “Presidential Malfeasance”, in edited by Mary Bloggs and Roger Taney, Crimes and Misdemeanors, New York: Holt, Rinehart and Winston,
whereas using the current logic, we get
2025, “Presidential Malfeasance”, in Mary Bloggs, Roger Taney, editors, Crimes and Misdemeanors, New York: Holt, Rinehart and Winston.
If you disagree with these aesthetic choices please let me know and give some alternative suggestions; I am totally open to reconsideration of this. Benwing2 (talk) 08:22, 20 August 2023 (UTC)[reply]
@Benwing2: It looks good to me from what I've seen so far, thanks. My only quibble is that I think "et al." should not generate a comma if there's only one named editor listed. —Al-Muqanna المقنع (talk) 09:17, 20 August 2023 (UTC)[reply]
@Sgconlaw Yes, I am updating the documentation currently to describe inline modifiers and indicate which fields can take them, and which fields (e.g. author fields) can take multiple semicolon-separated entities, each with their own inline modifiers. Benwing2 (talk) 19:38, 20 August 2023 (UTC)[reply]
@Benwing2: Another edge case—specifically for |translator=anonymous (or Anonymous) it might be better to render it as "anonymous translator" than "translated by anonymous". I think this is specific to translators since I can't imagine cases where you'd list anonymous editors or need special phrasing for other stuff. —Al-Muqanna المقنع (talk) 21:15, 20 August 2023 (UTC)[reply]
@Al-Muqanna OK sure. This is somewhat similar to |page=unnumbered (and similarly for |volume=, |issue=, |line= and other numbering parameters), which as a special case is written "unnumbered page" rather than "page unnumbered". Benwing2 (talk) 22:25, 20 August 2023 (UTC)[reply]
Hello, today I wrote a bot script which traverses the contents of rechnik.chitanka.info (downloaded locally as a word list) and can append anagrams to Wiktionary's Bulgarian entries. Would this be a good idea to deploy across the whole current Wiktionary, over all ~120k words that are available in the wordlist? (Of course, if we don't currently cover the word on Wiktionary, that page can't be updated; so in reality there are much fewer additions to be made. Further, the number of unique alphagrams I found was only 3944 total. The number of entries affected would be at most 8407, but many of these don't exist and so the number would be yet lower.)
The code can be found at the repository if you want to inspect—word list is included as well. I also ran a number of tests:
...you can also see all the others from Special:Contributions/KovachevBot, the above being just a selection. I ran around 40 or so and they all looked good so far (barring the first two and a subsequent minor error, which are now fixed).
Considerations:
If a page already has anagrams on it, I extend the template with any more that I have available locally; if there are no new ones to add, nothing is changed. Contrarily, if it has no anagrams at all, the whole section is generated and placed (hopefully) correctly. (This system of appending to the template assumes there is only one anagrams template in the entry; if there are none, despite an anagrams section, then nothing is changed. I think this is a safe assumption, because anagrams are not widely employed in Bulgarian editing so far and I reckon no one is using more than 1 {{anagrams}} even if they do.)
I'm using @JeffDoozan's technique of checking for categories at the end of the page, ensuring that the anagrams section isn't placed after them. You can see this succeeding at the first diff, on вол.
Like my previous script I am making unified diffs locally which can be used to revert any major errors, en masse if needed.
Also: the corpus from rechnik.chitanka.info is very large, but some of the terms are exceedingly rare and some have no definition entered whatsoever. For such cases, do we want to include words only if they have a known definition? This might cut down on the number of anagrams this method can include a lot, though.
Also, an example of the dubious anagrams being added is дом, where мод(mod) is added; this is hardly a Bulgarian word and I can't find a single explanation of its meaning. The obvious guess is just a borrowing of Englishmod. There are lots of terms like this that are entered on Chitanka. It is entirely possible to use a different wordlist, sourced only from words that are defined; it would just be a matter of re-generating it. Kiril kovachev (talk) 21:47, 31 July 2023 (UTC)[reply]
You mention putting anagrams at the end of the page. Do you actually mean language section? If not make sure to put them in the appropriate section. See also User talk:DTLHS#Anagrams vs. categories for some code I have that checks for and splits off categories at the end of a language section (along with the trailing separator, which is no longer present).
Instead of just keeping whatever anagrams are there already, you might want to check to make sure they're really anagrams and remove them if not.
How are you normalizing entries? E.g. removing spaces, hyphens and other punctuation and lowercasing? You might want to check with User:DTLHS (who is not very active anymore but used to do anagram runs for various languages) to find out what their procedure is for normalizing titles to produce anagrams.
As for including terms that don't exist in Wiktionary but are found in Chitanka, I would indeed try to exclude the really rare ones (e.g. without definitions).
Normally for anagrams we include non-lemma forms; I'm not sure if you're doing that currently but if not you should probably integrate the existing Bulgarian non-lemma entries with your script (and maybe even consider auto-generating the ones that don't exist, but that might take some work; note that the inflection modules I've written have a way of extracting the inflections from a given lemma and inflection specs, but in the older modules, including these, it's in a custom non-JSON format; it might be a good idea to have them return JSON, which is very easy; see Module:es-verb for how to do this).
Yes of course, my mistake, I misspoke there. At the end of the language section is where it's going.
Sure, that seems like a constructive feature, and ought not to take much effort to make work; the only thing is I doubt there'd be (m)any instances where that'd apply, but why not. Thanks for the idea ^^
Good point, I made as good as no effort to normalise anything; I think case-folding and removing non-alphabetic text would be the main point. (For some reason my current dataset has no multiword or hyphenated terms, might need to check that out.)
Indeed, it's probably best to avoid including terms with no definitely-known definitions. At least if they're given some kind of definition, then we'd have some confidence that the word is legitimate. Only that this method would not be able to detect when the definition is just a redirect to an empty word, but I figure we can just tank this loss as a minor edge case.
The word list already includes non-lemmas in its database, if that's what you mean? However, could you please elaborate on what you mean by auto-generating the nonexisting ones? How can I integrate the module with the local script I'm running? Also, if you remember we previously generally had opposition to generating Bulgarian inflections. I assume it's fine if it's just for providing links for the anagram section.
I'll check out that module tomorrow, but thanks for your feedback! Also, if we are happy to have this run, do I need to start a new vote to have the task approved? Kiril kovachev (talk) 22:19, 31 July 2023 (UTC)[reply]
The only reservation I have is about the inclusion of non-lemma forms. Bulgarian has a rich inflection, particularly for verbs (think Spanish verbs). My personal POV is that I'd rather bias the anagram section towards being amusing and informative, rather than towards exhaustiveness for its own sake. I'd imagine that for Bulgarians, non-lemma anagrams wouldn't be that interesting, and for learners of the language, it would present a potentially complicated slice of Bulgarian grammar without the context to make it make sense. Chernorizets (talk) 22:39, 31 July 2023 (UTC)[reply]
@Kiril kovachev There's no need for another vote. The vote was for allowing you to have a bot and the task was presented as an example of what you might do, as a sanity check. (I've probably done over 1,000 bot runs by now for various tasks with my bot; it would be impossible if every one required a vote.) What I mean by autogenerating the non-existing ones is going through the set of Bulgarian lemmas that have inflection tables, calling into the appropriate module code (Module:bg-verb or Module:bg-nominal) to get the list of all inflections, and adding those to the word list used to generate the anagrams. That way you can potentially have non-lemma forms in the anagram list without having to actually add any non-lemma entries to Wiktionary. In general if you're using pywikibot, you can expand a Wiktionary template using e.g. site.expand_text("{{bg-ndecl|дом<bd>|json=1}}", title="дом"), assuming the module is changed to support the |json= parameter; if you're interested in doing this, I can make the necessary module changes to support JSON output of the inflections. Benwing2 (talk) 22:48, 31 July 2023 (UTC)[reply]
@Kiril kovachev, Chernorizets That is fine, I am not wedded to the idea of including non-lemma inflections in anagram lists. This is what is normally done for Spanish, Italian and such, I think on the theory that people who are interested in anagrams (e.g. for crossword puzzles?) might want these forms included, but we can do what makes the most sense. Benwing2 (talk) 22:51, 31 July 2023 (UTC)[reply]
@Chernorizets; @Benwing2 If we've decided that we aren't doing non-lemmas, then don't feel obliged to make any module changes; thanks for that code sample, though, I didn't know such a thing was possible. I believe that'll be helpful for other endeavours quite shortly.
Now, I've re-generated the word list from the database (pushed if you want to see), excluding all non-lemmas (that I could filter out): even before my filtering, verb forms were already excluded from the dataset (they're part of a separate table in the database), and I further removed noun plurals and all undefined words. With this I believe we should be good, but any further thoughts/requests? I'll do some more test runs to ensure things are still okay. If we later decide we want non-lemmas, then that can be arranged as well anyway. Kiril kovachev (talk・contribs) 10:06, 1 August 2023 (UTC)[reply]
I know virtually nothing about Bulgarian, but I do think nonlemma forms should be included in lists of anagrams. Anagrams are an orthographic function, not a grammatical one, so whether a word is a lemma or nonlemma is irrelevant. Also, what would you do in the event that ABCD is a lemma but BDAC is a nonlemma form? Only list BDAC as an anagram of ABCD but not vice versa? Not list either one on the other's page? Listing both is really the only thing that makes sense to me. —Mahāgaja · talk14:31, 1 August 2023 (UTC)[reply]
@Mahagaja I'm not strictly against non-lemma forms, and I can understand the argument for them. I'm just a fan of incremental changes that add value. Wiktionary today doesn't have many non-lemma forms for Bulgarian, which is perhaps something we should rectify long-term, but what's significantly more pressing is that we're missing a ton of lemmas, even frequently-used ones. In my mind, an initial pass of the anagram bot could be a nice way of surfacing lemmas that people like myself should be adding.
This isn't a one-way decision - we could enhance the bot to include non-lemma forms in subsequent passes. Right now, even if we had that ability, there are few entries for non-lemma forms to add anagrams to. So you'd anyway end up overwhelmingly adding anagrams to lemmas, and having lots of red links for non-lemma forms we're unlikely to get to in the near term.
As I mentioned originally, this is just my POV. I won't stand in the way of including non-lemma forms if that's the consensus.
@Mahagaja To answer the practical question, non-lemmas have currently been excluded from consideration altogether, meaning neither lemmas will have non-lemma anagrams added, nor vice-versa. That's currently, though; I also don't have anything strictly against adding non-lemmas, but as Chernorizets pointed out, the utility of including these may be slightly tenuous now, at a time when our coverage even of lemmas is already quite sparse. The edifying function of providing valuable redlinks for us to gradually fill in is one potential benefit of excluding lemmas in this particular run. The issue is, I fundamentally agree with your opinion, and want virtually the same all-out anagram approach as we have in English, but the vast majority of anagrams created like this would be redlinks. I don't hate it-those would still be valid anagrams, after all-but what do you say to the idea of adding them later down the line? This idea doesn't conflict with yours: we will definitely add nonlemmas either way. The timing is the only contentious thing, I think. Kiril kovachev (talk・contribs) 00:58, 2 August 2023 (UTC)[reply]
I hadn't grasped that the plan was to include redlinks in anagram lists. I don't think we do that for any other language, do we? —Mahāgaja · talk06:34, 2 August 2023 (UTC)[reply]
@Mahagaja I guess not. If we were to include non-lemmas, but not ones that aren't already entered, then that could work just fine. Sorry, I didn't think of that at first... what do you think of that? @Chernorizets also what are your thoughts, because this is a good counter-point to the redlink issue. Kiril kovachev (talk・contribs) 12:55, 2 August 2023 (UTC)[reply]
@Kiril kovachev it's funny you bring this up, because I've been wanting to add more anagrams to Bulgarian entries, and I was thinking of writing my own local (non-bot) application to help me during editing :) I'll take a closer look at the bot code sometime later today (in my timezone).
A few questions:
how often do you envision the bot to run? As you mention, there are still quite a few Bulgarian headwords missing on Wiktionary, so I'd imagine that as people add more headwords that have anagrams, it would be a good idea to re-run the bot. The problem is that, unless we're adding Bulgarian headwords at great pace, the bot is mostly going to be a no-op in subsequent runs, as most anagram opportunities will have been realized in the initial run. I don't know OTOH how much of a problem that is.
what about a template that can auto-gen anagrams - if present - or throw an error if no anagrams are there? I'm not proposing that instead of the bot, but rather in addition - for the use case of incremental additions after the bot's first run.
Personally, I'd err on the side of being conservative and not include anagrams that don't have definitions in the dictionary. Consider the task of an editor who might want to create entries for some of those anagrams - they'd have to consult additional dictionaries to produce a decent References section. We can add more anagrams over time if we've grown the Bulgarian lexis on Wiktionary enough. Just my 2c.
Thanks again for being an innovator in Bulgarian language support on this site!
@Chernorizets Thanks for your quick answer, this time it's my turn to be about to go to sleep :') but I will answer for now: I did think about how often to set it running, but in truth we don't update the site ultra rapidly, so I could set it off every month or something like that (I would also put it on a cron job on my server, but I need to get the IP unblocked first...); I figure it wouldn't take overly long to run, but I could also have a local list of entries which have already been handled so we don't re-iterate the same entries: just the newly-created ones.
Also, I'm interested in what you mean by a template that can generate anagrams? I thought that that wasn't possible, which is why I made a bot for it, since I just can't figure out how to get the template to know what anagrams could possibly exist (we would have to check what other articles have the same alphagram, at runtime, and that seems very expensive, right?). If you have an idea I would love to hear it though. I'm personally stumped, and I saw this is how it's done on English entries as well, which is another reason I went about it like I did.
It sounds about right to avoid the undefined terms; I may, just for our curiosity, count up how many articles would be made with and without the exclusion; what's more is we could also leave the excluded ones on a big list, which we could then manually approve or reject.
Yeah, not sure if it's possible to create a template to autogenerate anagrams; that would require autogenerating all N! permutations and checking each one to see if it exists and has a Bulgarian section. Benwing2 (talk) 22:10, 31 July 2023 (UTC)[reply]
@Benwing2 I was thinking of something simpler - using a word list (e.g. the one from Chitanka) as module data, and writing a thin layer of code on top to generate the anagrams of a given word. A putative template like {{bg-anagrams}} would then invoke the module, using the page title as input, and expand to an invocation of {{anagrams}}. It sounds straightforward in my head, but I'm new so I might be missing some complications. Chernorizets (talk) 22:24, 31 July 2023 (UTC)[reply]
Yeah this is possible, although it might not be worth it to implement this, as new lemmas aren't being added that often. Benwing2 (talk) 22:52, 31 July 2023 (UTC)[reply]
More problematically, it wouldn't be symmetric - anagrams added to one entry wouldn't cause that entry to appear in other entries' anagram lists. I retract the template idea. Chernorizets (talk) 00:11, 1 August 2023 (UTC)[reply]
@Chernorizets If you index the module data by alphagram, wouldn't this symmetry issue be sorted? Each entry would generate the alphagram from its own letters, which would map to a list of all words with that alphagram by indexing the module data. In order to generate that page's anagrams we just exclude the entry itself, and then invoke the anagrams template with the rest. I still prefer statically generating them in advance, but this might still work if we wanted to do it. Kiril kovachev (talk・contribs) 10:12, 1 August 2023 (UTC)[reply]
since, by code inspection, the second arg is indeed just the alphagram corresponding to the title. It's also a good reminder that optional typing in Python is, erm, optional :) If you want to get "compile-time" warnings/errors for type mismatches, you could use something like MyPy. Chernorizets (talk) 01:47, 1 August 2023 (UTC)[reply]
@Chernorizets Whoops, thank you for the spot. I was originally using that type, i.e. a convoluted list of (letter, frequency) pairs instead of alphagrams as the dictionary key. Hence why it was called "counted" in various places in the code, because I was counting up all the letters in order to form the key, etc., but alphagrams made way more sense and were simpler to calculate... Also thanks for the MyPy suggestion. I'll give it a try, perhaps it can eliminate "errors" like these in the future :x
Additionally, I forgot to push this earlier, but I also included code for saving the backups (diffs) as mentioned above, and for normalising words (i.e. lowercasing them, removing spaces, etc.) which brings the alphagram count up to 5426. Kiril kovachev (talk・contribs) 08:47, 1 August 2023 (UTC)[reply]
@Ioaxxere Sure can do. I take it the best word list to base the anagrams off would just be the Wiktionary data set itself? I could maybe download one of the dumps and get all the English words from there. This would also ensure all the anagrams linked to are already entered in and have definitions, etc. Thanks for the idea, I'll get back to you these days when I have it set up Kiril kovachev (talk・contribs) 10:08, 1 August 2023 (UTC)[reply]
@RichardW57m Do you mean what decides whether two words are anagrams? In my script I treat words as anagrams if they have the same numbers of letters, ignoring everything that isn't alphabetic (one of the 30 letters of the Bulgarian alphabet). (Case is ignored as well, so А=а as expected.) I guess this excludes terms with numbers in them, but I've tested this and it applies to no entries in the dataset. Do you have any other suggestions for improvements to this definition? Kiril kovachev (talk・contribs) 14:15, 1 August 2023 (UTC)[reply]
@Kiril kovachev: It doesn't seem right that your script is the definition. You've already highlighted a potential issue with digits. The moist improtant question was the status of 'Й'. Why do you exclude 'Ѣ' and 'Ѫ'? Didn't Bulgarian have anagrams before 1945? Do words have to be lemmas to be considered anagrams? If not, does Bulgarian inflect logograms? I believe English has distinct plurals for nouns written as logograms, and if Bulgarian does they could come unstuck with the definition in your reply. --RichardW57m (talk) 15:05, 1 August 2023 (UTC)[reply]
@RichardW57m With respect, I believe you're over-thinking it. Here are my thoughts:
What do you suggest we should do about the definition of an anagram? Is there a place where anagrams are "defined" in English? To me it seems pretty trivial, any letters that are possible in a Bulgarian word (like those we enter on here) is a possible anagram component.
What do you mean about й? Unless I've made an omission, it is part of the list of letters in the script. It is properly accounted for, like every other letter.
The characters you mention are not available in the source I'm using, unfortunately. We also don't cover almost any terms that use the old orthography... thus nothing will be missed out by this omission. I guarantee you zero anagrams will be found even if I included this as part of the definition and script.
Please forgive my ignorance, but what are logograms in this context? At any rate, we appear to have decided that non-lemmas will be excluded from the generation for now. They can be included later if we decide, but @Chernorizets has suggested we leave them out for now. The anagrams may be more informative if we keep out the less essential, and more trivial anagrams formed by non-lemmas.
You are right, maybe there are improvements that can be made to the definition, and this could also be formulated somewhere formally. But none of these problems will fundamentally affect this generation. Even if we have numerous omissions (although, as I've mentioned, due to the source being used, we aren't affected by the orthography question nor due to non-lemma omissions), the script will never generate incorrect anagrams, and anything that's left out from this run can be filled in later. For example, we can add in non-lemmas later, take in old-orthography headwords as part of an expanded list, etc. If you'd like I can crawl our coverage of old-orthography terms and add them to the data too before I run anything.
But why not let it churn for now? I estimate a good 80% of the anagrams we could want will already be covered by this basic effort. The rest will surely require some more rigor, which I just think is superfluous.
The logograms I was thinking were the one using in maths, such as ∠(“angle”) and for ○, which I'm sure I've seen inflected in English. If you compare such words with the symbol deleted, one will get nonsense. Perhaps we haven't got round to documenting such thing in Wiktionary, and perhaps they don't exist in Bulgarian. --RichardW57m (talk) 11:48, 2 August 2023 (UTC)[reply]
@RichardW57m Fair enough, we could take a different approach to anagrams, then. Currently, I delete all letters that aren't alphabetic, but really this could be re-formulated: normalise words not by blanketly removing non-alphabetic characters, but just various features that are not significant, e.g. spaces, hyphens, punctuation, etc. The problem is that this creates more complicated rules, e.g. what is the full list of punctuation marks we need to remove?
For the record, I've tried the following schemes:
Remove absolutely all non-alphabetic characters
Remove all non-alphanumeric characters (as above, but keep numbers significant)
Remove all whitespace, as well as the characters [- . ; : ? ! ‒ – — ]
...and they all produce the same number of anagrams (3274) altogether. What I'm trying to get at is, there appears to be no real difference no matter which (sane) definition we use for anagrams. What definition do you think is principally the most sound?
@Kiril kovachev: I think you're using a database that excludes the difficult cases. I also think the boundaries are culturally specific. The following count:
Letters
Symbols
Numerals
Letter-like punctuation
There need to be rules on when letters are not distinguished.
The following don't count:
Sentence punctuation (this may be cultural)
Diacritics that don't distinguish letters (and that includes stress-marks, very relevant for Bulgarian inflected forms extracted from inflection tables).
Line-breaking controls
I think that many of the rules are going to be culture-dependent. I find it hard to accept that "'eard" and "read" are anagrams in English, but Wiktionary currently declares them to be anagrams. I also find it hard to accept "face" and "café" as English anagrams. --RichardW57m (talk) 15:38, 2 August 2023 (UTC)[reply]
Good thought, I will check what BG Wikipedia and other sources call an anagram. I doubt it will differ greatly, but we should be certain. On the other hand, this is English Wiktionary, so I argue in principle we should keep to the spirit of anagrams as we know them in English. That seems like the way that'd cause the least confusion. Kiril kovachev (talk・contribs) 17:36, 1 August 2023 (UTC)[reply]
I have seen some very weird claimed anagrams in other languages, e.g. Thai, but there I'm not even sure that Thai really has the concept of anagrams in Thai. --RichardW57m (talk) 15:05, 1 August 2023 (UTC)[reply]
@RichardW57m Bulgarian doesn't have its own definition for what an anagram is - it's the conventional meaning of rearranging the letters of one word to make one or more other words. There are no restrictions beyond that.
Bulgarian is written using the Bulgarian Cyrillic alphabet as of its 1945 orthographic reform, which among other things excluded certain obsolete Cyrillic letters. The fact we have a small handful of examples of pre-1945 orthography on Wiktionary should have no bearing on the anagram discussion - nobody under the age of 85 knows how to write in the old orthography, and even educated young (< 45 y.o.) people today occasionally have a hard time reading it. So it's not a practical concern, but perhaps a fun academic one.
There's no issue with "й", because an anagram has to be a valid spelling of a Bulgarian word (or wordform) using the original letters. So even though one could rearrange май(maj, “May”) to йам which is a homophone of ям(jam, “to eat”), it's not a valid anagram, and we don't consider ям(jam) an anagram of май(maj) - the letters don't match.
That's because in English and French, letters with diacritics are considered variants of the base letter, not separate letters (thus in French, soupçon is alphabetized exactly the same as if it were soupcon). Is that true in Bulgarian as well, or are й and и considered two different letters? —Mahāgaja · talk10:09, 2 August 2023 (UTC)[reply]
@RichardW57m this doesn't apply to Bulgarian. The letter "й" represents IPA /j/, whereas the letter "и" represents IPA /i/. The example you provided is an invalid anagram in Bulgarian, because the two letters are distinct, denoting a consonant and a vowel, respectively. Chernorizets (talk) 10:12, 2 August 2023 (UTC)[reply]
@Chernorizets, CitationsFreak: That is what we discussed yesterday. A general purpose bot would have to know what count as letters, and, going further afield, what count as the same letter, and possibly when. In that respect, "й" is an issue, but a simple one - it just has to be counted as a separate letter to "и". --RichardW57m (talk) 10:43, 2 August 2023 (UTC)[reply]
@RichardW57m I've already explained that an anagram of word X in Bulgarian is a permutation of the letters of word X that is also a word (a lemma or a non-lemma form). There are no special rules as to what constitutes a letter, and Kiril's bot's code does what it needs to do. Please have a look at Bulgarian alphabet for further questions. Chernorizets (talk) 20:03, 2 August 2023 (UTC)[reply]
The requirement that an anagram be a word is a significant simplification. Or did you mean term? Do remember, though, that for inflected forms, we need a definition of what is to be stripped off from the forms in the tables. I suspect the database will exclude terms with digits. (Or does Bulgarian actually lack non-translingual terms with digits that satisfy CFI?)
The bot does what it does, which I think would be good enough for a first pass, but as our coverage of Bulgarian expands, it will probably require revision of its operational definition of 'anagram'. --RichardW57m (talk) 09:08, 3 August 2023 (UTC)[reply]
I assume it's the same, "a rearrangement of letters (or letter-like-things)". The only real question is what is considered a letter. cf (talk) 19:09, 1 August 2023 (UTC)[reply]
Just to chime in, I'm also in favour of including non-lemma forms. Statistically, anagrams aren't actually that common outside of very short words, and Bulgarian has comparatively few inflections compared to many languages (though admittedly more than English). As such, I don't think the inclusion of non-lemmas would flood the anagram sections in a problematic way. Theknightwho (talk) 13:06, 2 August 2023 (UTC)[reply]
At least Bulgarian seems not to have the problem that not all lemmas with a normal part of speech consist of words. (Languages whose lemmas are stems may have that issue.) RichardW57m (talk) 14:44, 3 August 2023 (UTC)[reply]
@CitationsFreak: Depends on what you mean by 'rearrangement'! In our English anagram entries, CAFE, cafe and café are all anagrams of face and vice versa, but CAFE, cafe and café are not anagrams of one another! This subtle exclusion feels right to me. --RichardW57m (talk) 13:50, 3 August 2023 (UTC)[reply]
@Kiril kovachev @Benwing2 I'm going to summarize my position here, FWIW, since the discussion has gone a bit into the weeds, and people unfamiliar with the Bulgarian writing system insist on there being complicating factors that simply aren't there.
I like the idea of a bot to add anagrams to Bulgarian entries. Expanding it to cover other languages is scope creep.
I like the bot's approach of using a set wordlist to determine anagrams. In Bulgarian, an anagram is simply a permutation of the letters that yields a grammatically correct word. The approach naturally lends itself to expanding the scope of the bot in the future, if necessary, by growing the wordlist as appropriate.
I have reservations about including non-lemma forms as anagrams initially, but I'm happy to defer to Kiril on that.
The bot code I reviewed was simple and easy to maintain, and I support it being the first iteration.
@Chernorizets Thanks for your comments. I agree that there aren't any complicating factors involved in this; in general it's not a complex process. It's only User:RichardW57 trying to over-complexify this (which BTW is par for the course). Yes, the bot needs to have a small amount of language-specific knowledge about Bulgarian (specifically what counts as a letter), but that's it. Benwing2 (talk) 21:25, 2 August 2023 (UTC)[reply]
@Chernorizets Thanks for your summary. I think it's time we agree on what to do. Re everyone who's given their input here: @Benwing2 @CitationsFreak @Ioaxxere @Mahagaja @RichardW57m @Theknightwho: thank you all for your feedback on this question, it's been great to hear your thoughts. Given that it will be primarily I, Chernorizets and other Bulgarian editors who will get to see most of these changes, I think it's a good idea to first follow what we believe to be best. I'm neutral on the matter, but Chernorizets has asked that we exclude non-lemmas for the time being, which I would like to respect. I believe it would be best to give it some time and evaluate the utility of anagrams on entries before deciding when to do the rest.
The only thing I'm still wondering is: are non-created entries allowed as anagrams?
If so: there is good sense in using a fixed list of lemmas, because this provides editors exposure to redlinks which can lead to connecting many of the project's entries when these are created.
If not: I don't see what reason there is vis-a-vis the above considerations (that anagrams are an "an orthographic function, not a grammatical one") to prefer lemmas besides that they may be more interesting to readers (this is a legitimate consideration; there is also the aforementioned use by crossword/Scrabble players, though).
This is basically the final consideration before the bot can run in full. Overall, once this is sorted, if you don't mind, I would be happy to launch this running on just lemmas for now at any rate. Kiril kovachev (talk・contribs) 14:42, 3 August 2023 (UTC)[reply]
The following is just my opinion; I don't think we have a policy, or even a set of pleas. For a first pass in what I think is the not very incremental assembly of lists of anagrams, I think red links should be acceptable. The red links should be eliminated by creating the entries, but I suspect that won't be faster than any more of the red link elimination. Words that fail CFI by their very nature will require modification of the database, but I see a potential problem with phrases. I hope that phrases that fail CFI won't be generated, but I'm not certain that they won't be.
Problems will arise if editors add inflected forms and the next bot-driven update does not support them.
I'm seeing lots of awkward edge cases to resolve. I've just taken a good look at the Bulgarian verb conjugation tables, and processing them seems more difficult than I expected. There appear to be some phrases that inflect, and handling them would not be fun. I've not tackled inflected phrases for Pali inflection tables; it seemed to need a lot of mark-up. For Bulgarian it might even be simpler to create all their inflected forms as non-lemma entries, though I don't like that solution, and it might be argued that some of the inflected forms do not meet CFI, but are instead sums of parts. --RichardW57m (talk) 15:09, 3 August 2023 (UTC)[reply]
I think it's a generally good idea to generate derived forms automatically, but I don't think I'm up for doing it in the same breath as generating anagrams; if possible I'd like to keep them separate, because of the issues you mentioned, such as what if the declension is entered wrong, as well as the matter of deciding what is a non-lemma anyway (now the task would expand to checking pages to see if they have declensions, which is just too bloated for what is supposed to be an anagram job).
What we know for certain with the current data set is that every word is well-defined (so it should easily pass CFI) and it is a relatively frequent word, regardless of whether we currently have it here or not. Therefore I feel fine, be it with non-lemmas included or not, running it on the data we have. Kiril kovachev (talk・contribs) 19:19, 3 August 2023 (UTC)[reply]
I would say include nonlemmas, and red links to Bulgarian words you know pass CFI. (And even redlinked Bulgarian nonlemmas.) cf (talk) 15:29, 3 August 2023 (UTC)[reply]
For English, at least: include nonlemmas but not redlinks. I also think we should automatically create entries for inflected forms. Ioaxxere (talk) 15:31, 3 August 2023 (UTC)[reply]
How do you propose to track down and eliminate incorrect inflected forms? What about their misspellings? (Pali endings in -iṃ have a frequent misspelling in the Thai script writing system with implicit vowels at least in part because of Thai spelling constraints being imposed on Pali.) It was only two months ago that we realised that an accusative singular of Pali masculine nouns in -in (equivalently, in -ī) had the wrong vowel length in the inflection tables. (The error had been copied from Thai Wikibooks.) Are you suggesting that we should categorise inflected noun and adjective forms by case, number and paradigm? --RichardW57m (talk) RichardW57m (talk) 16:23, 3 August 2023 (UTC)[reply]
I am cleaning up some Arabic entries and am running into the issue of grouping etymologies. There are sometimes different types of terms derived from the same root (e.g. under آمر there are different etym sections for a form-III verb; a form-I active participle that has some derived lemmas, including the one that produced emir in English; and various non-lemma forms), but sometimes also different terms derived from different roots (e.g. under بان there's four etymology sections reflecting entirely different roots). Sometimes both co-occur. There are lots of different solutions that people have attempted to make sense of this. I would like to clean some of these entries up by using extended etymology numbering; so for example, I could have ===Etymology 1.1===, ===Etymology 1.2=== and ===Etymology 1.3=== for different lemmas derived variously from one root, and ===Etymology 2.1===, ===Etymology 2.2=== for different lemmas derived from a different root. I am not proposing using this universally but simply allowing it as a possibility when it makes sense. Any objections?
(Similarly, WT:EL doesn't specifically allow ===Pronunciation N=== sections but sometimes people use them; IMO it just creates a big mess by doing this but that's another discussion. IMO what I'm suggesting would be a lot cleaner than ===Pronunciation N=== sections and would require no changes anywhere except for in the occasional bot script, e.g. if User:JeffDoozan has a script to sort etym sections.)
@Benwing2 I'm not an Arabic editor, but it seems fine to me. If you do go ahead and do this, would it perhaps be a good idea to reflect this in Wiktionary:Entry layout? I think it would be good documentation for bot authors, as almost everywhere else it's all integer etymology numbers.
Also separately: would it be possible to address the problem by having multiple separate lines in the etymology section for each individual derivation from a given root, e.g.:
(*) I apologise I know virtually nothing about how Arabic forms its words, but I'm guessing the verb is formed from the root here.
This may just end up more cluttered than the original idea, though... Your idea preserves the roots under the same "major etymology" number but still keeps the separate derivations separate. Kiril kovachev (talk・contribs) 19:07, 2 August 2023 (UTC)[reply]
@Kiril kovachev This is a good idea although it's indeed likely it will get cluttered, esp. when for a given root you can potentially have multiple verbs, participles, verbal nouns and non-lemma forms with the same non-vocalized spelling (but different pronunciations), and each participle can potentially have a derived adjective and/or noun. With everything stuffed together like this, the etymology section can get confusing, and likewise the pronunciation section if there is one. (And yes you are correct; in general, Arabic verbs are formed directly from roots.) Benwing2 (talk) 19:32, 2 August 2023 (UTC)[reply]
As I understand it, part of the problem is the way we handle inequivalent pronunciations. Primarily for that reason, the homographic lemma and past form of English read are put in separate etymology sections. By contrast, the homonymous lemma and past form of English put don't even have separate entries.
Consequently, my understanding is that the indicative, subjunctive and jussive of آمر currently in Etymology 3 are likely to end up under separate etymologies, just because they have different Classical Arabic pronunciations. --RichardW57m (talk) 14:25, 3 August 2023 (UTC)[reply]
@RichardW57 I would probably not do that. My script for generating inflections for Russian, for example, puts all the non-lemma forms of a given lemma in the same etymology section even if they have different pronunciations (in Russian, this is due to stress differences). I think it makes sense to do the same for Arabic. Benwing2 (talk) 23:52, 3 August 2023 (UTC)[reply]
@Benwing2: So how do you handle the pronunciation sections? Do you subordinate them to part-of-speech sections? (The problem goes away while one does not give pronunciations.) --RichardW57 (talk) 05:20, 4 August 2023 (UTC)[reply]
I mean, he is probably him who bears the technical consequences of it, and assesses them insignificant in view of possible (readability?) advantages, so it is probably okay. He specifically asked about objections and since I did not object there were none from me—even habit is barely an objection as it would be an occasional alternative. Subnumbering etymology sections whose numbering is already allowed is not comparable to pronunciation section numbering however, meseems. We need to see how he will clean issues, I may not clearly see. Fay Freak (talk) 20:00, 2 August 2023 (UTC)[reply]
No I agree, I think your suggestion is a great Idea, based on what you said. I just wanted to make sure the Arabic editors would be aware if a change happened. I also don't know a lot about Arabic which is why I didn't join the conversation. سَمِیر | sameer (مشارکتها • با مرا گپ بزن) 20:53, 2 August 2023 (UTC)[reply]
@Fenakhay Wow, that page is insane. I might have to give you a different example to start off with because I have no idea how many homophonous roots there are under عرض, and there's no etymology to speak of for any of those terms. Benwing2 (talk) 23:56, 3 August 2023 (UTC)[reply]
No strong opinion on this and will rework my bot to handle whatever is decided. There's no need to worry about the bot causing problems in the interim, if it were to find ===Etymology 1.1=== today it would consider it "unhandled" and avoid making any changes. JeffDoozan (talk) 23:32, 4 August 2023 (UTC)[reply]
Strong support. This will make life so much easier for showing descendants into the likes of Korean & Japanese. Descendants at entries like 北京 (Běijīng) or 點心／点心 (diǎnxīn) are a mess. We should also focus on having proper usage examples for different lects and make sure that categorization works properly. Etymological information like for Cantonese at 圜 (won1) should not be hidden away in the expanded pronunciation section. I had to search around to find out where the categorization was coming from. Similarly, entries like 㗎 (jià) should not have Mandarin categorization if they are not used in it. (Also hopefully this will make some push towards separating out historical lects as well.) AG202 (talk) 04:20, 26 July 2023 (UTC)
Word of the year on WikipediaWikipedia we have GFC, NFT, -ism, and 😂. Ioaxxere (talk) 18:22, 14 July 2023 (UTC)
Support. I wonder if we should also change tone 9 from double acute accent to a breve. — justin(r)leung { (t...) | c=› } 14:18, 17 July 2023 (UTC)
Oppose I gave up counting the number of posts in some of these discussion threads. I would suggest using a rule that I have adopted for myself: absolute maximum of four separate posts in any given thread of comments, with emergency fifth comment in the limit (including explanation that no more comments will be made). The "was explicitly told that if he didn't stop posting harassing messages, I would call this vote" comment happened at approximately post FIFTY. Having a four comment limit makes every comment more meaningful, both to yourself as author and to the other readers. You may say "oh, it's necessary to have in depth discussions". Perhaps. But if there's any hint of confrontation, I would say just break it off and submit any lingering issues to the community at large, and don't participate in whatever outcome there is. Then just follow whatever the decision was. --Geographyinitiative (talk) 16:11, 19 July 2023 (UTC)
Oppose grouping together "present participle" and "gerund" in any circumstance. Gerunds are nouns. Ioaxxere (talk) 06:04, 21 July 2023 (UTC)
Woah woah woah. ѵ still pointed to ижица (ižica) and had entries for both Old Cyrillic and Cyrillic, so I don't see how someone could possibly not know that was the case. I did check that. If someone wants a translingual section, fine, but it should be something other than blindly copying the Unicode name as the definition, when we know that may very well be wrong. Though admittedly this entry was better than most; it was really just a matter of that section being redundant. kwami (talk) 21:39, 24 July 2023 (UTC)
Support or