WikiProject Linguistics collaboration on linguistics in Wikipedia	Task forces:	Applied Linguistics Etymology Philosophy of language	Phonetics SFL Theoretical Linguistics
Main page Discussion Assessment Article alerts Recognized content Portal

Skip to TOC
Skip to bottomSkip to bottom

	This project page does not require a rating on Wikipedia's content assessment scale. It is of interest to the following WikiProjects:
Linguistics This page is within the scope of WikiProject Linguistics, a collaborative effort to improve the coverage of linguistics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.LinguisticsWikipedia:WikiProject LinguisticsTemplate:WikiProject LinguisticsLinguistics

Archives

Index Archive 1 Archive 2 Archive 3 Archive 4 Archive 5 Archive 6 Archive 7 Archive 8 Archive 9 Archive 10 Archive 11 Archive 12 Archive 13 Archive 14 Archive 15 Archive 16 Archive 17 Archive 18 Archive 19 Archive 20 Archive 21 Archive 22 Archive 23 Archive 24

Etymology Phonetics 1, 2, 3 Phonology template Theoretical linguistics

This page has archives. Topics inactive for 180 days are automatically archived 1 or more at a time by Lowercase sigmabot III if there are more than 4.

Welcome to the talk page for WikiProject Linguistics. This is the hub of the Wikipedian linguist community; like the coffee machine in the office, this page is where people get together, share news, and discuss what they are doing. Feel free to ask questions, make suggestions, and keep everyone updated on your progress. New talk goes at the bottom, and remember to sign and date your comments by typing four tildes (~~~~). Thanks!

tl;dr Nexus and junction are much used by Jespersen. We currently have a dire stub about "Nexus grammar" and a strange article about "Junction grammar". Should I do something drastic about either or both, or can I just pretend not to have noticed them (perhaps while discreetly doing my bit to keep them hidden)?

-- Hoary (talk) 06:18, 1 August 2025 (UTC)[reply]

I'm glad to hear someone is working on this topic. I would like to add that from the Danish perspective, the concept of nexus was influential, and it may often be baked in various kinds of grammatical thinking without explicit mention. There's an article by Lars Heltoft (co-author of the 2011 important academic reference grammar) arguing explicit for the concept of "neksus". You probably also see that in the existence of NEXG, but unfortunately, I am unaware of any historical account of nexus within Danish linguistics (or any overall account of Jespersen's theory). Also note that Danish Wikipedia has a separate article on nexus the individual concept (rather than the concept), though I'm not sure that works very well. But that article has a number of interwiki links (of various size, note that the Russian one is called "nexus and junction").

However, my point is that there is potential for an article, and I'm not sure I support deletion of Nexus grammar - though I have not found the optimal source (as I haven't been looking that seriously beyond putting this concern in the back of my mind, and expecting such sources to be somewhere in the slump of not easily available stuff). Alternatively, Nexus grammar could be merged into Otto Jespersen (or a section could be written about it). On the question of what Nexus grammar is/should be about, I would probably prefer it to be about the theory (rather than one concept of the theory), but that may just be wishful thinking. Letting it "sleep" is fine by me (but not sure why removing it from the category matters for that?).

I agree that Junction grammar is odd, but it's unrelated and thus a separate problem (that I can't help with). //Replayful (talk | contribs) 23:55, 1 August 2025 (UTC)[reply]

Replayful, thank you for your stamina in reading and digesting my original post. I've taken the liberty of removing surplus ( ) from a link within your comment (the link didn't work, now it does); I hope you don't mind. I'm pausing for a time while I consider the meat of your comment. -- Hoary (talk) 02:49, 2 August 2025 (UTC)[reply]

Replayful, (or anyone), how about the following suggestion. On my hard drive, I create a short article about nexus (in Jespersen's sense). When I'm fairly certain that it's better than the current Nexus grammar, I blank the latter, copy my creation into the blank, save, and move the result to "Nexus (linguistics)" (on the grounds that "Nexus grammar" was probably a mistake for "Nexus (grammar)" and that "grammar" is unnecessarily over-specific here, and unfortunately so). In order to prevent a charge of "original research/synthesis" (and perhaps also out of laziness), I wouldn't mention the term as used in Role and Reference Grammar (which, incidentally, is an article that would greatly benefit from well-informed expansion). Within the article Otto Jespersen, I'd add a brief explanation of nexus and junction and I'd link from the former to "Nexus (linguistics)". None of this would preempt anybody's addition to "Nexus (linguistics)" of material about later Danish, R&RG or other related uses of the concept expressed by neksus/nexus. (Meanwhile, I'll forget I've ever seen the article Junction grammar.) -- Hoary (talk) 04:42, 2 August 2025 (UTC)[reply]

That's fine by me (though not sure why you'd need to blank before overwriting with something better). I wish I could help with the RRG article, but I think you're right in keeping that separate. //Replayful (talk | contribs) 19:27, 2 August 2025 (UTC)[reply]

Thank you, Replayful. I'm open to other editors' perhaps contrary opinions, but if I don't receive any in the next couple of days I'll go ahead as described above. -- Hoary (talk) 07:39, 3 August 2025 (UTC)[reply]

Now done. -- Hoary (talk) 08:53, 26 September 2025 (UTC)[reply]

I'm not sure what to do with this biography of a fringe theorist, which has been unsourced for 10 years. Please discuss. Bearian (talk) 18:18, 16 August 2025 (UTC)[reply]

There's an article in the Treccani encyclopedia, which seems like an important source. Unfortunately, my Italian isn't really in a shape to take on a biography. That suggests that other sources in Italian exists (check the article on Italian Wikipedia). //Replayful (talk | contribs) 12:59, 25 August 2025 (UTC)[reply]

This has been unsourced for 10 years. There is a list of criticisms on the talk page. If someone can please add reliable sources, that would be great. Bearian (talk) 11:55, 19 August 2025 (UTC)[reply]

Tagging Rua, Peter238, and Hoary. Bearian (talk) 11:57, 19 August 2025 (UTC)[reply]

the extipa chart has the consonant wildcards, and i believe voqs has the 'x'. but the lower-case are just broad phonetic transcription, and there is of course more variety than just those letters and meanings, esp cross-linguistically.

the wildcards are listed in the IPA article, with refs [some embedded as comments]; not sure if that's the better place or if it should be covered in a separate article like this. if the latter, we need some incoming links.

per the comment on the talk page, we should of course distinguish para-IPA wildcards from broad-transcriptional use of IPA letters with their normal meanings. the latter are open-ended and probably pointless to try to list, apart from a couple examples in our coverage of broad vs narrow transcription — kwami (talk) 12:54, 19 August 2025 (UTC)[reply]

Some of these are in Pullum and Ladusaw's Phonetic Symbol Guide (2nd ed) -- a copy of which is conveniently within arm's reach -- with the meanings attributed to them here. I looked up ⟨C⟩, ⟨N⟩, ⟨L⟩, and ⟨R⟩ and found the first pair (with such meanings) but not the second pair (or anyway, not with such meanings). The book of course doesn't claim to be a phonological symbol guide. My guess is that ⟨L⟩ and ⟨R⟩ and most if not all of the others are used and will be explained somewhere, but I am very poorly equipped to search for such explanations. (Rua, one of the two main contributors, last edited Wikipedia as recently as one week ago.) -- Hoary (talk) 08:19, 20 August 2025 (UTC)[reply]

Hoary, I'm very impressed. Bearian (talk) 08:24, 20 August 2025 (UTC)[reply]

Bearian, by my possession of the book, or by the length of my arm? Incidentally, the book doesn't have an entry for ⟨Q⟩, which also isn't listed in the article. I've seen it used for a geminate, in the context of the phonology of Japanese. So perhaps it too is a "cover symbol". Indeed, ⟨Q⟩ (together with ⟨N⟩) is used in en:Wikipedia's explanation of gemination in Japanese. (And no, I'm not suggesting that one WP article should cite another.) -- Hoary (talk) 00:01, 21 August 2025 (UTC)[reply]

Your possession of the book. I haven't seen your length. Bearian (talk) 01:20, 21 August 2025 (UTC)[reply]

A concerning and, as it stands, misleading article. I have left concerns on the talk page. --Trɔpʏliʊm • blah 19:35, 2 October 2025 (UTC)[reply]

The article Leiden school has been proposed for deletion because of the following concern:

The article is based only on a primary source. I couldn't find any independent sources discussing this school of thought. There are some independent mentions of the Leiden school in linguistics literature, but it's in relation to historical linguistic models, not the memetic approach to language that this article is about. Therefore, notability for the topic can't be established.

While all constructive contributions to Wikipedia are appreciated, pages may be deleted for any of several reasons.

You may prevent the proposed deletion by removing the {{proposed deletion/dated}} notice, but please explain why in your edit summary or on the article's talk page.

Please consider improving the page to address the issues raised. Removing {{proposed deletion/dated}} will stop the proposed deletion process, but other deletion processes exist. In particular, the speedy deletion process can result in deletion without discussion, and articles for deletion allows discussion to reach consensus for deletion. Antibabelic (talk) 08:19, 27 August 2025 (UTC)[reply]

Re-notifying about this discussion which ultimately got stuck in a quagmire for stupid reasons. I believe some fresh input would help us get to a conclusion in either direction. ~ IvanScrooge98 (talk) 01:03, 30 August 2025 (UTC)[reply]

An editor has requested that Swahili language be moved to another page, which may be of interest to this WikiProject. You are invited to participate in the move discussion. Kowal2701 (talk) 20:57, 10 September 2025 (UTC)[reply]

Just alerting you that this BLP of a linguist has been proposed for deletion. Bearian (talk) 00:40, 17 September 2025 (UTC)[reply]

Please address the multiple issues tagged. This was proposed for deletion, and could be nominated for deletion, so this is urgent. Bearian (talk) 03:51, 23 September 2025 (UTC)[reply]

There is a requested move discussion at Talk:Hangul orthography#Requested move 23 September 2025 that may be of interest to members of this WikiProject. grapesurgeon (seefooddiet) (talk) 20:43, 25 September 2025 (UTC)[reply]

Despite the countless inbound links as a subject frequently referenced on Wikipedia, the article itself is...not very good. C-class seems, at least to me, to be a charitable understatement. It may need another week of AFI, and it definitely needs to be reorganized. T3h 1337 b0y 23:14, 2 October 2025 (UTC)[reply]

Can someone who understands the concept of finiteness in linguistics have a look at the section Draft:Finiteness#Mathematics and linguistics? I want to be sure that I am correctly relaying that the linguistic concept of finiteness actually relates to the general concept of finiteness (i.e., things being finite, bounded, having an end). I copied some content there from the finite verb article, but I'm not entirely sure I understand it well enough to have conveyed it correctly. Cheers! BD2412 T 00:17, 18 October 2025 (UTC)[reply]

As part of cleanup for a lot of AI articles recently generated (see WP:ANI#URGENT: Mass draftifications by User:Asilvering (100+ in like a half hour) despite great sources; targeted at one editor and User talk:PaulHSAndrews) there are a couple of new linguist biogs. These could use any assistance available from people more familiar with the field of Indo-European linguistics.

• Dirk Boutkan
• Guus Kroonen

Also, does Marc Pierce (U. Texas, Austin) warrant a biog? (Hopefully not AI-spawned!)

Thanks Andy Dingley (talk) 11:00, 22 October 2025 (UTC)[reply]

There is a discussion about the Nişanyan Dictionary and Nişanyan Names websites on the reliable sources noticeboard. Input from anyone with relevant knowledge would be helpful. See WP:RSN#Are Nişanyan Dictionary and Nişanyan Names considered user generated?. -- LCU ActivelyDisinterested «@» °∆t° 14:09, 27 October 2025 (UTC)[reply]

There is a requested move discussion at Talk:Voiceless alveolar taps and flaps#Requested move 25 October 2025 that may be of interest to members of this WikiProject. TarnishedPath^talk 15:09, 3 November 2025 (UTC)[reply]

There is currently a discussion at Template talk:IPA pulmonic consonants regarding the transcription and placement of certain consonants in the current chart. The thread is Template-protected edit request on 31 October 2025. The discussion is about the topic Creaky-voiced glottal approximant. ~ oklopfer (💬) 15:21, 3 November 2025 (UTC)[reply]

An editor has requested that Altaic (disambiguation) be moved to Altaic, which may be of interest to this WikiProject. You are invited to participate in the move discussion. PK2 (talk; contributions) 22:35, 14 November 2025 (UTC)[reply]

There's currently an AfD open about an alleged "revertive case" in Manchu that doesn't seem to show up in any published sources. This could use attention from knowledgable editors, in case the current article is a garbled version of something that appears in sources under another name.

But I've noticed that the author of that article also created others that are similarly questionable. I'm particularly curious about ergative-genitive case, which is sourced to two papers that talk about the genitive case in ergative languages, but don't seem to mention a unitary genitive-ergative case. The alleged glossing abbreviation "EGN" doesn't seem to exist either as far as I can tell. So other eyes would be appreciated. Botterweg (talk) 16:18, 16 November 2025 (UTC)[reply]

I replied to the AFD. I agree with your suspicion here. I will try to look over the other articles when I am back where I have access to resources. -- LWG ^talk 16:35, 16 November 2025 (UTC)[reply]

Regarding ergative-genitive case, there are languages in which genitive and ergative are identical (see this paper for one example). But I don't see evidence for the term "genitive-ergative" or the glossing convention EGN. -- LWG ^talk 16:50, 16 November 2025 (UTC)[reply]

Another term for a case with ergative and genitive function is relative case. But I don't believe this thing is able to carry its own article independent of the ergative article. //Replayful (talk | contribs) 18:24, 16 November 2025 (UTC)[reply]

Thanks for this! I think I'll go ahead and nominate this one for deletion too. Botterweg (talk) 23:52, 16 November 2025 (UTC)[reply]

The AfD link is here by the way, if anyone wants to weigh in. Botterweg (talk) 19:14, 19 November 2025 (UTC)[reply]

I think all these go back to the 2006 addition of a "less used cases"-section in the Manchu article, and it's descriptions (or interpretations thereof) then spread to list of grammatical cases and individual articles. I spent a bus ride yesterday going through them to check which of them are in Erich Haenisch's old-ish grammar. In most cases there is no proper equivalent, or there is a very creative analysis (or misunderstanding) based on something in the language. The two things closes to how they're described in the Manche articles are the "(essive-)formal case", which in reality is a postposition, and the "terminative", which in reality is an adverbial verb form. I think those descriptions were already somewhat creative or unfortunate readings of the sources, and then were spread with possible reformulation to other article - so probably not a hoax, but some unfortunate repeated reanalyses. On Wikipedia:Articles for deletion/Initiative case it's suggested that some of it may go back to a 1879 grammar, which is probably a bit older than what we want modern case descriptions to be based on. Given that the Manch article has at some point been updated with modern sources (which I guess then don't support the less used cases-section) and that some of the relevant suffixes are actually already treated better somewhere else in the same article, I think it should be safe to remove the "less used cases" from the Manchu article, and then go through the relevant entries in the case list article, and maybe double check if articles based on the Manchu article still deserve to exist. //Replayful (talk | contribs) 16:48, 17 November 2025 (UTC)[reply]

I'm asking the community's opinion about an Italian IPA: Margherita Hack. It should be [marɡeˈriːta ˈak] because Italian the sound /h/ isn't pronounced. It may be forcedly pronounced in foreign names or words to underline their foreigness, but: 1) not for names of Italian persons (nor for stably naturalised words) even if of foreign origin; 2) a realiable source proving such a possible pronunciation must exist. Instead, a user is continuously adding a /h/ between parentheses for the surname Hack: [ˈ(h)ak]. His point is: every /h/ can be either pronounced or not pronounced in Italian. This is false, you can search all the articles of Wikipedia about Italian or the sources cited in the articles and you won't find anything supporting this point. He bases his assertion on a footnote of the help page (Help:IPA/Italian) where it's said that "/h/ is usually dropped", which clearly means what I've explained, not what he claims. Note that this sourceless footnote was added by him years ago, exactly as the /(h)/ added to the original IPA of Hack. But a user himself isn't a source, much less a reliable source (and much less if banned from the Italian Wikipedia for his behaviour incompatible with Wikipedia itself...). A reliable source is needed to prove that the pronunciation with /h/ is possible, but the user just added a random YouTube video and a Forvo audio by a random person. Everyone could add any pronunciation of any Italian IPA if he goes cherry picking through the Internet to find an unreliable source which agrees with his POV. Reliable sources are for example the 2 Italian phonetic dictionaries cited in the help page, for example one certifies that "Heidi" can be pronounced with /h/ while "hotel" can't. All I'm explaining here is explained in a discussion in the talk page of Margherita Hack, where I invite any interested user to say his opinion about this matter, i.e. the Italian pronunciation of "Hack", not the help page containing the foreing sound /h/. I hope you'll agree that, if the general rule is that /h/ is dropped, for any exception an authoritative source is necessary. The discussion is here: Talk:Margherita Hack#Pronunciation. ~2025-34431-41 (talk) 23:30, 17 November 2025 (UTC) @Moyogo, Tropylium, Kjoonlee, Buidhe, Phinumu, Paintspot, PharyngealImplosive7, and Citation unneeded: (WikiProject Linguistics participants) ~2025-34533-51 (talk) 12:12, 18 November 2025 (UTC)[reply]

(moved here from User talk:Beland)

Hi, I can see that you've moved the articles Æ (to "letter ae") and Ø (to "O with slash") (among others, but not Å). Has there been any discussion of this anywhere? You cite WP:TITLESPECIALCHARACTERS, but from what I can read, it recommends the redirect situation ("redirects from versions of the title that use only standard keyboard characters") that was present before your moves. In the case of Æ, I don't think the term "letter ae" is used anywhere else than the Unicode description. Do you want to elaborate a bit? I would have thought this should be discussed at Wikiproject Linguistics before implementation. //Replayful (talk | contribs) 21:20, 29 November 2025 (UTC)[reply]

I have reverted both bold moves. Please open a move discussion if you would still like these pages to be moved. You did not perform necessary post-move cleanup by fixing the redlinked hatnote at Ø. I will also note that Æ has been through an RM before; you may find inspiration for other alternative titles there. Toadspike [Talk] 00:35, 30 November 2025 (UTC)[reply]

Beland, I see you have done a rather large number of these moves in quick succession. They appear to be based on a misinterpretation of TSC. Please stop performing similar moves and undo all of them, as they are controversial. Toadspike [Talk] 00:39, 30 November 2025 (UTC)[reply]

There's no particular rule requiring discussion before moving articles if the move hasn't been discussed before. If there had been incoming redirects at the move destinations, I would have looked for previous discussions, but there weren't. I couldn't find any move discussion for Ø; digging into archived discussions, it appears a move to AE ligature was suggested two decades ago, before the current title policies were written and before Wikipedia even used Unicode. I was going to move Å to A with overring, but since there is now an objection to this type of move, I'll pause for the suggested discussion here on the WikiProject talk page. Thanks for pointing out the broken link, BTW; it never occurred to me that a page move could cause a broken link if a redirect was left behind. I'll check for that if I end up doing any future moves.

It's true, WP:TITLESPECIALCHARACTERS says to use redirects for characters that can't be typed on a standard keyboard if those special characters are part of the most appropriate title. I'm going through systematically and making sure that either those redirects exist or that articles are renamed to more appropriate titles. WP:ENGLISHTITLE says that titles should reflect the common English names of our subjects. The naturalness criterion at Wikipedia:Article titles also seems relevant; if a title can't be typed on a standard keyboard, that doesn't seem like a natural way to search for or link to an article (though we may find evidence to the contrary). I question whether "Ø" and "Å" are the most common English names for those letters. Looking at the Unicode standard, which has widely used names for every letter in English, for U+00D8, it doesn't use the name "LATIN CAPITAL LETTER Ø"; the standard calls it "LATIN CAPITAL LETTER O WITH STROKE". This to me reinforces the notion that "Ø" is not the English name for this letter. most people would say this letter isn't the English alphabet, even though it's retained in some imported words.

One thing that made me pretty comfortable making these page moves is that most of the articles in Category:Latin-script letters use English words as titles rather than the Unicode characters for those letters, for example: Thorn (letter), Hwair, and Turned A. H with stroke seems to be directly parallel to O with stroke. WP:CONSISTENT tells us that we should either use English words or Unicode characters for letter titles, but not have a mix of the two. If there is consensus for Unicode characters, that's fine, but English words seem to be somewhat better for navigation. It's not possible to make the redirects required by WP:TITLESPECIALCHARACTERS for many letters. For ligatures there's usually a disambiguation page, which isn't so bad for cases like AE. For Latin letters with diacritics, there's always going to be a letter article in the way. So for example to get to Å, you need to go to A and either notice that Å is in the infobox, or click through to A (disambiguation) then Å (disambiguation) and finally to Å. -- Beland (talk) 04:52, 30 November 2025 (UTC)[reply]

Doing bold stuff is fine, but you should be ready for a bold response. I don't think the Unicode descriptions are enough to argue that they are common in English sources, so that more has to be established on a per-article basis. That makes them a bit different from thorn (a specifically English letter for which there are many English sources). Ø and Å are not treated as "letters with diacritics" in the languages that use them (which also is the case for Æ, but other languages use it as a genuine ligature), so the "with overring/slash"-title seems misleading from that perspective. I'm not sure which redirects you say are impossible to make - A with overring, O with slash works, and you say the AE disambiguation page isn't so bad. But generally, I don't think one simple way to title letter articles can be established, so it has to be discussed on a per-article basis. //Replayful (talk | contribs) 08:41, 30 November 2025 (UTC)[reply]

Sure, I'm happy for the bold response!

WP:TITLESPECIALCHARACTERS envisions e.g. a redirect to Å from an ASCII version with diacritics removed. That would be A, but it is impossible to make a direct redirect from there because it is the article for the Latin letter A. I'm just a bit worried about readers looking for Å, typing A, and getting lost on the way from one to the other. I suppose readers who don't know what title we've chosen will have that experience anyway. But it would be easier on the second visit if readers knew they could type "A with ring" or whatever they vaguely remember and get to the right place. Or they learn this naming pattern from other letters. Given our guidelines, perhaps these concerns are neither here nor there, and we should just look to source evidence.

In that spirit, what sources do you have that show "H with slash" should be treated differently than "O with slash", which would outweigh Unicode treating them the same and also outweigh WP:CONSISTENT? -- Beland (talk) 09:07, 30 November 2025 (UTC)[reply]

Allan et al. 2000 Danish, an Essential Grammar simply refers to Æ, Ø and Å, at most in the formulation "the letter(s) X" (p. 185) - never uses formulations like "with overring/stroke" or ligature. In this review of a dictionary (p. 334), æ and ø are also just referred to with the character. This is in line with my experience reading about is (as far as I remember). It comes off to me as very odd to refer to those with the "X with Y" description - and if the articles were so titled, people would be misled to think they should refer to them like that. I'm not saying they should be treated different from "H with slash" as such (but that one doesn't exist so it shouldn't have a title, I guess), but whatever name they're moved to, should be a term actually used in sources where people might (have) read about it (in line with WP:COMMONNAME).

I don't think moving to "X with Y" is going to change whether people get lost when looking at article X; as you say, the derived character cannot take the place of what it's based on. Whether an "X with Y" page redirects to the unicode title or opposite doesn't change that. And as pointed out below, some of the characters have multiple possible descriptions, and different sources use different names. For instance, "letter tone two" for Ƨ seems odd given its use in the Metelko alphabet where tone is not relevant. "Turned m" isn't a great description of what the capital form looks like. Using the character as the title seems the most consistent (and will probably also be in line with WP:COMMONNAME), and many people are smart enough to search by copying (arriving from search engines will also look more relevant with the character as the title, I'd say). The gist of WP:CONSISTENT also seems to be that consensus can trump consistency. Using the character in the title is also more consistent with the many articles about people with those characters in their names (definitely the case for ÆØÅ; the IPA characters are a bit different since the sound they denote will have a separate article - also a reason why it doesn't make sense to compare "Ə" with "schwa", since we don't know which are used to refer to the vowel or to the character).

I think the Unicode descriptions you mention are written specifically to avoid the use of Unicode characters - a practice we don't need to copy. Better sources are needed to establish how the letters are referred to. //Replayful (talk | contribs) 22:02, 30 November 2025 (UTC)[reply]

Seconding Replayful – the possible alternative names are many and varied, the situation for each letter is complex and different. "There's no particular rule requiring discussion before moving articles if the move hasn't been discussed before." – Yes, for uncontroversial moves. It doesn't require a previous RM to see that something could be controversial, but moving pages that have been through one, or have been boldly moved several times, is obviously controversial. Toadspike [Talk] 10:00, 30 November 2025 (UTC)[reply]

Now that the discussion has been moved here, I would like to point out some other characters relevant for linguistics that were moved, such as IPA characters (I put them first here):

• Ə
• Ʊ - talkpage has a discussion of the term "Latin upsilon"
• Ɯ
• Ɵ
• Ɪ
• ⱺ (Swedish phonetic symbol)
• Ꞩ, Ꞡ, Ꞧ, Ꞥ (used in former Latvian orthograpy)
• Ƨ, Ƅ, Ƽ - "tone letters"
• Others: Ʋ, Ỽ, Ȼ

I don't think the IPA symbol pages work well at another title than the symbol (at least the first 4).

(And sorry for being nitpicky, but "untypeable letters" is a somewhat misleading title. They are perfectly typeable on keyboards designed for it, and by changing your keyboard layout, you can also get to type Æ, Ø, Å. The others here, like the IPA symbols, probably need plugins or more complex though.) //Replayful (talk | contribs) 09:08, 30 November 2025 (UTC)[reply]

I mean "untypeable" in the sense of WP:TITLESPECIALCHARACTERS - not able to be entered on standard keyboards, by which it seems to mean ANSI US QWERTY give or take a £. English Wikipedia is designed for English speakers with English keyboard layouts, not Swedish, etc. Most people have no idea how to change that, and wouldn't and shouldn't do so for a one-off web query. This makes non-ASCII characters de facto untypeable for many readers in our core audience.

What sources would you point to show that general English sources would use the symbols rather than names with words for these characters? Google Books Ngrams shows "schwa" surpassed "Ə" in frequency around 1970, and is currently much more common.

I was going to check the fourth symbol against "theta", but it turns out to be a barred "O". Visual confusion like that seems like a good reason to prefer English words as titles. -- Beland (talk) 09:34, 30 November 2025 (UTC)[reply]

"letter ae" and "O with slash" can just be useful redirects, but I absolutely do not support moving letters/symbols to descriptions like these.★Trekker (talk) 10:15, 30 November 2025 (UTC)[reply]

Well, "barred o" or "o with stroke" are very confusing as they could very well be ø or ɵ. For example Robert Bringhurst’s The Elements of Typographic Style uses "barred o" for ø but Unicode uses "o with slash" for that letter, the IPA handbook uses "slashed o". Both Unicode and the IPA handbook use "barred o" for ɵ, but Unicode uses the name "o with middle tilde" for the capital Ɵ. This is not to say articles should or shouldn’t be renamed, just that it’s not straightforward and that there are pros and cons either way. --Moyogo/ (talk) 14:37, 30 November 2025 (UTC)[reply]

Would y'all then support moving articles in Category:Latin-script letters that previously had English-word titles to Unicode character titles, to follow WP:CONSISTENT? -- Beland (talk) 20:52, 30 November 2025 (UTC)[reply]

I wouldn't mind more articles having a title just consisting of a character, but it needs to be taken into account whether the article is written from the perspective of the character or something else (handwriting variation, orthography-specific terms, probably others) - so it's not just a question of moving, but possible rewriting. My opinion is that the ones I listed above could be moved back, but I won't wouldn't say it can be done for all, as it depends on why it was moved to its current position. //Replayful (talk | contribs) 13:04, 6 December 2025 (UTC)[reply]

I probably would yes.★Trekker (talk) 08:50, 9 December 2025 (UTC)[reply]

User:Botterweg believes that the articles Conditional logic and Ramsey test may have some indeterminate sort of issue regarding their prose style, or support from sources, or plagiarism, or what have you. I do not see the issues that Botterweg is pointing at, and I believe that the articles have no such issues, hence I would like project members to take a look at both articles and give feedback, thank you. Thiagovscoelho (talk) 14:33, 2 December 2025 (UTC)[reply]

It looks like Botterweg's concern is that the articles appear to have been generated by a LLM like ChatGPT, which is strongly discouraged on Wikipedia (see WP:LLM for more details). Can you please describe the process by which that text was written and verified? -- LWG ^talk 15:03, 2 December 2025 (UTC)[reply]

For "conditional logic", first I read (1) SEP on "logic of conditionals" (2) this SEP supplement to "defeasible reasoning" that had a nicer axiomatization of VC than the one in Lewis's own works on counterfactuals (3) Handbook of Philosophical Logic, edited by Gabbay and Guenthner, volume 4, p. 82 on Stalnaker's logic and following pages on other mentioned logics (4) Stalnaker's own original "Theory of Conditionals" paper, featured in the book "Ifs". I took notes and made an initial draft of my edit based on them, then I did a web search for additional sources to use and incorporated those in the final product. For "Ramsey test", I first read about it in Gordian Haas's book on verificationism (§9.3) a while ago, which made me aware of there being a lively academic literature dedicated to improving it; so recently I looked at the section under "Belief revision" and produced more content expanding on it and did a web search for additional sources and information. I verified that all content was supported by citations and all citations supported the content, of course. Thiagovscoelho (talk) 15:30, 2 December 2025 (UTC)[reply]

I replied on the article talk page. From a quick glance over, it appears to me that many/most of your citations lack sufficiently specific page numbers to enable someone to verify them without reading the entire book in question, which is time-prohibitive. Please update the article with citations based on your notes so other readers can follow your thought process and verify the content. -- LWG ^talk 15:54, 2 December 2025 (UTC)[reply]

I will replace the complaint with "Page numbers needed" since this is a more helpful criticism, and I will begin to provide specific page numbers wherever they seem lacking. Thiagovscoelho (talk) 16:19, 2 December 2025 (UTC)[reply]

No, my concern is simply that these are LLM-generated articles. I added the LLM template to indicate that they will require more careful inspection before being marked as reviewed. Surely you're not disputing that these are LLM-generated? Botterweg (talk) 20:43, 2 December 2025 (UTC)[reply]

It doesn't seem possible to address your complaint when you are not pointing to specific flaws in the content that can be improved by further edits, such as improving tone, sourcing, etc. Per its page, the LLM template is intended for articles that have problems, and potentially need similar treatment to "cleanup rewrite", because they have actual issues in tone, sourcing, etc.; and is intended to be removed when the problems (it is understood that there are known problems) are removed. You are misusing it if there are no specific issues with the content. Since I had been working on the articles, I am interested in ensuring that they have no issues, and I thought I had successfully ensured this; so the template communicated the wrong message to me, if it was meant as indicating nothing in particular about the content, when this is not what it is for. Thiagovscoelho (talk) 23:12, 2 December 2025 (UTC)[reply]

I have read over the article on conditional logic and it does not look like an LLM generated article to me. I pasted it into the ZeroGPT checking tool and the result was "Your text is authentically human-written 95.85%". Then I used QuillBot and it reported "0% of text is likely AI". Then I tried Surfer AI Content Detector and it reported "0% chance that your text was generated by AI". I know these tools are not always accurate, but 3 out of 3 ain't bad. I am familiar with the subject matter of this article and it seems to be well-written and accurate. The only issue I can see is that it could benefit from page numbers being supplied in some of the referenced books and papers. I don't see any reason for the LLM tag to continue. Dezaxa (talk) 21:13, 4 December 2025 (UTC)[reply]

The user hasn't denied that the article is LLM-generated. They're just claiming that they thoroughly fact-checked it. That could be true, since I agree (as someone with background in the topic) the LLM's text does pass a sniff test. But I think the template should remain until the article has been thoroughly vetted by other users. Botterweg (talk) 21:57, 4 December 2025 (UTC)[reply]

Regarding AI detectors, the only one that's remotely reliable at the moment is Pangram, though I think that's a moot point for this discussion unless the user actually wants to go there. Botterweg (talk) 22:03, 4 December 2025 (UTC)[reply]

Fair enough. I checked with Pangram and it said 30% AI generated. When I get some time I'll make some edits and try to improve it further. Dezaxa (talk) 11:30, 5 December 2025 (UTC)[reply]

For WikiProjects that participate in rating articles, the banners for talk pages usually say something like:

• "This article has been rated as Low-importance on the importance scale."

There is a proposal to change the default wording on the banners to say "priority" instead of "importance". This could affect the template for your group. Please join the discussion at Wikipedia talk:WikiProject Council#Proposal to update wording on WikiProject banners. Stefen 𝕋ower ^{Huddle • Handiwerk} 19:45, 6 December 2025 (UTC) (on behalf of the WikiProject Council)[reply]

There is a requested move discussion at Talk:*Dʰéǵʰōm#Requested move 5 December 2025 that may be of interest to members of this WikiProject. Vestrian24Bio 13:36, 12 December 2025 (UTC)[reply]

Is Marcantonio considered a fringe theorist? Contents referencing Marcantonio is being removed by a user on the grounds that her views are fringe. Oumuamua8 (talk) 01:48, 13 December 2025 (UTC)[reply]

The person who removed various citations to her was me. She is definitely fringe; she disputes both the Indo-European and Uralic language families, which are the two best-studied and most firmly established ancient language families there are. See this paper for a criticism of her work: [1]. Stockhausenfan (talk) 12:09, 23 January 2026 (UTC)[reply]

Hello, Please note that Spoken language, which is within this project's scope, has been selected as one of the Articles for improvement. The article is scheduled to appear on Wikipedia's Community portal in the "Articles for improvement" section for one week, beginning today. Everyone is encouraged to collaborate to improve the article. Thanks, and happy editing! Delivered by — MusikBot talk 00:05, 15 December 2025 (UTC) on behalf of the AFI team[reply]

So lately, I have been the subject of various debates on whether users should be allowed to publish phonology charts with some allophones in them. Furthermore, do these charts necessarily need to display only the phonemes? I feel that we should build a consensus regarding allowing the display of phonological charts to show a small amount of allophones, specifically the ones more notable or common, and properly transcribed as well. According to Wikipedia:WikiProject Linguistics/Phonetics/Phonology template, it states that allophones "are not normally shown", which almost implies that it isn't necessarily professionally recommended, but still could be accepted and still legible for viewers and readers out there. And there are quite a few charts I have seen (even in linguistic publications) that do include them, making them not necessarily phonemic or phonetic, but rather more generally "phonological". To be clear, I am well aware that these rules or guidelines (on not displaying any allophones in charts) are considered to be incredibly necessary for licensed publishing linguists doing actual linguistic fieldwork where they do *need* to strictly publish both a distinguished phonetic chart and phonemic chart. But here, this is more of a summarized or generalized phonology section within a language article on a public online encyclopedia. No actual linguist out there would ever use an article from our site as a source or citation.

As for why I think *some* allophones (specifically notable or common ones) should be allowed in the charts, in my honest view as a linguistics major and as someone who has analyzed many thousands of phonemic charts over the years (even among some of them with a couple or so allophones in them), I view this as a way of previewing either more of, or the rest of the sounds of a language in a chart, not just only seeing the phonemes. It can be seen as a way of convenience among some readers, because then they can easily get an idea of what more of the sounds of the language are by reading a chart. Although I do believe with that being said, it is very important to transcribe them and mark them correctly and don't forget the accurate transcription. For example if they are allophones, then use phonetic brackets [] or tildes (~), or even parentheses () to mark them as allophones. But I also believe that the use of footnotes could also be permitted to explain the phonetic context of the sound as well. For that, I also believe that maybe we should be referring to them as “phonological charts” rather than “phonemic” or “phonetic” ones, to avoid confusion. But no matter what it is, at the end of the day, as long as the chart is properly sourced and written, and with the correct sounds and correct transcriptions, and made easy and understandable for readers and viewers across the aisle to comprehend, it should be at least acceptable to display these charts. And also provided that there are at least notes on the allophones below the chart, for context and detailed explanation.

That is my full viewpoint on this subject matter, feel free to agree or disagree. Fdom5997 (talk) 02:01, 27 December 2025 (UTC)[reply]

This question is more complicated than people often imagine, because it is not always clear what are phonemes and what are allophones. Two phones may have semantics-distinguishing functions in one variety of a language and not in another. The best route is always to ask what information will help the reader. Including things they are likely to stumble over is helpful, cluttering the article with every possible detail is unhelpful. I suspect that is why the guidelines say "are not normally shown" (which clearly implies "but are sometimes shown"). It might be prudent to refer to these as "phonetic charts" rather than "phonological charts", so that flexibility is not ruled out by definition. Doric Loon (talk) 10:34, 27 December 2025 (UTC)[reply]

Important context is missed by not looking at the full guidelines and just focusing on the first part of the statement, "not normally shown":

Parentheses may be used to set off marginal phonemes, or ones found only in loanwords, for example (m); the meaning of the parentheses should be explained in the text. Allophones are not normally shown, being relegated to the section on allophony below, but may be presented when it is not clear which should be taken as the default allophone of the phoneme. In such cases they should be clearly marked as variants of the same phoneme, for example m ~ mb, and not set off in separate cells.

While what we call the charts by is somewhat irrelevant, I think calling them phonetic charts is the least preferable of the three suggestions and in fact would make what we call them unnecessarily relevant, since that would imply you can throw in all of the realizations and would introduce confusion, leading a naive reader to believe most or all of the phones have equal status when looking at the basic chart. They are of course in phonology sections of pages (making them phonological charts, broadly), and typically we have stuck to only showing phonemes (making them phonemic charts, more narrowly), just as the actual field does, keeping allophones out unless the lines are not clear (i.e. usually in free variation, as opposed to in complementary distribution/positional variants, where in the former there may be no clear overarching phoneme, while in the latter there typically is; the former case is exactly what the tilde is used for).

I firmly believe that the unnecessary inclusion of pure (allo)phones in these charts, when they are better explained in footnotes and sections below, diminishes a chart's legibility, particularly for a naive reader. If the guidelines exist with good reason for publishing linguists, I see no reason for us not to stick to them either.*

As for representation, clearly parentheses cannot be serving double meaning, they need to be consistently used in their symbology, to avoid further confusing when a sound is a marginal phoneme or just an allophone, something which is unfortunately currently being conflated in our presentations, sometimes or even often both being encompassed in parentheses in the same chart or language page. This is a disservice to the naive reader.

*Unless an author themselves which we are deriving from appears to have good reason for including allophones in their charts (not our translations). In those cases, the brackets seem acceptable, to avoid conflation with parentheses, but adding more allophones than the author shows in their own chart is again diminishing legibility, and accuracy. Who are we to pick and choose which others would be worthy?

To suggest no "actual linguist" would use Wikipedia as a source, I have already proven that otherwise in another discussion, with a recent example from Martin J. Ball, former president of the ICPLA, unquestionably an "actual linguist", directly using WP as a source to make a claim about categorization of sibilants, using that page, undeniably a WP page on linguistics: doi:10.3138/jcspeech.29303; so I don't know why that was repeated here after being shown to be untrue. As the encyclopedia matures, more and more academics will continue to open up to it, and the recent example from Ball, a rather esteemed linguist, is further normalizing that.

Finally, other editors have already suggested to Fdom that if there must be a chart of phonetics or allophones, then it ought to be kept separate from the chart of phonemics. Both may exist, but they should not be conflated and fused. ~ oklopfer (💬) 16:20, 27 December 2025 (UTC)[reply]

@Doric Loon I see your reasoning here, but to be clear, what I am saying here is not cluttering the article by adding every possible detail (part of the reason why I emphasize *some* allophones, not all of them). Also, I disagree with saying "phonetic" charts, because that would imply that we actually should add every allophone in the chart. So "phonological" in this case is at least more accurate. Fdom5997 (talk) 21:54, 29 December 2025 (UTC)[reply]

Hi @Fdom5997, yes I can see my first thoughts don't conform to a consensus here, and that's OK. :-) Doric Loon (talk) 11:28, 30 December 2025 (UTC)[reply]

Sure it is! I’m just stating my thoughts too. Hope that helps. Fdom5997 (talk) 17:03, 30 December 2025 (UTC)[reply]

I believe that allophones should be presented in a separate chart from the phonemes, rather selectively put with the phonemes in one mixed chart. This way, the reader can view both the phonemic and phonetic aspects of a language. This is present in many linguistic publications. 🪐Kepler-1229b | talk | contribs🪐 20:37, 27 December 2025 (UTC)[reply]

@Kepler-1229b I think the meaning of your comment may be a bit misconstrued by what appears to me to be a missing word; did you mean rather than selectively put…in one mixed chart, or truly just rather selectively put…in one mixed chart? ~ oklopfer (💬) 21:39, 27 December 2025 (UTC)[reply]

We sometimes make charts for languages that are too poorly documented for us to know what the phonemic inventory is; the researcher may have done their best, but with the caveat that the list may not be entirely phonemic. In such cases we should make that obvious to our readers. That could be done in the table header, but IMO the best option would be to delimit all phones with square brackets.

If we have two tables, one phonemic and one allophonic, the allophones should probably be enclosed in brackets as well. In fact, I think that convention should be added to the MOS. If there are two tables, we might also dab the first with virgules; would that be worth adding to the MOS as well?

If there's a single table, it should be phonemic. If we include multiple allophones, I think they should share a cell. Multiple allophones would certainly be appropriate when it's not clear which is the 'elsewhere' allophone, say "m ~ mb" when it's debated whether the language has phonemic nasals, but we might also add allophones where they're remarkable enough to warrant special attention. IMO we should not do this for run-of-the-mill palatalization or changes in voicing -- if we want that much detail, it should be in a second table as Kepler suggests.

@Kepler-1229b:, do you think an allophone table should be organized spatially by phoneme, or just by the articulation of the phone? If the latter, how would you suggest clarifying which phones are allophones of each other? — kwami (talk) 22:14, 27 December 2025 (UTC)[reply]

+1 for standardizing sqbr delims across the board in lesser documented/ambiguous situations and for designated (allo)phonetic charts

imo slashes/virgules would also probably make sense for phonemic charts when phonetic charts are provided alongside, but would they remain omitted in cases where only one is provided? should the 'remarkable enough' allophones be sqbr wrapped (and what determines when they are 'remarkable enough')? ~ oklopfer (💬) 00:01, 28 December 2025 (UTC)[reply]

I suppose it would be a good idea to recommend using virgules in inventory tables by default. I wouldn´t mind adding that to the MOS. It would address the issue of readers not always being familiar with what a consonant etc. table is in practice.

I would recommend not marking allophones with square brackets. I suspect that readers would soon expand them to the messy layouts that Kepler objects to below. Instead, I think we should continue the convention of using a tilde. If we mark phonemes with virgules, this would result in e.g. /m~mb/. (Or maybe spaced /m ~ mb/.) That should make it clear that they belong together and maybe discourage people from adding more.

'Remarkable enough' would be problematic, but there are cases that are so bizarre that IMO it would be a good idea to show them in the table. These wouldn't just be assimilation to the vowel or a neighboring consonant, but oddities where we might want to create a dedicated article for the sound. Though, granted, I don´t know how we´d defend a cut-off point if someone wanted to add dozens of additional allophones. — kwami (talk) 04:35, 28 December 2025 (UTC)[reply]

The allophone tables I have seen simply list all the allophones of all the phonemes in one table similar to the phonemic one. A single table would end up very cluttered and reader-unfriendly. The issue of arbitrarily selecting one "best" allophone is completely eliminated with two separate tables. 🪐Kepler-1229b | talk | contribs🪐 01:11, 28 December 2025 (UTC)[reply]

Do you mean that they would be clustered together, so that it is clear that they're a set?

We´d still need to choose a 'best' allophone for the phoneme chart, unless we're using orthography or basic IPA (as with English /r/). — kwami (talk) 04:27, 28 December 2025 (UTC)[reply]

Perhaps my point has been unclear. What I mean is to have a second table listing all the phonetic realizations of all phonemes, and distinguish them in a note below the charts. An example of this is in the Waninawa language article for the vowels, where I have not yet been able to clarify the distribution of the allophones. 🪐Kepler-1229b | talk | contribs🪐 18:02, 28 December 2025 (UTC)[reply]

A somewhat different approach could be like how the consonants phones are currently distributed among the phonemes in Pirahã language#Phonology, which avoids placing the phones on an arranged consonant chart altogether. Though this may be a bit of an exceptional case, as Pirahã has such a small phonemic inventory, and may get messier in languages with larger ones. Just pointing out another example. ~ oklopfer (💬) 18:42, 28 December 2025 (UTC)[reply]

If a table is titled "consonant phonemes" (see Choni language#Phonology), it shouldn't include allophones. For instance, we definitely don't want to suggest in any way that Turkish possesses a phonemic velar nasal. Those are mutually exclusive propositions. If I see a table of phonemes, I want to see a table of contrastive units. At least such tables should be renamed, if we want to show allophony. Then, there's the obvious question - which allophones? The most obvious candidates are those that are perceived by natives as separate sounds (the /i-ɨ/ (non-)contrast in Russian comes to mind here) but can be analyzed away as allophones of the same phoneme. Of course, there are more candidates. Sol505000 (talk) 11:09, 28 December 2025 (UTC)[reply]

@Sol505000 I definitely approve of that idea for sure. I say we just do that. Fdom5997 (talk) 04:21, 29 December 2025 (UTC)[reply]

You still need to use the talk page, not edit war. See WP:BRD. Sol505000 (talk) 18:26, 29 December 2025 (UTC)[reply]

The redirects:

• Voiced palatal trill
• Palatal trill
• Voiced post-palatal trill
• Voiced palatal fricative trill
• Voiced labiodental trill
• Labiodental trill

have been listed at redirects for discussion to determine whether their use and function meet the redirect guidelines. Readers of this page are welcome to comment on these redirects at Wikipedia:Redirects for discussion/Log/2025 December 22 § Trill consonants until a consensus is reached. ~ oklopfer (💬) 17:24, 29 December 2025 (UTC)[reply]

The redirects:

• ˀ
• ˁ
• ˤ

have been listed at redirects for discussion to determine whether their use and function meet the redirect guidelines. Readers of this page are welcome to comment on these redirects at Wikipedia:Redirects for discussion/Log/2026 January 9 § ˀ until a consensus is reached. ~ oklopfer (💬) 17:33, 29 December 2025 (UTC)[reply]

As a clarification note:

I opened this one not to delete (unlike the trills RfD above) but to retarget, as currently the former points to Glottal stop (letter), while the latter two both point to Pharyngealization. Meanwhile, Glottalization and Reversed glottal stop have been left out of the equation, and I find the inconsistency bothersome.

Secondary opinions on resolving the inconsistency are invited (there has not been any activity in 2 weeks, besides the relisting). Currently, I have supported two other commenting editors' suggestions to disambiguate, though they had suggested through hatnotes, which still leaves the question of which ones should be the canonical pages that the redirects point to. ~ oklopfer (💬) 02:32, 30 December 2025 (UTC)[reply]

I think this is what many of our phonology articles are sorely lacking. I'd like to see where e.g. the vowels of Luxembourgish come from. Yes, we already feature some of the information, such as the /æːɪ–ɑɪ/ and /æːʊ–ɑʊ/ splits of vowels that were formerly the same 2 phonemes (when the language was still tonal). There's still nothing about schwa insertion, breaking up historical Volk to Vollek and that it's by no means restricted to Luxembourgish. You get this in Ripuarian, Limburgish, Brabantian and Hollandic dialects as well (and also in Irish English!). Up until a year ago, I didn't know that /ie, ue/ are perfectly analyzable as the six and seventh vowels of Spanish, patterning with /e, o/ in verb conjugation and whenever the stress shifts, as in nieve vs. nevada (where /ie, ue/ become /e, o/). That's because the diphthongs come from Latin short lax /ɛ, ɔ/ and monophthongs from the long tense /eː, oː/ (to me, this is so audible especially for /e/, it's often so close and front by Polish standards!) This information wasn't even on WP, I put it there. I think many of our articles could be improved. Using Keller (1961) cited in Luxembourgish phonology seems to be a good start for at least some Germanic languages. Danish phonology could show progression from Old Norse, same with other articles. This will also help fight the idea that languages are kind of separate islands, which some people still seem to have. Sol505000 (talk) 18:38, 29 December 2025 (UTC)[reply]

I don't think there's anything stopping anyone from writing historical sections? But it might just overlap a lot with other content - "main" language articles usually have a history section, some languages have their own history article (e.g. History of Danish), and some historical language stages also have their own article; and sound changes are probably a very normal topic within the history of a language. I'm worrying a little that it will be difficult to keep it "strictly phonological" or distributed unevenly. //Replayful (talk | contribs) 21:53, 29 December 2025 (UTC)[reply]

There is a requested move discussion at Talk:Romanization of Serbian#Requested move 11 December 2025 that may be of interest to members of this WikiProject. TarnishedPath^talk 21:30, 29 December 2025 (UTC)[reply]

Speech-generating device has been nominated for a good article reassessment. If you are interested in the discussion, please participate by adding your comments to the reassessment page. If concerns are not addressed during the review period, the good article status may be removed from the article. Z1720 (talk) 02:50, 30 December 2025 (UTC)[reply]

Syntactic Structures has been nominated for a good article reassessment. If you are interested in the discussion, please participate by adding your comments to the reassessment page. If concerns are not addressed during the review period, the good article status may be removed from the article. Z1720 (talk) 20:55, 2 January 2026 (UTC)[reply]

Historical reconstruction of this language is a very new field for many reasons, especially because (IMO regrettably and erroneously) R.M.W. Dixon opined that it was impossible. But I have discovered that a book-length reconstruction was published in 2024, and I am very keen to document it. I will be adding content to the article from the grammar that I have but all linguistically-informed contributions are more than welcome! Cheers. Ikuzaf (talk) 01:21, 8 January 2026 (UTC)[reply]

Ikuzaf, thank you for taking on this project. Bearian (talk) 04:19, 2 February 2026 (UTC)[reply]

I have just started a talk discussion at Talk:/æ/ raising#page name breaks non-mainspace linking regarding a potential retitling of the w:/æ/ raising page, as the initial / forces the necessity to prefix with w: in any namespace but main; otherwise, it redlinks: /æ/ raising. Suggestions at that discussion are invited. ~ oklopfer (💬) 03:20, 15 January 2026 (UTC)[reply]

Nardog reminded me of WP:COLON, discussion is moot ~ oklopfer (💬) 05:27, 15 January 2026 (UTC)[reply]

There is an ongoing discussion at Talk:Sinti regarding how to apply WP:RS, WP:NPOV, and WP:RSUSE to minority or closed-language communities when academic and community-grounded perspectives diverge.

The issue touches on questions of epistemic authority, linguistic boundaries, and how Wikipedia evaluates sources for ethnonym etymology and ethnic classification.

Input from editors familiar with ethnic group representation, linguistics, or anthropology would be appreciated.

Talk:Sinti#Linguistic Authority, Closed Languages, and Asymmetry in Academic Classification

All perspectives are welcome. BlackDutchSinti (talk) 15:54, 16 January 2026 (UTC)[reply]

I think Portmanteau (and its talk and disambiguation pages) could do with some help. I suspect some editors know too little about linguistics and others know too much. I'm in the former category. Thincat (talk) 10:39, 17 January 2026 (UTC)[reply]

I tried reading it but got rather lost in the second paragraph, which rambles, not always relevantly, and includes the nugget "don't is also an example of a portmanteau morph" (whereas I'd thought it was a word, one that could be considered constructed of a root and a portmanteau [present, non-3sg, negative] morph). -- Hoary (talk) 11:53, 17 January 2026 (UTC)[reply]

Yeah, I think that's because it tries to relate portmanteau words to portmenteau morphemes, and the two should be kept apart. //Replayful (talk | contribs) 14:55, 31 January 2026 (UTC)[reply]

13 years later, this still needs your wisdom and expertise. Bearian (talk) 21:54, 18 January 2026 (UTC)[reply]

It's probably more worthy of deletion. It has very little real content, and most of it is not about numerals as a word class, but just different ways to express numbers. I couldn't search up anything about the allusions. And the "unique system" mentioned is only done so very vaguely. //Replayful (talk | contribs) 13:04, 28 January 2026 (UTC)[reply]

I PRODded it, feel free to remove if in disagreement. Stockhausenfan (talk) 19:51, 28 January 2026 (UTC)[reply]

 You may be interested in a discussion I created on WikiProject Languages about creating a template for formatting Jawi text. — W.andrea (talk) 22:13, 18 January 2026 (UTC)[reply]

Hello WikiProject Linguistics, I have an open request for the learning language app Preply. I am a paid representative so I make requests rather than directly edit. Let me know if you have any questions. Cristina Preply (talk) 11:31, 19 January 2026 (UTC)[reply]

The following discussion copied below was triggered by a large change I performed yesterday, removing what I consider to be erroneous interpretations of glottalic consonants - that is, supposed [ʔʲ ʔʷ] are impossible sounds, as they go against the definition of a glottal stop. I am centralizing the discussion here as it is clearly about a scope broader than the language article it was originally started on.

Courtesy pinging @Snowman304 & @ThaesOfereode for attribution & to let the both of you know I have moved the discussion to here.

Whether they are interpreted and transcribed as any of [ʔj ʔw], [ˀj ˀw], [jˀ wˀ], or [j̰ w̰] I have no prejudice, but transcribing as [ʔʲ ʔʷ] feels to me that it is just perpetuating a fundamental misunderstanding of what makes a glottal stop, a glottal stop. Others' opinions on the matter are invited and much appreciated. ~ oklopfer (💬) 17:18, 20 January 2026 (UTC)[reply]

Are there any sources to support the claim that these transcriptions "go against the definition of a glottal stop"? They are obviously possible to articulate, since the lips and tongue can be moved independently of the glottis. Whether there is a perceptual difference is another question of course, but there should be no rule against them if sources support their use. I would interpret them as simply indicating that the glottal stop and some other articulatory gesture overlap in time. Freelance Intellectual (talk) 17:46, 20 January 2026 (UTC)[reply]

I would interpret them as simply indicating that the glottal stop and some other articulatory gesture overlap in time.

yes, and that is essentially what glottalics are. ~ oklopfer (💬) 18:26, 20 January 2026 (UTC)[reply]

So we are agreed that they are not impossible. I would still like a reply to my question about sources supporting the original claim. We should not be attempting to standardise analyses across the descriptions of all languages, which would be original research. Freelance Intellectual (talk) 21:28, 20 January 2026 (UTC)[reply]

I'm going to start by walking back my hard stance claim, as I cannot find any explicit statement of it in any literature; I admit that I crossed the line into WP:SYNTH, and deserve several WP:TROUT thwacks. That being said, I have been sourcing quotes from numerous authors to demonstrate what led me to make such a statement; importantly, what they really point to is large ambiguity and uncertainty in attempting these definitions, not a conclusion in either direction.

I am going to do my best to refrain from further SYNTHing here, as my intention now is to provide reasonable basis for what might be included for a generalized footnote (as mentioned in the comments below) to be added to pages marking transcription of glottal stops with purported secondary articulation; but if I cross that line again, please do not hesitate to call me out for it. In any case, I promise to clean up any mess I have made, but would prefer to resolve this before doing so, so that I don't have to scrub through the pages multiple times. I've rolled them back now and removed my disputed tags. There is no reason for me to hold the corrections hostage. If we can find consensus for the footnotes, great, but that's on me to insert later.

 ——— 

1. Laver (1994), Principles of Phonetics doi:10.1017/CBO9781139166621, p. 206:

Since a glottal stop (mentioned in the earlier section on nil phonation) by definition requires complete closure of the vocal folds, it is physiologically impossible to make a voiced glottal stop.

Takeaway: Laver treats a glottal stop as a state of complete glottal closure, excluding other phonation types as a physiological impossibility

 ——— 

2. Ladefoged & Maddieson (1996), SOWL, pp. 354-355:

As noted at the beginning of this chapter, the standard phonetic definition of a secondary articulation is that it is an articulation of a lesser degree of stricture accompanying a primary articulation of a higher degree ... It therefore follows, given a three-way partition of degree of stricture, that only the combinations of closure + open approximation, and close approximation (frication) + open approximation remain as ways of combining a primary and a secondary articulation. In other words, secondary articulations will always be approximant or vowel-like in their degree of stricture. ... The same considerations apply to secondary articulations, but the difficulties of demarcating comparable boundaries between a primary and a secondary articulation are greater than for doubly-articulated stops. Formation and release of a closure provide definite landmarks, but approximant articulations lack comparable boundaries. Nonetheless we feel that a useful distinction can be drawn between a consonant with a secondary articulation and a sequence of a consonant and an approximant. In practice, this distinction can be difficult to make, and many published descriptions of languages are written without such a distinction in mind.

Takeaway: Ladefoged & Maddieson treat secondary articulation as additional constriction which displays vocoid features on top of the consonants they modify, while also noting the challenge and ambiguity of distinguishing secondary articulation with consonant-approximant sequences

 ——— 

3. Esling, Fraser & Harris (2005), Glottal stop, glottalized resonants, and pharyngeals: A reinterpretation with evidence from a laryngoscopic study of Nuuchahnulth (Nootka) doi:10.1016/j.wocn.2005.01.003, p. 386:

In most modern phonetic literature, therefore, a glottal stop is defined simply as a closed glottis or tightly closed glottis without any reference to the ventricular folds, arytenoid cartilages, or supraglottal cavity activities. In our ongoing laryngoscopic research, glottal stops have been observed to occur on a continuum from a weakly constricted glottal stop to a strongly constricted glottal stop, but the most common realization of glottal stops that we have encountered across a number of languages includes an adduction of the arytenoid cartilages, a complete adduction of the vocal folds, a partial adduction of the ventricular folds, and moderate narrowing of the laryngeal vestibule through its epilaryngeal sphincter mechanism.

Takeaway: Esling et al. note prior literature definitions of glottal stops lacking supraglottal features, but also have found that the reality is more complicated than those reductive definitions; regardless, it remains confined to laryngeal space

 ——— 

4. Borroff (2005), Articulatory Phasing of Glottal Stop, p. 77:

Oral consonantal gestures are temporally anchored with respect to surrounding gestures by means of the phasing relations holding between it and those surrounding gestures. Because glottal stop does not participate in these phasing relations, it is not temporally anchored and is free to vary in temporal position.

Takeaway: Borroff notes glottal stops are fundamentally different from oral consonants and their relation to surrounding sounds, having significant freedom in when they may occur

 ——— 

5. Shaw, Campbell, Ehrhardt & McKay (2005), Patterns and timing of resonant glottalization in Nɬeʔkepmxcin, p. 222:

...although phonetically grounded conditions and markedness constraints undoubtedly play important roles, the basic generalization is that glottal timing is independent of oral articulation in the phonetic realization of glottalized resonants in both onset and coda positions.

Takeaway: Shaw et al. corroborate the previous point; glottal timing is independent from oral articulation, particularly in resonants

 ——— 

6. Borroff (2007), A Landmark Underspecification Account of the Patterning of Glottal Stop doi:10.7282/T34F1PMJ, p. 82:

Consequently, glottal stop will also often lack the burst generally associated with release of constriction at other places of articulation. This lack of a release burst and formant transitions for glottal stop is interesting, particularly in the light of the important role these cues play in stop identification at other places of articulation.

Takeaway: Borroff notes glottal stops lack the release aspect of typical stops, distinguishing their constriction type from oral consonants

 ——— 

7. Bird, Caldecott, Campbell, Gick & Shaw (2008), Oral–laryngeal timing in glottalised resonants doi:10.1016/j.wocn.2007.10.005, p. 3:

Since continuants do not need to rely on a short release burst to transmit cues (unlike obstruents), there is no need to strongly anchor the two articulatory gestures, and timing should be variable.

Takeaway: Bird et al. again corroborate the previous points; glottal timing is independent from oral articulations, and continuants further allow for this freedom

 ——— 

8. Maddieson (2013), WALS Online doi:10.5281/zenodo.13950591, Chapter 7:

Glottal stops may also occur with what are called "secondary articulations" such as labialization, so that /ʔʷ/ has a similar modification to that found in /kʷ/; similarly for palatalized /ʔʲ/. The sound /ʔʷ/ is obviously quite similar to the glottalized resonant /w'/, and either of these sounds might be analyzed as a sequence of two consonants, /ʔ/ and /w/. Arguments must be sought to determine which of the potential interpretations yields a more satisfactory overall view of the language’s structure. Because it has a contrastive glottal stop and other palatalized consonants, Hausa is here interpreted as having a palatalized glottal stop rather than a glottalized palatal resonant; thus the word for ‘daughter’, orthographic ’ya, is interpreted as /ʔʲaa/ rather than /j'aa/.

Takeaway: Maddieson notes a labialized glottal stop vs. a glottalized labio-velar resonant is largely an analytical matter of a particular language's phonological structure, rather than a necessarily strict distinction; in either case, corroborating what Maddieson states with Ladefoged in SOWL, the sound may also be analyzed as a consonant-approximant sequence

 ——— 

9. Ladefoged (posthumous) & Johnson (2015), A Course in Phonetics 7th edition, pp. 158 & 286 (also 6th edition, 2011, pp. 150 & 275):

The IPA diacritic to indicate creaky voice is [ ̰] placed under the symbol. Hausa orthography uses an apostrophe (’) before the symbol for the corresponding voiced sound, thus contrasting y and ’y. The Hausa letters y and ’y correspond to IPA [j] and [j̰].
Glottal Stricture specifies how far apart the vocal folds are. Languages make use of five possibilities: [voiceless]; [breathy voice], as we saw in languages such as Hindi; [modal voice], which is the regular voicing used in every language; [creaky voice] in languages such as Hausa; and [closed], forming a glottal stop.

Takeaway: Ladefoged & Johnson analyze the same consonant Maddieson treats as a palatalized glottal stop instead as a creaky-voiced palatal approximant; they also note the glottal stop as a distinct phonation type, being complete glottal closure, and contrast it with creaky voice

 ——— 

In sum: the given sources reflect a lack of consensus on a singular definition, highlighting the difficulty in differentiating glottal stops and laryngeal settings in relation to oral consonants, and indicate that different phonetic analyses may arise depending on how one chooses to segment based on contrast of other sounds within a given language ~ oklopfer (💬) 03:49, 21 January 2026 (UTC)[reply]

I have to say, I don't understand how thus is any different from the ambuguity between say [ⁿt], [nᵗ] and [nt]. I don't think anything special is needed here other than following the transcription of the source. Stockhausenfan (talk) 12:18, 23 January 2026 (UTC)[reply]

[ⁿt], [nᵗ] and [nt] are all supralaryngeal and take place at the same point of articulation, plus they have clear acoustic landmarks. The ambiguity there is primarily surrounding segmentation within a single articulatory system.

In the case I've highlighted here, by contrast, this is about laryngeal constriction paired with a supralaryngeal articulation, specifically lacking clear acoustic timing and cues like oral consonants (including release bursts), and with the laryngeal and supralaryngeal articulations decoupled both mechanically and temporally from one another (lack of interdependence). There are significant differences in those separate processes in comparison to prenasalization, stop release, and nasal-stop clusters; the ambiguity that arises is quite different, as is the analysis required.

I agree that we should follow source transcriptions, but given that the literature itself highlights major ambiguity that arises from potential analysis of laryngeal behaviors, including specifically noting that glottal stops do not behave like oral consonants in a variety of ways, a brief explanatory note seems reasonable if not important to alert readers that alternative analyses exist. ~ oklopfer (💬) 13:33, 23 January 2026 (UTC)[reply]

My own two cents are that we should follow the standards of transcription already established in the literature. This calls to mind the use of /r/ in traditional transcriptions of English-- although most dialects of English do not have that phoneme as it is specified in the IPA chart, that transcription is still what is conventional. I feel like this is similar, because even if [ʔʲ ʔʷ] are phonetically impossible, their use in phonemic transcriptions is still conventional for many languages, and it's not up to us as Wikipedia editors to decide which academic transcriptions are "wrong" and which ones are "right". Perhaps a middle-ground could be achieved by adding a footnote to these sounds explaining this: while they are technically impossible to produce by the definition of a glottal consonant, they are still notated as such for convenience's sake? ellaminnowpea ^{(371 💬)} 20:01, 20 January 2026 (UTC)[reply]

I would be in favor of a footnote explaining this, perhaps showing and describing alternate transcriptions/interpretations. It might be best to turn it into a template as well to avoid bloat; I made this alteration on just under 50 articles yesterday between the palatalized and labialized versions, so a standard insertion that can be placed on all of them would be good. I am willing to go through and revert all of them (so that others don't have to dig through and clean up after me) in place of said new template if we can agree on this course of action. ~ oklopfer (💬) 20:10, 20 January 2026 (UTC)[reply]

I highly recommend that. Especially since most publications and most linguists do still transcribe these sounds using the "impossible sound" transcriptions. Use the footnotes instead of changing the symbols, and do revert all the other edits where you changed these symbols. LanguageNerd876 (talk) 00:15, 21 January 2026 (UTC) Blocked sock. Stockhausenfan (talk) 16:29, 23 January 2026 (UTC)[reply]

I'm not sure I agree. There are plenty of instances where it is not convention to transliterate or interpret the given linguistic phrasing. The first thing that springs to mind is the Proto-Slavic yers (ь and ъ) even though they are typically understood to be /ɪ/ and /ʊ/, respectively. This contrasts with Americanist notation being "IPA-ified" by Indo-Europeanists. If the convention of literature is to transcribe the sound as /ʔʷ/, then it should remain; if other independent sources transliterate, I would be more open to that. It's worth noting that I don't think Oklopfer has adequately established why their interpretation of the sound is true; my read is that the "little imagination to interpret" this sound as [ʔw] is tantamount to original research, especially since they didn't use that same interpretation in the original edit; they used [ˀw], which appears to me as a singleton, not a cluster. ThaesOfereode (talk) 01:17, 21 January 2026 (UTC)[reply]

I agree with the last part, but I'm a bit confused. It sounds like you agree with me but yet don't. I have seen literatures that transliterate and transcribe those sounds as /ʔʷ, ʔʲ/, despite their phonetic "impossibility" here. And like I said, I do agree with him in that the solution should be a "middle-ground" by using the footnotes, instead of just a broad, prescribed transcription. LanguageNerd876 (talk) 01:33, 21 January 2026 (UTC) Blocked sock. Stockhausenfan (talk) 16:29, 23 January 2026 (UTC)[reply]

I apologize for the lack of clarity on my part. What I'm saying is that I'm not certain footnotes should be used as a middle ground because they will still apply the concerns I have about OR when Oklopfer has not firmly established that the sound is impossible or, if so, what sound it actually is. As something of an aside, I'm not really convinced that this sound is strictly impossible either. I never got into articulatory phonetics in undergrad beyond the basics, but it seems to me that the sound would matter more on the release rather than on the closure. It feels like it's akin to arguing that /pʷ/ is an impossible sound because /p/ is already a labial sound so rounding it would not make it phonemically distinct; the literature, however, disagrees. ThaesOfereode (talk) 02:10, 21 January 2026 (UTC)[reply]

Hopefully my comment above with source examples clears up some of this. I agree that my claim of impossibility is too close to OR per SYNTH, since no source states it directly. Laver (1994) does say 'impossible' for phonation in general on a glottal stop, but not secondary articulation specifically. I do still think a general footnote explaining ambiguity of the analyses, as Maddieson (2013) states outright, is important. ~ oklopfer (💬) 06:49, 21 January 2026 (UTC)[reply]

@Oklopfer I did just find the citation for the Telue Gelao language. In fact according to it, you were correct in its transcription listed as /ˀj/ and not /ʔʲ/. Here is the link if you're curious. LanguageNerd876 (talk) 09:23, 21 January 2026 (UTC) Blocked sock. Stockhausenfan (talk) 16:29, 23 January 2026 (UTC)[reply]

"phonation in general... but not secondary articulation specifically" – phonation and secondary articulation (using the lips and tongue) are independent of each other, not general and specific. But thank you for referring back to sources. Freelance Intellectual (talk) 09:51, 24 January 2026 (UTC)[reply]

I used a poor choice of wording, my point was just that Laver does use the word 'impossible', but not in the case of secondary articulation. As the other sources I provided showed and I acknowledged above, glottal stops are largely independent from supralaryngeal articulation. ~ oklopfer (💬) 13:49, 24 January 2026 (UTC)[reply]

I have created a template in my userspace for this: {{User:Oklopfer/Template:2art?}}

I need to improve the documentation before it is ready for mainspace, but I would appreciate reviewers for the content, and any suggestions to improve if necessary. ~ oklopfer (💬) 17:52, 23 January 2026 (UTC)[reply]

This seems pretty reasonable; it explains the issue concisely without editorializing or imposing a particular transcription. Stockhausenfan (talk) 18:14, 23 January 2026 (UTC)[reply]

I think this is a good idea. I suggest making this compatible with {{sfn}} and {{harv}} templating for those (like Nizaa language) which use that as their standard referencing. ThaesOfereode (talk) 22:19, 23 January 2026 (UTC)[reply]

I will try to make an additional parameter or parameters for that level of separation; grouping will get a little complicated, but I have some ideas for resolving that. I could have a bibliography-only parameter, but it would mess with alphabetization of sources in those sections. ~ oklopfer (💬) 23:19, 23 January 2026 (UTC)[reply]

I think in most cases, this footnote would be excessive WP:DETAIL and WP:SYNTH. It absolutely would be in the case of Hokkien phonology, which is what brought me to this discussion, where it's not even about a phoneme but just a side comment. However, the content in the proposed template should absolutely be included in Wikipedia and I think a much better place to put this would be in the articles glottal stop and glottal consonant, neither of which currently mention these issues. Freelance Intellectual (talk) 10:02, 24 January 2026 (UTC)[reply]

@Freelance Intellectual Can you explain what makes it SYNTH atp? Maddieson makes the broad generalization:

Glottal stops may also occur with what are called "secondary articulations" such as labialization, so that /ʔʷ/ has a similar modification to that found in /kʷ/; similarly for palatalized /ʔʲ/. The sound /ʔʷ/ is obviously quite similar to the glottalized resonant /w'/, and either of these sounds might be analyzed as a sequence of two consonants, /ʔ/ and /w/. Arguments must be sought to determine which of the potential interpretations yields a more satisfactory overall view of the language’s structure.

That covers the first sentence of the template, and really could be applied anywhere which notes secondarily articulated glottal stops. The second sentence just summarizes how glottalized resonants are described in the literature, which underlies the analytic ambiguity Maddieson explicitly notes. ~ oklopfer (💬) 14:11, 24 January 2026 (UTC)[reply]

Separate from my previous reply, I understand that the second sentence of the template may be more DETAIL than necessary for many articles, but not particularly the first sentence. Perhaps as you suggest the whole block could be placed in a section of glottalization, glottalic consonant, or glottal stop, and we could have a redirect glottalized resonant which the first sentence of the template uses instead of glottalized and resonant separately. Then the further detail summarizing the ambiguity can be kept to a main section, and the footnote becomes only a reference to Maddieson. ~ oklopfer (💬) 14:30, 24 January 2026 (UTC)[reply]

In the case of Hokkien phonology, the first sentence of the footnote would be nonsensical, suggesting that [ʔʲj] can instead be analysed as [ʔjj]. Maddieson actually uses // not [], but changing the footnote would mean it's still nonsensical, giving information about phonemic analyses at a point where the article is discussing phonetic detail of allophones.

SYNTH: '"A and B, therefore, C" is acceptable only if a reliable source has published the same argument concerning the topic of the article.' In this case, A is a phonetic/phonological analysis of a particular language, B is Maddieson's argument, and C is the suggested existence of an alternative analysis for the given language. To put it another way, Maddieson's point also applies to glottalised consonants as well as phonemes following a glottal stop – should those also require this footnote?

More broadly, pretty much all scientific concepts have some degree of controversy, not just specialised ones like "glottal stop with secondary articulation", but also foundational ones like "phoneme" or "word". These do not require a footnote in every article where they are used, but instead controversial aspects are discussed in an appropriate section of an appropriate article. This isn't specific to linguistics, either, and also applies to concepts like "species" and "planet". For example: Phoneme#Non-uniqueness of phonemic solutions, Word#Definitions, Species#The species problem, Definition of planet#Ongoing controversies. Freelance Intellectual (talk) 12:53, 25 January 2026 (UTC)[reply]

First, I've changed the notation from phonetic to phonemic in the sentence regarding Maddieson's statement as to not be presumptively reanalyzing what he is saying.

However, I think your SYNTH argument does not hold here; nowhere is there a definitive claim of "therefore C", only that Maddieson notes the possibility of such analyses, explicitly. More importantly, Maddieson does not say it applies for glottalized consonants in general: https://wals.info/chapter/7; that interpretation extends beyond the scope of what Maddieson actually discusses, and is trading out his narrow ambiguity for a universal one, when he never makes such a claim. He is very specifically talking about the analytic overlap and ambiguity of secondarily articulated glottal stops and glottalized resonants.

That aside, I recognize the desire to use the footnote conservatively (if at all), and I agree we should figure out the proper place to put it in a main article. ~ oklopfer (💬) 13:40, 25 January 2026 (UTC)[reply]

WP:SYNTH applies to implied conclusions. Maddieson accepts the validity of glottal stops with secondary articulation, and this source should not be used to cast doubt on analyses of particular languages. However, I'm glad we've agreed on putting the discussion of analytic overlap and ambiguity in a main article. I would suggest a section on secondary articulation in glottal stop. @LWG, ThaesOfereode, EllaMinnowPea371, and Stockhausenfan: any further comments? Freelance Intellectual (talk) 07:39, 26 January 2026 (UTC)[reply]

I think I agree that glottal stop would probably be the best place to put it, but want a few more second opinions. Glottalic consonant and glottalization both feel like they could potentially be viable but don't feel like they have as much room in their scope to have it placed.

@Kwamikagami as well, any opinions on where the content from {{User:Oklopfer/Template:2art?}} best belongs? ~ oklopfer (💬) 15:26, 26 January 2026 (UTC)[reply]

I apologize for not keeping up with this very closely over the past few days. That said, reading over, I am a little leery of adding these addenda to the page. Our job as editors is not to interpret for the audience but to make legible extremely technical speech. Especially as Freelance here has pointed out a defensible SYNTH concern and there was a puppet in our midst, I think the template footnote is a little overly ambitious as it stands. I think it would be a good idea to add these citations to the glottal stop (and/or associated pages, as above) for interested readers though. ThaesOfereode (talk) 15:30, 26 January 2026 (UTC)[reply]

Just looking at Mufian language, I see that the source the article cites transcribes that phoneme as [ʔʷ], described as "labialized glottal". Are you saying that the source is erroneous? If so, that's not Wikipedia's role to correct, unless another source exists that transcribes it differently (unlikely for a language this obscure). -- LWG ^{talk (VOPOV)} 20:23, 20 January 2026 (UTC)[reply]

I think there may be some confusion of 2ary articulation and phonation. A glottal stop cannot have any other phonation (unless it is lax and not actually a stop) because the glottal closure precludes it. But it could have various releases such as aspiration, and there is no restriction on what other articulators do elsewhere in the mouth, apart from they can't be too close to the glottis. We might conceivably have labialized, palatalized and velarized glottal stops. (I don't know if pharyngealization would work or not, but expect that it could.) There is therefore no reason to not take claims of e.g. [ʔʷ] at face value. — kwami (talk) 08:00, 26 January 2026 (UTC)[reply]

Khan (2015: 157) describes an 'emphatic' glottal stop in Aramaic that is presumably uvularized or pharyngealized. Krause (2018: 3) describes the same for the glottal stop and /h/ in the Arabic word Allah.

(Khan: Domains of Emphasis, Syllable Structure and Morphological Boundaries in the Christian Urmi Dialect of Neo-Aramaic. In Khan and Napiorkowska, eds., Neo-Aramaic and its Linguistic Context.

Krause: Transkription Deutsch-Arabischer Diskurse. MuM-Multi II, TP Hamburg.) — kwami (talk) 08:11, 26 January 2026 (UTC)[reply]

Please add reliable sources. Bearian (talk) 04:07, 23 January 2026 (UTC)[reply]

Hello all! I've been working on the Glossary of sound laws in the Indo-European languages for about two years now and as I close in on bringing it to featured list candidacy (pressing forward on my goal of getting Indo-European sound laws to Good Topic status), I want to ask other editors what they think the Indo-European sound laws page's purpose is. It's currently just a massive, mostly uncited set of tables and the useful content seems overall better suited for a page like Proto-Indo-European phonology. I've been expanding the glossary with the idea that it will be used to define terms for the lay audience and find use as an anchoring tool for links either before a page is made or because a full page may not be warranted (e.g., Kuiper's law), so it would be redundant to put that information on the "Indo-European sound laws" page. I also don't think the glossary contents should be migrated to the article page; I think the glossary/list format is appropriate for the content.

My two cents on the issue is that the page doesn't really have a purpose. Given the size of the page and the relative traffic (3,186 pageviews this month), I think it's best being stripped for its best parts to be merged with Proto-Indo-European phonology and used as a redirect to the glossary. Open to other ideas though; some more experienced editors or scholars might have a better idea to do with the page. ThaesOfereode (talk) 20:46, 26 January 2026 (UTC)[reply]

I think the purpose of the page is well served by your commendable project. Indo-European sound laws should be deleted. The only work with such a title is, as far as I know, Neville Collinge's the Laws of Indo-European which does much the same as your Glossary. The only other meaning of the term I could imagine is one of Pre-Indo-European sound laws or sound laws within PIE. There is such a section in the aforementioned article, but it can definitely be merged into PIE phonology, unless that article were to expand to an unreasonable size. I'll mention that the phonology article is woefully inadequate. Aspets (talk) 12:34, 28 January 2026 (UTC)[reply]

Thanks for the kind words. There are a few other works that cover the topic – Lubotsky's Sound Law and Analogy (R. S. P. Beekes's Festschrift) and Probert's Laws and Rules in Indo‐European come to mind (both cited in the glossary) – but they do not cover it as a cohesive whole, rather different authors typically tackle one or two laws at a time. For the other meaning, I don't think that it is captured by that title (maybe "Pre-Proto-Indo-European sound laws" or "Proto-Indo-European sound laws" would work), but even if it were, this is still captured by the glossary at this point.

Yes, the phonology article is in poor shape; frankly, most Proto-Indo-European articles are in pretty rough shape, which is part of the reason I have the Dura Lex project. Looking at it, I'm uncertain what I would even want to keep from there. My initial thought was maybe the tables, but they're completely uncited and don't capture the very scope most sound laws capture: narrow circumstances where changes occur (e.g., Cowgill's law of Greek). Curious if you think there's much worth salvaging. ThaesOfereode (talk) 22:59, 28 January 2026 (UTC)[reply]

Indo-European sound laws could be renamed to Indo-European sound correspondances with only the tables kept. But as you mention, they would be a lot of work to reference completely (but it's probably possible, the information exists in handbooks). Of course, "[a] fact or claim is "verifiable" if a reliable source that supports it could be cited, even if there is no citation for it in the article at the moment". Aspets (talk) 12:00, 29 January 2026 (UTC)[reply]

Possibly, but Indo-European sound laws also seems like something that could be on Wiktionary (in the appendix namespace)? But I agree a lot with your earlier comment. I'd say that Indo-European sound laws has been successfully couped by the glossary, and might as well redirect there, unless someone really likes the idea of sourcing the table. //Replayful (talk | contribs) 15:08, 31 January 2026 (UTC)[reply]

I'm not terribly attached to the tables to be honest, though I ran a few PIE terms blind through the table and they mostly work. Some of it doesn't make much sense though, like why are the consonants measured to Old Irish and modern English, but the vowels map to Proto-Celtic and Old English, respectively? Also not distinguishing the Tocharian languages is not great either. I may keep the tables somewhere in the draftspace for a little while to work on them for Wiktionary or maybe the phonology page, since someone worked hard on them and they mostly work, but they are going to need some restructuring. If no one objects to the move within a week of my original post (i.e., by the 2 Feb), I'm just going to be bold and go ahead with the change up: move tables to a user/draftspace (probably with Aspets's "sound correspondences" title) and then redirect the page to the glossary. All links headed to the page can remain per WP:NOTBROKEN so the transition should be fairly easy. ThaesOfereode (talk) 18:02, 31 January 2026 (UTC)[reply]

I'm fine with the move if you also link the draft here, so I could source some of the material. Aspets (talk) 16:28, 1 February 2026 (UTC)[reply]

Yep, I was planning to. I appreciate the willingness to help out. ThaesOfereode (talk) 16:31, 1 February 2026 (UTC)[reply]

NP. Aspets (talk) 16:34, 1 February 2026 (UTC)[reply]

I've taken care of this now. I've adopted your naming recommendation and moved the tables (and most of the text) to User:ThaesOfereode/Indo-European sound correspondences. If you forget the name for whatever reason, it can be access through my to-do list under "Miscellaneous". Feel free to let me know if there's anything I missed. ThaesOfereode (talk) 17:20, 3 February 2026 (UTC)[reply]

@Replayful: You may be interested in the above as well. ThaesOfereode (talk) 17:21, 3 February 2026 (UTC)[reply]

Yes, thank you. Aspets (talk) 08:48, 4 February 2026 (UTC)[reply]

Is it possible I can be of any help? Just today, coincidentally before knowing this thread existed or that a move had even occured while I was viewing the page, I decided to adopt the table to my own private version where I've added a few extra languages such as Proto-Italic, Proto-Brythonic, Welsh, Proto-Celtic consonants, Proto-Germanic, Norse, Icelandic, Old English consonants, Proto-Hellenic, Modern Greek, and planning more. AlphabetRevivalist (talk) 20:50, 3 February 2026 (UTC)[reply]

Absolutely! Feel free to add reliably sourced material to the new draft page here. If the page becomes well-sourced enough, I will move it to the mainspace under the new title. I'm happy to hear other editor's opinions, but I think we should plan to make the table somewhat more even than it was before; that is, using all proto-languages (using the neo-traditional model) rather than mixing Modern Greek, Proto-Italic, and Old Norse. I think that will make it more legible/accessible to the lay reader. ThaesOfereode (talk) 21:06, 3 February 2026 (UTC)[reply]

I'm not entirely sure how to source them other than Wiktionary examples, I'm very new at this game.

Do you propose including only proto-languages, without including attested languages? My version includes everything on the table before (or rather it will, it's still incomplete and I probably won't work on the rest today as I've been working on it all day already) in addition to the languages I listed above. AlphabetRevivalist (talk) 21:12, 3 February 2026 (UTC)[reply]

Also, my version is entirely in IPA, so a bit less cluttered and more consistent with sounds between languages than the original which primarily uses orthography and Romanizations. AlphabetRevivalist (talk) 21:21, 3 February 2026 (UTC)[reply]

I recommend you read WP:V which is sorta required reading for editing Wikipedia. The reason we're moving the tables off of the "main space" is because they're unsourced. If we can find sources, then we can put it back up. You can help, but Wikipedia does not allow any original research (also good reading), so you'll have to find something which backs your "claims" up. Aspets (talk) 08:47, 4 February 2026 (UTC)[reply]

This was a super dumb change. Users and readers alike have been consulting those tables for 20 years. If the concern was that the content was unsourced, the appropriate fix was to add sources. And why was this done without even a WT:AFD? --Victar (talk) 18:48, 6 February 2026 (UTC)[reply]

I can understand your frustration. For what it's worth, development of the original page (with sources) is underway here, if you'd like to contribute. We believe the tables may be useful, either here or on Wiktionary, in a page dedicated to sound correspondences rather than sound laws, provided that they can be sourced. The page was functionally unsourced, lacked real clarity in the topic at hand, and may be better suited for consultation on Wiktionary rather than as a distinct Wikipedia article. We are happy to get your input on the linked draft's talk page. As for AfD, I suppose you can consider the move something of a bold edit, but there was, I believe, reasonable consensus here. I didn't think to take it to AfD because it wasn't really a deletion per se; in any event, it seemed better to get input from editors who would be familiar with sound laws rather than non-specialized editors who would see several sources with the general topic of "sound laws" and conclude that the page should remain as is. Sound laws are not coterminous with sound correspondences. There was agreement that the tables might be worth something in the long term, which is why we maintained the data therein. I understand you may have objections and I'm happy to discuss them here or at the draft's talk page in a respectful manner. ThaesOfereode (talk) 19:13, 6 February 2026 (UTC)[reply]

In the interest of ecumenism, would it be better if we simply moved that page to say sound correspondences and redirected the laws page to the glossary (with a hatnote about the redirection on the glossary). Would that be sufficient to address your concerns? ThaesOfereode (talk) 19:21, 6 February 2026 (UTC)[reply]

In milling this over a bit, I think this might be the right approach. @Aspets, Replayful, and AlphabetRevivalist: thoughts? ThaesOfereode (talk) 19:47, 6 February 2026 (UTC)[reply]

I definitely agree that it should be titled "Sound Correspondences" rather than "Sound Laws". AlphabetRevivalist (talk) 19:49, 6 February 2026 (UTC)[reply]

Yes, I agree. In retrospect, I wish I had considered this option, but hindsight is 20/20, right? ThaesOfereode (talk) 19:57, 6 February 2026 (UTC)[reply]

Anything is fine by me, I don't have too strong opinions about anything //Replayful (talk | contribs) 01:05, 7 February 2026 (UTC)[reply]

I think it'd be better to keep it off the main space, since it is sorely lacking in sources. Ideally we would bring it back piece-meal, as each language column gets sourced. For the material which cannot be sourced, I agree that wiktionary could be a resting place for it (I'm thinking of medial stages between two well-documented or reconstructed stages). Aspets (talk) 17:50, 7 February 2026 (UTC)[reply]

I disagree with bringing it back piecemeal. We need that table somewhere. It can be populated with sources as we go. I honestly don't care where it lives but the table was of immense usefulness. I'm going to paste what I sent @ThaesOfereode below:

The loss of the sound law table is tremendous. It was incredibly useful for research, at least for an at-a-glance conspectus of correspondences. I disagree deeply and absolutely with the redirect to the glossary which is by no means useful to navigate. Even if the table was largely unsourced, it was for the most part highly accurate. The difficulty with citing all the changes in the table is precisely proportional to the difficulty an individual must now undergo in order to find the sound correspondences in the very disparate literature. It served a great purpose and its loss is an unthinkable injustice. Consider that the helpful links to the old page here are now broken and no longer answer the question of where to find a comprehensive chart of sound correspondences. This is not the only example. These tables were used by a large amount of people, and for good reason. I implore you to reconsider the decision to redirect for the time being. (Or at least move the old tables elsewhere) Arxandr (talk) 00:16, 13 February 2026 (UTC)[reply]

We have a page for the near-close near-back unrounded vowel [ɯ̽], which has low visibility due to its exclusion from our {{IPA vowels}} chart. A handful of pages link to it, but it largely flies under the radar: [2] (pageview stats, with some others for comparison)

I think we should do one of two things with it, and I'd like to know thoughts on which option others find preferable.

• Option 1: we add it to the vowel chart as is shown in {{IPA vowels/sandbox}}, adjacent to its rounded counterpart
• Option 2: we merge it into a section of its close back counterpart [ɯ]; this would be similar to the near-close centrals [ɪ̈] & [ʊ̈] residing in sections of their close central counterparts and the near-open backs [ɑ̝] & [ɒ̝] in sections of their open back counterparts, and none of which we list in the chart

From my perspective, both options have caveats:

• Option 1 feels it may be giving undue weight, as the occurrence listing currently only has 7 languages; though [ɤ̞], which we do have in the chart, has the same amount in its occurrence table, and of course [ɶ] has less than half of those. However, despite their low occurrences, the pageview stats I linked above show they still get far more visibility than [ɯ̽] does.
• The other examples I gave in option 2 are differing from their main pages in height alone, while [ɯ̽] is differing from [ɯ] in both height and backness. That being said, the difference between [ɯ̽] and [ɯ̞] is a fine line, as explained on the current page, and the sections for [ɑ̝] & [ɒ̝] also have mid-centralized [ɑ̽] & [ɒ̽] occurrences listed in them, so making the distinction of both height and backness may be negligible.

~ oklopfer (💬) 06:12, 29 January 2026 (UTC)[reply]

Better IMO to add it as a section to ʊ, which is often undefined for rounding and includes this vowel. — kwami (talk) 07:40, 29 January 2026 (UTC)[reply]

Handbook pp. 169-170 & 180-181 defines [ʊ] as rounded so I am hesitant about flattening it to include all three protruded, compressed, and unrounded ~ oklopfer (💬) 08:34, 29 January 2026 (UTC)[reply]

They do, but it's commonly used with agnostic or ambiguous rounding regardless. I think that's mentioned in Ladefoged's stuff, but I couldn't tell you where. — kwami (talk) 08:45, 29 January 2026 (UTC)[reply]

I'll try looking around for it tomorrow (anecdotally I know what you're saying is true). If I can find a cite and that's the decision from this discussion, I suggest we adjust the template chart to use {{IPA vowels/vowelsingle}} instead of {{IPA vowels/vowelpair}} for the vowel; that's approximately how it appears on the actual chart anyway ~ oklopfer (💬) 08:58, 29 January 2026 (UTC)[reply]

@Kwamikagami perhaps this is the passage you were thinking of?

[u] and [ʊ] are unrounded, [ʊ] often being pronounced with spread lips.

— Handbook, p. 43; Ladefoged's American English Illustration of the IPA

~ oklopfer (💬) 01:45, 5 February 2026 (UTC)[reply]

That's specifically for English, and we wouldn't want to say that in general [u] may be unrounded! No, I've come across other comments that the IPA doesn't make a rounding distinction here, because languages don't have a contrast, so ʊ may be used for either, and often may be ambiguous. — kwami (talk) 01:49, 5 February 2026 (UTC)[reply]

Keep digging then, I will... ~ oklopfer (💬) 01:58, 5 February 2026 (UTC)[reply]

I've found a few other examples, but they are all language-specific. Most other pieces talking about ambiguity of [ʊ] are about compressed vs protruded, not rounded vs unrounded. If anyone has an idea where to look, it would be much appreciated.

Or perhaps it would just be easier to merge to [ɯ], where rounding ambiguity need not be taken into account. There is already apparent overlap in the occurrence tables with the former and [ɯ̽]. ~ oklopfer (💬) 15:09, 8 February 2026 (UTC)[reply]

That might work better. The IPA does [currently at least] define it as a rounded vowel. — kwami (talk) 21:42, 8 February 2026 (UTC)[reply]

Just link it from [ʊ] in "see also". Aspets (talk) 12:03, 29 January 2026 (UTC)[reply]

Linking it at the bottom of an already long page is going to leave us in the exact same place we are now. We should either give it more predominant visibility in the chart template, or merge it to a section of a closely related vowel. ~ oklopfer (💬) 15:02, 29 January 2026 (UTC)[reply]

No, there's nothing which compels us to do so. Merging is only allowed if the vowel belongs to the same topic. The IPA vowel chart on wikipedia is already too cluttered compared to reliable sources, and this very rare vowel definitely does not belong on the template. Aspets (talk) 16:31, 1 February 2026 (UTC)[reply]

Merging is only allowed if the vowel belongs to the same topic.

I'm not really sure what this means. Both [ɯ] and [ʊ] can be argued to be relevant to the "same topic" as [ɯ̽], or on the contrary [ɪ̈] [ʏ̈] [ʊ̈] [ɑ̝] [ɒ̝] can all be argued to be not the "same topic" of the pages they are currently sections on.

The IPA vowel chart on wikipedia is already too cluttered compared to reliable sources, and this very rare vowel definitely does not belong on the template.

As I pointed out, [ɤ̞] is listed in the chart, and is arguably even rarer, as it is mentioned for less dialects (though the same number of languages, wherever we draw that line). Why include it and not [ɯ̽]? If the argument is along the lines of "it keeps the chart balanced", then adding the latter would only maintain that balance: {{IPA vowels/sandbox}}

Again, it is the only vowel with an independent page that is not listed in the chart. ~ oklopfer (💬) 23:26, 1 February 2026 (UTC)[reply]

I definitely think it should be merged somewhere. I guess I don't really care where, as long as it's cross-ref'd. I think ʊ makes the most sense. — kwami (talk) 02:56, 2 February 2026 (UTC)[reply]

Well, I think [ɤ̞] could definitely be removed.

Since all information must be sourced, we do not decide on our own as Wikpedians if we merge articles or not. It will depend on how the topic is handled by reliable sources. The two options you gave above seem not to be based on any treatment of the vowel by an authority, but I'd be happy to read any source you're able to find about the topic. I don't really have the time to look for one myself, but if we are to treat these vowels (and regarding on what pages) then there better be some. Cheers. Aspets (talk) 16:23, 3 February 2026 (UTC)[reply]

I think you have a misunderstanding of the merging process. See WP:MERGE, WP:PMG, and WP:PAM; we absolutely decide on our own as Wikipedians if we merge articles or not — who else would?

And again, we already treat the near-close central vowels and the near-open back vowels this way, rather arbitrarily. [ɪ̈ ʊ̈] are often treated as reflexes or realizations of [ɪ ʊ], but that is not where we have them listed. Is your suggestion that we WP:SPLIT these, creating even more hidden-from-view vowel pages?

Finally, both merge options are in fact treated as related by reliable sources; this is clearly provided on the current page, where [ɯ̽] is most often a realization of /ʊ/ or /ɯ/. Asking for evidence which already exists and is already provided, and instead turning a blind eye, is not an argument. ~ oklopfer (💬) 17:01, 3 February 2026 (UTC)[reply]

Well, I guess this brings us to an interesting dilemma: are the articles about phones or phonemes (or phonetic realisations of phonemes?)?

I'm sorry if I've come across as agressive or rude, that certainly wasn't my intention. But I do find my words to not have been terribly well chosen and phrased, reading them through again. Aspets (talk) 08:55, 4 February 2026 (UTC)[reply]

They can't be about phonemes, because those are defined by the language. There is no cross-linguistic /ʊ/. Basically, we cover phones that have dedicated IPA letters, plus some others we find notable; the IPA assigns letters to phones that are phonemically distinct in [some undefined number of] languages, again with exceptions. — kwami (talk) 09:01, 4 February 2026 (UTC)[reply]

I added some sources. She sounds fascinating. Bearian (talk) 04:21, 2 February 2026 (UTC)[reply]

Cross-posting from Wikipedia talk:WikiProject Apps — I'd appreciate input from editors here given the applied linguistics angle.

BoldVoice is an AI-powered pronunciation and accent coaching app for non-native English speakers. I work at the company (WP:COI disclosed), so I'm asking here rather than creating the article.

Key sources:

• Wired: "AI and the End of Accents" (Oct 2025)
• Inc.: "This Startup Is Putting an Accent Coach in Your Pocket" (Jan 2026)
• TechCrunch: In-depth profile (Jul 2021)
• Hawaii TESOL Journal (2023): Peer-reviewed app review
• National Foreign Language Resource Center (2025): Academic review covering segmental/suprasegmental features

The app focuses on phoneme-level pronunciation feedback and has been the subject of L2 acquisition research. Does this seem like it would meet WP:GNG? Full discussion at Wikipedia talk:WikiProject Apps#Notability check: BoldVoice (accent coaching app). ~2026-81303-3 (talk) 18:33, 5 February 2026 (UTC)[reply]

The article Generative principle has been proposed for deletion because of the following concern:

Tagged for notability concerns for almost 15 years. No other language has a reliably sourced article from which to translate. There's a bibliography, but not a single citation about this hypothesis. This has also aged poorly as an idea. Fails the relevant notability guidelines.

You may prevent the proposed deletion by removing the {{proposed deletion/dated}} notice, but please explain why in your edit summary or on the article's talk page.

Please consider improving the page to address the issues raised. Removing {{proposed deletion/dated}} will stop the proposed deletion process, but other deletion processes exist. In particular, articles for deletion allows discussion to reach consensus for deletion based on established criteria.

If the proposed deletion has already been carried out, you may request undeletion of the article at any time. Bearian (talk) 01:56, 12 February 2026 (UTC)[reply]

@Oklopfer and Kwamikagami: Should we mention palatalized velars as an alternative term or representation of post-palatal or pre-velar consonants, but mention they can have a palatal release instead, as I've done here and here? - Bᴏᴅʜı ***** Hᴀᴙᴩ** 02:02, 18 February 2026 (UTC)[reply]

I think we need better sources: I believe that [c] is not the same as [kʲ], but don't know how robust that distinction is in practice. Like the difference between [ɲ] and [n̠ʲ], it may just be a matter of the transcriber not being very precise. If that's the case, then IMO the palatal template should stick to describing a palatal articulation. — kwami (talk) 02:12, 18 February 2026 (UTC)[reply]

And the line in that distinction becomes even thinner in the case of [c̠]/[k̟] which BodhiHarp is referring to, but [kʲ] may still just be a velar plosive with approximant-like release. It is rather ambiguous. ~ oklopfer (💬) 04:34, 18 February 2026 (UTC)[reply]

If this is in your "wheelhouse", please add citations where needed. Bearian (talk) 16:18, 19 February 2026 (UTC)[reply]