Wikipedia talk:WikiProject Linguistics/Archive 20

Archive 15

←

Archive 18

Archive 19

Archive 20

Archive 21

Portmanteau vs Contraction

Each of the articles for portmanteau and contraction mentions the same distinction, that "contractions are formed from words that would otherwise appear together in sequence, such as do and not to make don't, whereas a portmanteau is formed by combining two or more existing words that all relate to a single concept". The article for contraction notes that there is some overlap, but neither article is sourced. There is some debate regarding the list of portmanteaus (which is a mess for a variety of reasons) on whether to include terms like vocoder (voice encoder), Juneteenth (June nineteenth), and and Spam (spiced ham). There are many such examples there which, taken literally as "words that would otherwise appear together" would qualify as contractions. Looking more closely at contractions, however, most of the examples suggest that they are mostly used in less formal speech rather than the formation of new words. Is there a source to clarify the distinction, and can the distinction be made more clear? HalJor (talk) 15:10, 5 May 2021 (UTC)

@HalJor: Off the top of my head, I think contractions are grammatical (e.g. "I'm'onna") and blends are lexical, but not sure of portmanteaux. — kwami (talk) 01:20, 6 May 2021 (UTC)

Crystal's dictionary has:

contraction (n.) A term used in linguistics to refer to the process or result of phonologically reducing a linguistic form so that it comes to be attached to an adjacent linguistic form, or fusing a sequence of forms so that they appear as a single form. The first kind of contracted form (or contraction) can be illustrated by I’ve from I have, haven’t from have not, and WANNA-contraction. The second kind is seen in French du, des from *de le and *de les respectively.

portmanteau (adj./n.) A term used in morphological analysis referring to cases where a single morph can be analysed into more than one morpheme, as in French au, aux, etc. (=*à le, *à les, ‘to the’). The item is called a ‘portmanteau morph’ (‘a portmanteau’), and sometimes, when it is equivalent to a word, a ‘portmanteau word’.

blending (n.) A process found in the analysis of grammatical and lexical constructions, in which two elements which do not normally co-occur, according to the rules of the language, come together within a single linguistic unit (a blend). In grammar, the process is illustrated by such syntactic blends as It’s his job is the problem, a combination of the sentences It’s his job and His job is the problem. In lexis, ‘blending’ is a common source of new words through abbreviation (though not all become standard), e.g. brunch, Interpol and Eurovision. The term is also used by some psycholinguists for a type of tongue-slip involving the fusion of two target words, e.g. swurse for swear + curse.

Bussmann's dictionary has:

contraction: Process and result of the coalescence of two consecutive vowels into a single long vowel: Gmc *maisōn > OE māra > Mod. Eng. more (→ synaeresis). Also generally, every form of lexical shortening, e.g. Eng. don’t for do not, Fr. au for *à le.

portmanteau morpheme [Fr. portemanteau ‘clothes-stand’] Term introduced by C.F.Hockett to denote phonomorphological units that blend several otherwise distinct morphemic units together, cf. Fr. au (= blend of à+le) which contains the meanings of ‘dative,’ ‘definite,’ ‘masculine,’ and ‘singular.’

blend (also amalgam, fusion, hybrid, telescoped word): In word formation, synchronic or diachronic crossing or combining of two expressions into a single new one. Blends may develop from an unconscious or unintentional misspeaking (→ speech error), e.g. in the blend of innuendo and insinuation to insinuendo, or through stylistic intent. In the latter case, a distinction may be drawn between (a) haplological blends (→ hapology [sic]) in which the last part of the first word and the first part of the second word are identical (networkhorse, californicate) or in which sound and syllable elements overlap (tragicomic, guestimate); (b) neologisms involving word splitting (= true blends) (motel, eurocrat, telethon); (c) analogous formations in which a base word is replaced by a similar sounding lexeme (vidiot < video+idiot); (d) orthographic variants that are recognized as blends only from their spelling (Ronald Raygun). Blends, in comparison with more usual compounds. tend to be formed spontaneously through the close association of two words and do not themselves usually serve as models for further compounds. Because most blends can only usually be understood in context, only a very few of them (e.g. the linguistic term Franglais), are adopted into everyday language. On syntactic blends, see Paul (1880) and Bolinger (1961).

So maybe some of the 'portmanteau' stuff should be moved to blend word? — kwami (talk) 01:34, 6 May 2021 (UTC)

Honestly most of it probably should be merged with blend word, save the section on portmanteau morph(eme)s. And honestly perhaps portmanteau morp(eme)s could be its own article altogether if properly expanded. Blends and the non-linguist/lay sense of portmanteau are the same thing and because of that the two articles have a ton of overlap. Umimmak (talk) 03:41, 6 May 2021 (UTC)

'Portmanteau morpheme' would replicate the 2nd sense of 'contraction' per Crystal, so probly either merge that into 'contraction', or move the 2nd sense out. — kwami (talk) 06:58, 6 May 2021 (UTC)

Portmanteau morphemes often originate from contractions, but not necessarily so. They can also be suppletive. It's a nice topic by itself, but the current section is a bit short for a standalone. –Austronesier (talk) 07:16, 6 May 2021 (UTC)

(edit conflict) I mean you might see French du be called a contraction of de+el, but that feels distinct from say the -o in Lat. amo being a portmanteau morpheme bundling together 1.sg.pres.ind.act as that would never be called a contraction. So I don’t think contraction should be the general article for portmanteau morphemes. Umimmak (talk) 07:18, 6 May 2021 (UTC)

They could go to either, as long as we don't have a content fork. Whether French aux should be under 'contraction' or 'portmanteau' I don't know. It should be at least mentioned under both, but one should be an explanation of a 'main topic' or 'see also' redirect to the other. — kwami (talk) 07:32, 6 May 2021 (UTC)

The blend vs. portmanteau portion of this topic has been previously discussed at Talk:Portmanteau and at Talk:Blend. (Pinging @DavidWBrooks, Vikom, Florian Blaschke, and TAKASUGI Shinji: from those discussions.)

Usage maven Bryan Garner has blend and portmanteau as synonyms (emphasis added):

A portmanteau is a type of luggage with two separate sections. A portmanteau word is formed by combining the sounds and meanings of two different words. Linguists also call such a word a blend.

Most portmanteaus merge the initial part of one word with the end of another: smog (smoke + fog) and infomercial (information + commercial). Others combine one complete word with part of another: docudrama (documentary + drama) and palimony (pal + alimony). Sometimes words with the same sounds are combined to create a pun: shampagne (sham + champagne).

— Bryan Garner, Law Prose #194

Garner goes on to mention Lewis Carroll's well-known portmanteaus chortle, galumph, and slithy (as well as the fact that Carroll coined the phrase portmanteau word as well, though not portmanteau of course). I could see Blend word and Portmanteau being one article, because I don't know what distinction there is to draw between the two, although Contraction feels different enough to be separate.

As for the first two, I believe that blend word is primary. I seem to have a sentimental attachment to portmanteau as a stand-alone article, although I'm not sure I can articulate a valid, policy-based reason to keep it separate. The only reason I can think of, is the Lewis Carroll connection, but that's probably a pretty weak reason, unless a lot more material could be found specifically about that. On Wiktionary, yes, separate articles; but here? Sadly, probably not. Perhaps this should be formalized with a Merge request (for Blend + Portmanteau; not Contraction); I'll probably create one, if someone else doesn't. Mathglot (talk) 19:51, 10 May 2021 (UTC)

Hi Mathglot! It sounds like portmanteau word should be merged with 'blend', but that portmanteau morpheme (in the sense of affixes and grammatical words) should be merged with 'contraction'. One as a device to create new lexicon, the other a phonological reduction of existing forms without necessarily a change in meaning. — kwami (talk) 00:44, 11 May 2021 (UTC)

Hi, kwami, yes, possibly; hadn't considered that aspect; perhaps Portmanteau will end up on the disambig page with '~ word' and '~ morph' as two of the entries, "Portmanteau morph" ends up merged to Contraction (grammar), and "Portmanteau word" ends up as a redirect to Blend word or to a section of it devoted to Carroll. (And I also found plenty of attestations of "portmanteau sentence", meaning intersentential code-switching; e.g., here, so that could be another DAB entry.) Or maybe "Portmanteau" should become a WP:Broad concept article, comprising word, morph, and sentence.

Perhaps we can brainstorm an Rfc/Merge question here, so that the language of the question will have the optimal chance to achieve a good outcome. This might be one of those, that requires an A-B-C-D-E series of possible choices for respondents. For the time being, I've been making a list of prior discussions about portmanteau–blend similarities/differences (over a dozen; with half of them at Talk:Portmanteau/Archive_1) and a list of prior discussants to invite to a possible forthcoming discussion (roughly 25, but many may be retired; discussions go back to 2002!). A principal issue for me, regardless who writes it, is about how to formulate a good Rfc/Merge question, making sure to cover all the bases, but not overcomplicating it. Maybe we could discuss that here. Mathglot (talk) 01:58, 11 May 2021 (UTC)

Yes, the "syntactic blends" mentioned in the defs above. Maybe that along with code-switching blends should be a third article? Lexical blends like "smog" are intentional (though I make new ones accidentally when I change my mind halfway through a word), but syntactic blends tend to be unplanned. Phrases and clauses as well as full sentences. — kwami (talk) 03:15, 11 May 2021 (UTC)

kwami: Update: I added redirect Portmanteau sentence to code-switching before I saw your reply. Mathglot (talk) 08:55, 11 May 2021 (UTC)

I find myself in agreement with: "I could see Blend word and Portmanteau being one article, because I don't know what distinction there is to draw between the two, although Contraction feels different enough to be separate." And with the French vs. Latin reasoning in the "So I don’t think contraction should be the general article for portmanteau morphemes" comment.

Side comment: I hope that "such syntactic blends as It’s his job is the problem, a combination of the sentences It’s his job and His job is the problem" doesn't appear in one of our articles. That looks like folk etymology to me, as It's his job is the problem is much more obviously derived from It's his job that is the problem, not from mashing two separate sentences awkwardly together. There are several dialects that routinely drop that. I cannot count the number of times I have had to add a that to a sentence in Wikipedia itself because it was painfully ambiguous; some editors habitually write without that even when they think they are writing with especial formality and clarity, just because they're used to the lack of it and are accustomed to accept the ambiguity as normal. But the results can be very confusing, especially for non-native English speakers.
 — SMcCandlish ☏ ¢ 😼  07:01, 12 May 2021 (UTC)

I suspect you're correct about that sentence (not a pattern I'm familiar with), and if so it's a bad example. What's the term for when you change your target utterance halfway through? We've got a couple things going on here:

• Transparent optional contractions, such as did not ~ didn't
• Opaque obligatory contractions, such as Fr. aux

(and intermediate cases)

• Lexical blends, such as smog
• Target-switching during an utterance, whether within or between languages

if not more. It would be nice if we could outline all these for navigation. — kwami (talk) 23:22, 12 May 2021 (UTC)

Category:Phonetics articles needing expert attention has been nominated for discussion

Category:Phonetics articles needing expert attention has been nominated for possible deletion, merging, or renaming. A discussion is taking place to decide whether this proposal complies with the categorization guidelines. If you would like to participate in the discussion, you are invited to add your comments at the category's entry on the categories for discussion page. Thank you. Peaceray (talk) 05:04, 18 May 2021 (UTC)

IPA transcription for Robert McNamara

I just added an IPA transcription of Robert McNamara's surname based on the pronunciation in this ABC News video. I'm far from an expert at IPA transcription, so I would appreciate it if someone could double-check my transcription. (If there's a better page for this kind of issue, let me know.) Rublov (talk) 20:55, 19 May 2021 (UTC)

You pretty much got it. Two quibbles: (1) the US dictionary convention of 2ary "stress" on the mar, which is not actually stress and not used by e.g. the OED, and (2) the vowel of that syllable, ambiguous for those of us who have the merry-marry-Mary merger, which Lexico.com (from the OED and Random House) says is the marry vowel. — kwami (talk) 23:09, 19 May 2021 (UTC)

Random House has got nothing to do with Lexico. Nardog (talk) 23:17, 19 May 2021 (UTC)

They're partly based on Dictionary.com, which I thought was from Random House. — kwami (talk) 23:19, 19 May 2021 (UTC)

Random House licenses its dictionary to Dictionary.com, but they have no financial affiliation. Lexico is a website run by Dictionary.com that hosts dictionaries licensed from OUP. Nardog (talk) 23:27, 19 May 2021 (UTC)

Thanks! Rublov (talk) 00:10, 20 May 2021 (UTC)

List of languages by number of phonemes

The newly created List of languages by number of phonemes could do with some attention (here's a sneak peek: the language listed with the largest inventory is Norman French). – Uanfala (talk) 12:55, 20 May 2021 (UTC)

Odd that they would skip the high and low ends.

Missing a column for tone or stress.

How do we rank languages by inventory size, when a different analysis could double or halve the inventory of any particular language?

English listed with 11 vowels, no stress. I count 21 vowels in our RP article. Maybe we could give a range? Would there be any point to an article that doesn't give a specific number of phonemes?

What qualifies a language for inclusion in the list? — kwami (talk) 04:39, 21 May 2021 (UTC)

Or in other words: unmaintainable language-fancruft. With or without Ubykh and Rotokas, I'll ignore it. –Austronesier (talk) 13:09, 21 May 2021 (UTC)

[w ʍ ɥ ɥ̊] RM

Participation is appreciated at Talk:Voiced labialized-velar approximant#Requested move 24 May 2021. Nardog (talk) 12:17, 30 May 2021 (UTC)

Cosmetics: * in templates

Hi, I noticed the asterisks look different, namely more centered than in many fonts, when put inside Template:PIE. E.g. *outside/*inside. It's a minor issue and probably not equally noticeable to every user, but if bored I may want go through all the articles using it to fix inconsistencies; before that I wanted to make sure I am doing it right and hear about other minor possible fixes I could do on the way. My first edit in this direction was apparently well received, most others will be hopefully simplier. The italics part is even more nitpicking, my understanding is that "*" shouldn't be in italics, unless the surrounding text is. My rationale for keeping the asterisk outside is that it isn't part of the foreign world and that reconstructed forms are sometimes rendered without templates and so with the same asterisks as normal text. When multiple forms are given, this way the code gets a bit more complicated; making the template somehow ignore this character could be simplier, otherwise a mention to leave it out in the template documentation may help. Other templates may also show the same problem. Any suggestions are welcomed. Personuser (talk) 13:45, 1 June 2021 (UTC)

@Personuser: Can you show a screenshot of what it looks like for you? Because the two look exactly the same on my system and looking at the very simple source of {{PIE}} I can't spot anything that would change the formatting – it only sets the lang attribute, adjusts wrapping as described, and adds an (as far as I can tell unused) CSS class. – Joe (talk) 09:13, 2 June 2021 (UTC)

@Joe Roe: Thanks for the feedback. Screenshots are a bit complicated for me at the moment. The outside asterisk appears to me a bit higher than the "o", the inside one lower than the dot on the "i" (with Unifont). It seems that when I tried to change the browser font (Firefox) I skipped some advanced settings, now I can make it appear almost the same (and probably the same with some more fiddling). Unless someone else finds this annoying I guess this is even less worth fixing than what I thought, keeping templates simple has its benefits. Personuser (talk) 19:28, 2 June 2021 (UTC)

Example sentences that aren't terrible

I've been frustrated with how Wiki Text handles example formatting, and I know I'm not alone in this feeling. We often have to hack together the markup and it doesn't always turn out so good regardless. A while back, I submitted a feature request on Wikimedia Phabricator which would allow continuous example numbering and cross-referencing. However, it hasn't gotten a reply–– Phabricator has a bit of a backlog, and this request probably sounds a bit niche. If you would support this idea or if you would like to add anything to the feature request, you can leave a comment here.

Alternately, if anybody has any handy tricks or secret wisdom about hacking together markup for nice examples, please do share! Botterweg14 (talk) 15:13, 16 June 2021 (UTC)

Different SIL codes, but reliable sources don't cover them as distinct language varieties

What to do about Assyrian Neo-Aramaic [aii] and Chaldean Neo-Aramaic [cld], given different SIL codes on non-linguistic grounds per this source[1]? They are two different churches and ethnic identities, but not two different language varieties. Most reliable sources distinguish between dialects of Northeastern Neo-Aramaic (NENA) spoken by Jews and Christians, but not between Assyrian and Chaldean dialects. As far as I can tell, it is impossible to write articles on either of these topics without a large amount of original research, since the vast majority of reliable sources discuss NENA rather than "Assyrian" or "Chaldean". (t · c) buidhe 09:38, 17 June 2021 (UTC)

@Buidhe: I moved [aii] to Suret language per Moseley and tagged it for merger with Chaldean. Are you willing to take on that chore? — kwami (talk) 17:51, 17 June 2021 (UTC)

Thanks so much! I have gone ahead and completed the merge. The phoneme table might be useful but only if one could clarify which dialect(s) it refers to. (t · c) buidhe 18:03, 17 June 2021 (UTC)

Opinions and comments needed

Opinions and comments needed here.

It is a renaming proposal of the Italian municipality currently spelled Codogné. It was proposed to change the spelling to Codognè. The reason of the proposal is that the Italian state websites listing Italian municipalities use the same accent (è) for every accented town name and that on the website of the town itself this is the most common spelling. The reason to reject the proposal is that Italian orthography handbooks and encyclopedies which distinguish between two different accents prescribe the current spelling (é) corresponding to the prescribed pronunciation in Standard Italian. Since both options are reasonable a large number of opinions is desirable to have a shared consensus.

Tsistunaɡiska (talk) 06:39, 30 June 2021 (UTC)

Etymology versus "origin of the ..."

There's currently a discussion on the MoS about whether etymologies only apply to words: Wikipedia_talk:Manual_of_Style#MOS:SECTIONSTYLE,_"etymology"_and_User_talk:Catchpoke. Catchpoke (talk) 19:41, 12 July 2021 (UTC)

RfC at Category:Criticism of political correctness

There is a request for comment at Category:Criticism of political correctness. Please see the discussion if you are interested. --Animalparty! (talk) 02:04, 19 July 2021 (UTC)

RfC: Where should so-called voiceless approximants be covered?

The following discussion is an archived record of a request for comment. Please do not modify it. No further edits should be made to this discussion. A summary of the conclusions reached follows.

A large majority of participants were in favour of including voiceless approximants in articles about voiceless fricatives. A minority favoured including in articles on voiced approximants. Much of the discussion revolved around the position of the IPA, with the majority arguing the IPA supports their position. Those that opposed the majority position argued that the IPA is not authoritative in classification matters and merely provides symbols, so where symbols appear in their charts is not definitive. This analysis was even accepted by some on the majority side. All participants agreed that the issue is controversial and no firm consensus can be found in the literature at the present time. Wikipedia should therefore be careful to make clear, in each affected article individually, that our placement of these items in articles does not imply that we support a particular position in this debate. SpinningSpark 15:57, 13 September 2021 (UTC)

---

In which articles should occurrences of sounds that have been described as voiceless approximants, such as [ʍ] in Scottish English whether, be included? 10:44, 27 July 2021 (UTC)

• In articles about voiceless fricatives. First, not all phoneticians accept that there are such sounds as voiceless approximants. The sounds described as voiceless approximants by those who accept them are, to those who do not, fricatives. The International Phonetic Association clearly belongs in the "there are no such things" camp, defining the symbol ⟨ʍ⟩ as representing a "voiceless labial-velar fricative" on the official IPA chart and exemplifying it with Scottish English whether, which the other camp considers an approximant. So it would be in violation of the NPOV policy if we took it for granted that there are such things as voiceless approximants.
  Second, since not all linguists accept that there are such things as voiceless approximants, reports of voiceless fricatives cannot be taken definitively to mean they are the kinds of sounds that would be considered fricatives even by those who do (which is not always clear-cut, as we shall see below), unless the authors make it clear or instrumental studies on the specific nature of the sounds are available—which are hard to come by especially for underdocumented languages. If we took it for granted that there are such things as voiceless approximants and created separate articles or sections for them, then we'd have to list sounds that we don't know for sure would be considered fricatives by all scholars alongside the ones we do, thereby implying something the sources don't say.
  Third, even some of those who accept voiceless approximants have noted that some sounds are difficult to classify unequivocally as either fricatives or approximants. Again, if we took it for granted that there are such things as voiceless approximants and described them separately, descriptions of those ambiguous sounds would have no place to go. Whether we placed them among fricatives or among "approximants", either way would be misrepresentative of the sources.
  These considerations leave us no option but to list them in articles about fricatives. But that doesn't mean we, editors of Wikipedia, which strives for verifiable accuracy and a neutral point of view, have to disregard the other perspective entirely. We can simply incorporate both points of view by listing them among fricatives and explicitly noting that they are described as approximants by those who accept them, as done here.
  I acknowledge that, in the future, the supposition of voiceless approximants may very well become mainstream. I just do not see evidence that it currently is—the English [ʍ] is described as a fricative as recently as 2019–2020 in popular textbooks—and we have duty to uphold WP:NPOV and WP:CRYSTAL. Nardog (talk) 10:44, 27 July 2021 (UTC)

References

Agree. Thanks, Nardog, for this nice exposition of the facts. In principle, I would take the stance of the International Phonetic Association as authoritative in terms of defining the mainstream understanding. It is not that they cannot be reasoned with when new evidence comes up - for example in 2005 they included the labiodental flap sound into the accepted inventory, after sufficient evidence was provided for its real-life existence, and after a lively academic debate about that matter. So, if the approximant faction succeeds in convincing the IPA, then the idea will become mainstream. Until then, we would be well advised to only mention the alternative analysis of some linguists in both articles, without giving the matter too much space. LandLing 11:16, 27 July 2021 (UTC)

Concur with LandLing, who said everything I would have, but more concisely. :-)  — SMcCandlish ☏ ¢ 😼  14:41, 30 July 2021 (UTC)

I applaud the decision to raise this tricky issue for discussion. It is noticeably confusing in Voiceless labiovelar fricative when we read "Features of the voiceless labial–velar fricative: Its manner of articulation is approximant". I think it is wise at this stage not to go against IPA policy, but on the other hand I think it would not be out of place to ask the IPA to give an opinion on a problematical area that falls within its scope. Ideally, someone should write a short article for the Journal of the IPA outlining the problem as it confronts WP, asking the IPA membership to comment and the Executive to consider whether a change in the IPA chart is called for. RoachPeter (talk) 14:34, 27 July 2021 (UTC)

Indeed. For now I slapped a "citation needed" behind the approximant feature in the article. I wonder, though, how practical it is to interact with IPA directly. Has this happened before? LandLing 15:44, 27 July 2021 (UTC)

Hmmm...ideally, the someone should be a prolific contributor to JIPA and former Secretary of the IPA, no? ;) –Austronesier (talk) 16:15, 27 July 2021 (UTC)

Ah, I see. Good idea! LandLing 17:41, 27 July 2021 (UTC)

I agree that inviting comment from the IPA would be useful, and I think it would be an interesting experiment in how Wikipedia can engage experts in helping to fill gaps in our coverage. Presumably the article RoachPeter suggests would take a form similar to Keating, Wymark, and Sharif (2019)? There the authors review transcription practices present in the literature, conclusions on the phonetic and phonological status of these sounds, and then propose particular actions for the IPA given their review. An article like Peter suggests would probably include some coverage of how the conflict affects Wikipedia coverage which we can connect to the Journal's interest in articles on the "practical applications of phonetics to areas such as phonetics teaching" given how the conflict makes it difficult to educate the public on these topics in a consistent manner. I assume we wouldn't make any outright proposal like Keating, et al., but instead we would invite further papers and proposals. One alternative would be to propose a special issue; though it would be substantially more work, an issue curating perspectives on the topic would be useful not only for us but the field generally. If anyone is interested in working on something like this, let me know and I can help with the literature search. — Wug·a·po·des​ 20:28, 27 July 2021 (UTC)

That'd be great, but it's going to take years. (It took three years for the IPA to consider whether to add a symbol for an open central vowel and eventually veto it.) In the meantime, can we come to a consensus on how best to discuss these sounds with the sources available right now? That was the whole point of starting the RfC. Nardog (talk) 22:03, 27 July 2021 (UTC)

You trying to keep us on task or something? I haven't had time to look through the literature but will have more to say this weekend. — Wug·a·po·des​ 19:35, 29 July 2021 (UTC)

I doubt '[t]he International Phonetic Association clearly belongs in the "there are no such things" camp'. In figure 6 on page 35 of the Handbook the symbol ⟨l̥⟩ is used for an unvoiced lateral approximant where ⟨ɬ⟩ would be available for the corresponding fricative, and on page 136 Šuštaršič & Komar mention "a voiceless labial-velar approximant [ʍ]" in Slovene. Also, what is ⟨n̥⟩ used in the exemplification of the diacritic ⟨◌̥⟩ on page 24, if ⟨n⟩ symbolizes a nasal without friction a.k.a. nasal approximant? — On Fricative#Pseudo-fricatives we have a sentence that reads: 'In addition, [ʍ] is usually called a "voiceless labial-velar fricative", but it is actually an approximant.' Love —LiliCharlie (talk) 17:05, 27 July 2021 (UTC)

See Ohala & Solé (2010) quoted above. Just because someone uses ⟨l̥⟩ speaks nothing about whether they regard it an approximant, just as ⟨d̥⟩ is a handy way of signaling a devoiced allophone of /d/. (Also, some might call some allophones "devoiced approximants" but not accept "voiceless approximants" as possible segments, on, for example, phonological grounds.) ⟨n̥⟩ is a completely non-issue because there are no IPA letters for "nasal fricatives". Since a "voiceless approximant" means silence to those who do not posit them (see Wells, O'Connor), attaching the voiceless diacritic to a symbol for a voiced sonorant automatically makes it an obstruent in their view (Ohala & Solé). That sentence was added by Kwami back in 2005, presumably based on Ladefoged, and is unsourced to this day, and the sources I mentioned above clearly dispute it. Nardog (talk) 21:35, 27 July 2021 (UTC)

My point is that the International Phonetic Association doesn't belong in either camp but invites users to disregard even its fundamental assumptions, see below. Love —LiliCharlie (talk) 22:16, 27 July 2021 (UTC)

P.S.: This is not the place to discuss that, but it is quite possible that "silent segments" may still be identifiable by their transition to audible ones. Love —LiliCharlie (talk) 22:43, 27 July 2021 (UTC)

---

The only case where this is a problem is ⟨ʍ⟩, and then only because it has a dedicated IPA symbol, and that a historical relic. Any other voiceless approximants should be covered under the corresponding voiced approximant, but cross-linked to and from the equivalent fricative, with the potential ambiguity being noted in both. That is, whether an author transcribes a sound in the language they describe as [ɥ̊] or [çʷ] or [ɸʲ], or as [l̥] or [ɬ], we should follow them as the RS and not impose an ideological decision that they must be wrong because their transcription contravenes one or another statement by the IPA.

LiliCharlie makes a good point here. Once a convention is established, the IPA needs to get consensus to change it. A lack of consensus to change something does not mean they'd have consensus to establish it today. E.g. ⟨ɱ⟩, which violates the stated conventions of the IPA and I doubt would get consensus if it were proposed as a new symbol today (it would be ⟨m̪⟩), but which has failed to get consensus to retire. Similarly ⟨ɧ⟩, which is used as a phonemic rather than as a phonetic symbol, something the IPA says it doesn't do.

The IPA is not an arbiter of which phonetic distinctions there are in languages. Its job is to provide symbols for them, not to define them. How even members of the IPA use those symbols often differs from the description by the IPA. As LiliCharlie points out, this is true even in the Handbook. There are many more examples in JIPA. The IPA not only tolerates this, but only attempts to apply some order to it. That is, I think LiliCharlie is correct in doubting that the IPA has taken a stand on the issue of whether [ʍ] is a fricative, an approximant, or ambiguous. They simply need to call it something, and there's been no consensus to change what they call it.

As for Scottish (or at least Scots) /ʍ/, among older speakers it's indeed a velar fricative [xʷ]. Even among younger speakers, and in RP English, it's arguably [hʷ]. If [h] is called a fricative, as it is in the IPA, then so would both [xʷ] and [hʷ]. You also have languages like Kham which are analyzed as having [ɥ̊] and [w̥]; phonologically, analyzing these sounds as [çʷ] and [xʷ] runs into the difficulty that there is no /ç/ or /x/ in the language, nor any other labialized consonantal segments, but there are /y/, /ɥ/, /u/ and /w/ and other voiceless sonorants, such as the nasals.

I doubt most of the modern authors who call English /ʍ/ a 'fricative' have actually investigated the question. They may call it that because that's the tradition in the English lit (dating from before there was such a thing as 'approximants'), or because that's what the IPA calls it. With more obscure languages, an author is more likely to make an attempt to analyze the nature of the sound. For instance, with Hupa, Golla lists the fricatives /x xʷ h W/ (elsewhere ⟨ʍ⟩), but says that /W/ is a labialized glottal fricative, forming a labialization pair with /h/. He also says that while e.g. /ha/ is phonetically [ḁ], /W/ is phonetically [u̥]. Thus /ohW#/ is phonetically [o̥u̥]. If we're going to insist that voiceless approximants are 'fricatives', then we'd need to characterize voiceless vowels as fricatives as well. Better IMO just to follow our sources, and to note how differing descriptions may correspond to each other. — kwami (talk) 19:07, 27 July 2021 (UTC)

Sidenote: this has extra complication in that we use IPA as our standard phonetic notation, but authors might not, and could be using a different trancription system that differs on quite a few substantial grounds, e.g. perhaps does not recognize any distinction between fricatives and approximants (/j/ as a "palatal fricative"; such a claim still stands e.g. in Votic language), or describes plain velar consonants like /k/ as "palatal" versus uvular consonants like /q/ as "velar" or "postvelar" (to my understanding this problem appears e.g. in many sources on Pashto, whose "/ç ʝ x ɣ/" really are better considered [x ɣ χ ʁ] or maybe [x̟ ɣ̟ x̠ ɣ̠]). So this already creates often a requirement to "retranscribe" authors' usage to WP's own standard anyway. I don't see any reason to categorically exclude doing the same across different applications of IPA. --Trɔpʏliʊm • blah 16:01, 29 July 2021 (UTC)

kwami wrote: "The IPA is not an arbiter of which phonetic distinctions there are in languages. Its job is to provide symbols for them, not to define them."

That's why part 1 of the Handbook concludes with the following sentence (p. 38): "Nonetheless, the IPA should not be regarded as immutable, even in its fundamental assumptions, and there needs to be a continuing reappraisal of their appropriateness." Love —LiliCharlie (talk) 22:07, 27 July 2021 (UTC)

The discussion so far has focused too much on what the IPA's position is, which I agree is ultimately beside the point. The question I raised is how to discuss sounds described as voiceless approximants in accordance with our policies, particularly WP:NPOV, and I'd appreciate input in those terms. Nardog (talk) 22:19, 27 July 2021 (UTC)

I support kwami's proposal that "voiceless approximants should be covered under the corresponding voiced approximant, but cross-linked to and from the equivalent fricative, with the potential ambiguity being noted in both" and that "we should follow them [=the authors' descriptions and transcriptions] as the RS and not impose an ideological decision that they must be wrong because their transcription contravenes one or another statement by the IPA." And let me add: by the IPA or any other alleged authority. Love —LiliCharlie (talk) 23:00, 27 July 2021 (UTC)

But why in articles about voiced approximants, not voiceless fricatives, which are clearly acoustically/perceptually closer to and never contrast with "voiceless approximants"? I'd find it incongruous and reader-hostile if our article for ⟨ʍ⟩ is at Voiceless labial–velar fricative yet the sounds it typically represents were covered at Voiced labial–velar approximant. Nardog (talk) 23:13, 27 July 2021 (UTC)

Re-read what kwami and I wrote: "voiceless approximants ... should be cross-linked to and from the equivalent fricative" so users find that piece of information no matter which article they happen to visit. That's quite user-friendly, isn't it? Love —LiliCharlie (talk) 23:26, 27 July 2021 (UTC)

But the occurrences will still be listed in the voiced approximant articles, right? Or are you suggesting to list them in both voiced approximant and voiceless fricative articles? Nardog (talk) 23:37, 27 July 2021 (UTC)

My original idea was to have separate tables in separate articles, depending on how the sounds are described by authors, irrespective of transcription symbols they use. I expect many items will require remarks such as "also described as a frictive" etc. What exactly is the advantage of departing from Wikipedia practice and present a unified table in an article that is dedicated to a certain sound but also lists sounds that sources say belong elsewhere? Don't you trust reliable sources? Or is it because you think the expression "voiceless fricative" is overly ambiguous? Love —LiliCharlie (talk) 00:24, 28 July 2021 (UTC)

Separate between what? I wonder if you've mistaken my request for clarification for a suggestion. I ask again: What is your answer to the question at the top of this section? Nardog (talk) 00:43, 28 July 2021 (UTC)

I thought I had answered that: My idea was that voiceless approximants should be covered in articles on voiceless approximants, and voiceless fricatives in articles on voiceless fricatives. This would also apply to lists of occurrences; separate articles, separate lists, even if this requires lots of cross-references. — Should we follow your suggestion ("no option but to list them in articles about fricatives") a unified list should have separate columns for fricatives and approximants, so that either can be filled, or both, as the case may be. Love —LiliCharlie (talk) 01:16, 28 July 2021 (UTC)

IMO the articles on sonorant consonants should be moved so that they no longer have 'voiced' in the title, just as we do with vowels. The bulk of the text would cover the voiced sound, but we'd have a section for the voiceless sound if there's enough coverage in the lit to warrant one. (Probably not for individual vowels, but instead a section on voiceless vowels in the 'vowel' article.) I mean, do we really need the separate articles Voiceless bilabial nasal and Voiceless retroflex flap?

The one exception would be for [ʍ], because that has a dedicated IPA symbol that makes its treatment problematic. I wouldn't object to keeping the article at 'Voiceless labial–velar fricative', as long as we are clear in the heading that that label may be inaccurate. However, rather than saying something like 'arguably, the labial-velar fricative is often neither velar nor a fricative', another possibility would be to move the article to 'ʍ', and to cover the usage of the IPA symbol itself as the topic of the article. The sound [w̥] could then be covered in more detail at Labial–velar approximant (voiceless section), the sound [xʷ] could be covered at labialization, and disputes over the English sound could be covered at Pronunciation of English ⟨wh⟩. All of those articles would be listed in the 'See also' section of the 'ʍ' article.

For a parallel, note how we treat [ɧ] at Sj-sound, rather than under the IPA name 'simultaneous voiceless postalveolar and velar fricative', a description which a former president of the IPA has stated several times is inaccurate. — kwami (talk) 06:45, 28 July 2021 (UTC)

History of ⟨ʍ⟩ through Kiel

⟨ʍ⟩ was adopted specifically for the English wh sound. Usage for other languages has always been ambiguous. AFAICT, it was introduced in the 1900 chart. It was not defined there apart from its placement in the chart (with [ʍ w] as a voicing pair of fricatives under the column labiales and secondarily under vélaires), and in a statement that for both the back of the tongue is raised. ([j] was also a fricative, as was [ɥ], which had 2ary articulation in the palatales column.) The 1904 description says it "is one variety of Northern Engish wh." The 1912 description used the same wording, and the 1912 chart showed it as a fricative in the 'lips' column, with secondary articulation in the 'back' column. [j], [ɥ], [w] were still considered fricatives. In 1921, ⟨ƕ⟩ replaced ⟨ʍ⟩. In 1932 and 1947, [w] and [ɥ] are in a new row for 'Frictionless Continuants and Semi-vowels', under the bi-labial column with 2ary articulation under the velar and palatal columns. [ʍ] is not on the chart and is defined as "voiceless w". In the 1949 Principles, the chart is the same but the description (p. 14) says, "the letter for the voiceless fricative corresponding to w is ʍ, but it is generally preferable to represent this sound by the digraph hw." The 1979 revision places ⟨ʍ⟩ in the chart for the first time since 1904. It is placed in the fricative row, and was no longer presented as the voiceless partner to [w]. It is also labial-velar, in the same column as [k͡p]. That is, it's defined as a doubly articulated [x͡ɸ], which according to Ladefoged and Maddieson is for all practical purposes impossible. The chart presentation reaches it current treatment after the Kiel convention, with ⟨ʍ⟩ listed under 'other symbols' as a "voiceless labial-velar fricative" and ⟨w⟩ as a "voiced labial-velar approximant". However, the Report on the Kiel Convention (p. 70) reported that the convention had decided that the chart should include the wording that "[w] represents a voiced labial-velar approximant and [ʍ] its voiceless counterpart," but that for reasons of space it was not possible to comply with this decision.

— Preceding unsigned comment added by Kwamikagami (talk • contribs) 08:11, 28 July 2021 (UTC)

• I support voiceless fricatives, as proposed by Nardog. I comment more as a Wikipedian than as a linguist, since my own research is not in phonetics or a close sub-field, and I'm not particularly current on that literature. But sources such as textbooks seem to treat them as voiceless fricatives (at least the ones I'm familiar with do), and sources of that nature rather than cutting-edge research should in general guide Wikipedia, in my opinion. Cnilep (talk) 23:22, 27 July 2021 (UTC)

  Prominent authors like Catford and Ladefoged do make the distinction in their textbooks. This is from Catford (1988:66f.) A Practical Introduction to Phonetics: "The articulatory channel for an approximant, however, is a little wider than that of a frictive, just to the extent that airflow through the channel is non-turbulent (hence no hiss-sound) when it is voiced, but it is turbulent and hence noisy (though not strongly so) when it is voiceless." Love —LiliCharlie (talk) 01:40, 28 July 2021 (UTC)

  Yes, the fact that some phoneticians dispute the distinction between voiceless approximants and fricatives (or voiceless vowels and fricatives) does not mean that we should treat voiceless approximants and vowels as fricatives by default.

  Also, the argument is largely about the English wh sound. Without that, the lit would be largely limited to comments like 'a voiceless lateral approximant [l̥], which is perhaps not distinguishable from a fricative [ɬ]' – something we can easily handle with our normal use of RS's. We shouldn't hold the rest of the world's languages hostage to a dispute about English. — kwami (talk) 06:45, 28 July 2021 (UTC)

  This is the second time you talk about voiceless vowels. Leaving the question of air-pressure aside, in Catford's (1977; 1988) terminology, sounds with articulatory channels that result in friction when voiced are fricatives, those with channels resulting in friction only when voiceless are approximants, and those with channels that never result in friction are resonants. Consequently, [ḁ] and [u̥] which you mentioned above aren't both approximants, but have/represent different stricture types. Love —LiliCharlie (talk) 10:15, 28 July 2021 (UTC)

• A related background issue seems to be that it's not very clear if our "sound" articles are phonological or phonetic (the "feature lists" are a particularly clear residue of this question not having been recognized in the early years of linguistic coverage on Wikipedia). Currenly my own preference would be to treat phonemic voiceless approximants (those that contrast with a voiced one or are consistently voiceless) together with the fricatives but phonetic ones as footnotes to the voiced (and also to move these articles from e.g. Voiced alveolar approximant to just Alveolar approximant). It's not like we strive to have separate articles for every attested phoneme, most prominently not distinguishing articles based on secondary articulations or minor allophones, but also not for "minor" phonations like breathy voice. It seems to me that consistent application of this approach (don't make articles on rare phonemes that can be considered variants of a more common one) would call for also not distinguishing voiced and voiceless articles for any sonorants, given that all voicing contrasts on sonorants are cross-linguistically rare. --Trɔpʏliʊm • blah 11:27, 29 July 2021 (UTC)

  I support getting rid of 'voiced' from the titles, but think that treating [ç] and [j̊] in one article, and [j] in another might be confusing, esp. as [n] and [n̥] will still be under the same article. For other phones indicated by diacritics, we treat them in the article for the base letter. Also, if we did it that way, the article on j truly would be only about the voiced approx., so it would be more difficult to argue for it being renamed. — kwami (talk) 19:19, 29 July 2021 (UTC)

I don't think we should lump [ç] and [j̊] together on alleged phonemic grounds, either. They are realizations of different phonemes in, for instance, German ich /ɪç/ [ʔɪç] and Pierre /pjɛːr/ [pj̊ɛːɐ̯] (where [j̊] can be thought of as simultaneous [j] and aspiration). — Not specifying phonation in approximant article titles seems reasonable. Love —LiliCharlie (talk) 17:59, 30 July 2021 (UTC)

I'm afraid I don't understand the 'Pierre' example. Are you talking about an Anglophone pronunciation of the name? If so, I think it is much more often /pi'eə(r)/. When it is pronounced as a single syllable, I don't know how you can tell that the realization of the devoiced /j/ sound is phonetically different from the [ç] sound in German 'ich'. And, sorry, I can't see what "simultaneous [j] and aspiration" means.RoachPeter (talk) 17:25, 31 July 2021 (UTC)

@RoachPeter: No, I was talking about Standard German where this word doesn't seem to have an alternative disyllabic pronunciation (as opposed to words like Pietro, where orthograpic ie may be either /jɛː/ or /i'ɛː/, this variation being noted as /i̯ɛː/ in Mangold's Das Aussprachewörterbuch). — By "simultaneous [j] and aspiration" I meant a VOT after the approximant /j/. Love —LiliCharlie (talk) 11:29, 2 August 2021 (UTC)

P.S.: German /j/ [j̊] has a lot less loud friction than /ç/ [ç]. Some German speakers, especially ones who do not have the /ʃ~ç/ distinction in their native dialect but make an effort to distinguish the two, do pronounce /ç/ as [j̊], but that sounds much too "soft" and conspicuously non-standard to my ears. Love —LiliCharlie (talk) 12:01, 2 August 2021 (UTC)

The problem we seem to have here, besides the poorly defined outlier character ⟨ʍ⟩ in the IPA, is that some phoneticians/phonologists have argues that voiceless approximants do not exist. But others argue that they do exist, though no language seems to make a phonemic distinction between them and fricatives. I don't follow the argument that because one POV says a sound doesn't exist, we can't report on it following RS's that say it does exist. — kwami (talk) 23:41, 30 July 2021 (UTC)

Neither do I, because no one's making that argument. What I also don't follow is how the middle-of-the-road approach of discussing them under fricatives and noting that they are also described as approximants could constitute impos[ing] an ideological decision more than describing them definitively as approximants. Nardog (talk) 03:01, 31 July 2021 (UTC)

I don't understand. How is arguing that voiceless approximants found in the literature aren't really voiceless approximants, because some other sources say such things don't exist, NOT imposing an ideological decision? We should treat voiceless approximants as approximants, and voiceless fricatives as fricatives. When there's debate over the nature of a segment in a particular language, we can cover the debate. Anything else is imposing a minority viewpoint onto sources that don't support it. — kwami (talk) 05:59, 31 July 2021 (UTC)

But what I'm wrestling with is the logical problem of how to define a voiceless approximant that isn't circular. Those who believe voiceless approxx do not exist say that if they produce a noise that is audible, that is non-periodic noise that's typical of fricatives. If voiceless approximants consist of noise, they fulfil the conditions required for a fricative. I have tried to suggest a way of squaring the circle by (in the article on Approximant#voiceless approximants) citing the distinction between laminar and turbulent flow (or cavity and local friction), but this steers close to POV and nobody seems to think it relevant anyway. RoachPeter (talk) 17:25, 31 July 2021 (UTC)

I'm not sure we can. There are lots of things that people argue about how to define, such as stress, fortis/lenis, ATR/RTR. Distinctions that are clear to the ear do not always have consensus explanations. All we can do is to report RS's, and to try to give different approaches appropriate WEIGHT. But if a credible author argues that a particular language has voiceless approximants rather than fricatives, even if that's just a decision made by ear, then we should cover them as voiceless approximants unless we have other sources that investigated and found that they were wrong. Also, the conflict here seems to be between reports from the real world that they do exist, vs theoretical arguments that they can't exist. It's fine to argue explanations, but IMO reality trumps theory.

Thanks for the helpful section on voiceless approximants that you wrote. If you don't mind, I'd like a citation for "the term voiceless approximant is not used or recognized by the IPA", since it is used in the Handbook, and such phrase can always be treated as atomic, as 'labialized alveolar' or 'aspirated fricative' would be. I'd also like to see how voiceless vowels are treated in sources that reject other voiceless frictionless continuants. Per O'Connor's approach, it would seem they cannot exist. Are they all [h] with various secondary articulations? (Plus, some argue that [h] is often not a fricative, e.g. English /h/ vs the fricative /h/ in Arabic. That would mean, per Wells' argument, that English /h/ does not exist. Indeed, this debate reminds me of debates over the nature of [h].) The Handbook does mention voiceless vowels, and it's not clear to me how [i̥] would differ from [j̊] apart from syllabicity judgements, which are another can of worms. — kwami (talk) 20:51, 31 July 2021 (UTC)

@kwami I think your point about voiceless vowels is extremely important. If we agree that voiceless approximants exist (as I am more and more inclined to do), there has to be a clear relationship between voiceless vowels and voiceless approximants as there is between voiced vowels and voiced approximants. The relationship will be seen in their acoustic spectra - the voiceless items will have formants showing in the noise spectrum below about 3kHz at frequencies matching those of voiced approximants, whereas true voiceless fricatives are more likely to have diffuse noise spectra with the energy concentrated above the normal vowel formant region. Regarding the IPA: if it was me that wrote that voiceless approximants were not recognized by the IPA it was careless writing - I should have restricted the point to the official IPA chart/alphabet, where they are not mentioned. I am far away from my books at the moment so can't check on what the Handbook says, but I'm sure it would be worth quoting. I have contacted the President of the IPA to see if we could get some dialogue about this topic - no reply yet. RoachPeter (talk) 18:37, 1 August 2021 (UTC)

Not everything official is in the chart. At least one of the illustrative tone letters in the intro to the Handbook doesn't appear on the chart. The chart is obviously in need of an "etc." or "e.g." there, but that hasn't been fixed despite it being over 30 years. (As an unfortunate consequence, IPA Braille doesn't have full tone support.) And there are additional tone letters that didn't make it into the Handbook at all, yet they're still official IPA per the Kiel Convention. So taking silence to mean lack of recognition is dicey.

(Aside: the editors of the Handbook seem to have taken the other tone letters as redundant, apparently because of Chao's ambiguous wording in his article from the 1930s that was accepted by Kiel, and I suspect omitted them for this reason, but the missing letters encode a distinction that is crucial for unambiguous phonemic transcription of languages like Hokkien. There is even an editor on Wiktionary who vociferously rejects IPA transcription of Sinitic languages because he claims it isn't true IPA.)

For voicelessness, the Chart lists n̥ d̥, and [n] of course is a frictionless continuant. The intro to the Handbook (p.15) says the ring is "available to reverse the voicing value otherwise implied by any symbol. Voiceless trills or nasals, for example, ... can be notated as [r̥], [ŋ̊] etc. ... Vowels which occur without voicing can also be indicated, e.g. [e̥]." No mention of lateral approximants or semivowels (or diphthongs), but "any symbol" and "trills or nasals, for example" implies them. Then on p.24 it says, "the voiceless diacritic can also be used to show that a symbol that usually represents a voiced sound in a particular language on some occasions represents a voiceless sound, as a detailed transcription of conversational English Please say ... [pl̥iz̥ se ...]."

In the illustration for Slovene, [ʍ] as an allophone of /ʋ/ is described as an approximant (p.136). And in the illustration of Portuguese (p.129), it says, "in connected speech, ... consonants and vowels in unstressed positions may be devoiced." The vocalic example is viajante envolto /viaˈʒãti ẽˈvoltu/ [viaꜛˈʒɐ̃tɯ ɯ̥̃ꜛˈvolt], later corrected to [viaꜛˈʒɐ̃tɨ ɨ̥̃ꜛˈvolt]

As for sounds not being possible because there's nothing to hear, the same is true of tenuis [p t k ʔ]: they can only be heard in boundary effects they have on other sounds. Perhaps something similar happens with [l̥ j̊ w̥]? Or perhaps they are accompanied by glottal frication (or however is best to describe [h])? I just don't buy that well documented sounds can't really exist just because they aren't accounted for in someone's theory. — kwami (talk) 21:54, 1 August 2021 (UTC)

Just because some linguists posited voiceless approximants when describing specific languages doesn't make the category of voiceless approximants accepted in general linguistics. Wikipedia doesn't lead; we follow. If the IPA's description of ⟨ʍ⟩ isn't a clear indication of its ascription to a particular camp, it is at least a reflection of the lack of consensus for the category of voiceless approximants. The IPA of course doesn't prescribe any point of view, but it reflects consensus or lack thereof—precisely because it's under "continuing reappraisal"—and thus provides a pretty good yardstick for gauging what the consensus is or whether there is one on a given matter.

Any sound described as a voiceless approximant by some can be, and often is, also described as a voiceless fricative by others, so that opens up to WP:CONTENTFORK.

Separating the sounds described as voiceless approximants from the fricative articles would imply—falsely—that the sounds listed in the fricative articles are definitely non-approximant, which we can't guarantee because not all linguists make the distinction. That's something you should be wary of especially if you're an advocate for making the distinction. Nardog (talk) 18:43, 10 August 2021 (UTC)

But the IPA does posit voiceless approximants, as noted by others above. And not all linguists make a distinction between [h] and voiceless vowels. Should our coverage of voiceless vowels therefore be at voiceless glottal fricative? If I recall correctly, not all linguists accept affricates as a distinct category either. There are a large number of very good phoneticians who describe voiceless approximants, and a large number of RS's for various languages that do as well. We can't impose our POV on them through some legalistic reading of the Handbook that we have no reason to believe was ever intended to be an arbiter of Truth. — kwami (talk) 00:05, 11 August 2021 (UTC)

We don't have separate articles for stop-fricative sequences nor for voiceless vowels (I don't think we discuss voiceless vowels in any of our vowel articles). These issues are not the same as the one raised by the OP. As for that issue, I'd prefer to cover the so-called voiceless approximants (whether they actually exist or not) in the voiceless fricative articles simply to save space and to avoid possible WP:CONTENTFORK (which personally I really dislike). Sol505000 (talk) 08:43, 11 August 2021 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Old split proposal, is there interest?

Norwegian language conflict has an old split proposal from 2017. I am not an expert in the field nor an enthusiast so idk if this would be useful or a good idea, but would like to remove the split tag to clean the backlog. I'm messaging here hoping there will be some discussion on the matter, and if not please ping me back and I'll delete the split tag. Thanks! — Preceding unsigned comment added by A. C. Santacruz (talk • contribs) 12:14, 14 September 2021 (UTC)

External link 'not secure'

Maybe use this one instead?

https://mingle-ish.com/phrasal-verbs/ 213.205.194.52 (talk) 14:28, 10 October 2021 (UTC)

Faux, Defaux, Dufaux, ...

Can some people please take a look at the edits at Faux (surname), Defaux (surname), Talk:Defaux (surname), and Dufaux (surname). I and another editor disagree, so input from others is welcome. Fram (talk) 17:07, 20 October 2021 (UTC)

I agree, though @Fram: pinging me when talking about me had been nice.--Cassius Fury (talk) 17:42, 20 October 2021 (UTC)

While people are of course still welcome to take a look and improve the articles, this seems to have been amicably resolved between @Cassius Fury: and me. Fram (talk) 07:48, 21 October 2021 (UTC)

Yep. The purpose of my terminology (used in the articles) was avoiding siding. Unfortunately, I ended up doing it nonetheless. But @Fram: made me notice my mistakes, and I was able to change the articles to something to which we agree.--Cassius Fury (talk) 10:51, 21 October 2021 (UTC)

Where to redirect cased IPA letters?

Discussion at Wikipedia:Redirects_for_discussion/Log/2021_October_29#Ɥ for where to rd the character itself when we have an article on the letter, such as turned h for 'Ɥ' (the capital of which has no IPA use). — kwami (talk) 22:44, 29 October 2021 (UTC)

Kʰ

Resolved

I'm coming across redlinked phonemes, such as /Kʰ/ in Garhwali language#Allophony which I assume should link to the same target as Kh (IPA). Do we need some new redirects, or are these approximations to IPA which need replacing by some homoglyph? Certes (talk) 14:53, 2 November 2021 (UTC)

The redirect Kh (IPA) should probably go, it was created by a bot that apparently mass-produced redirects with kh for every title (or redirect) with χ (the redirect that prompted that was Χ (IPA)). In IPA, /kʰ/ refers to an aspirated k, which doesn't have its own dedicated article or section, but I guess is subsumed under Voiceless velar plosive. On an unrelated note, the Garhwali article is a bit of a mess. It should one day be rewritten from scratch, so if I were you I wouldn't worry too much about making small improvements to it now. – Uanfala (talk) 01:52, 3 November 2021 (UTC)

@Uanfala: Thanks for the reply. I agree that Garhwali language has bigger issues, so I'll leave it to a subject expert. A sample of similar articles doesn't show any significant problems. Certes (talk) 11:36, 3 November 2021 (UTC)

I listed Kh (IPA) at RFD, as I agree it's a bad target, based on transliteration from non-Latin scripts and nothing to do with the IPA. – filelakeshoe (t / c) 🐱 12:08, 3 November 2021 (UTC)

Discussion at Wikipedia:Articles for deletion/Pashyanti

 You are invited to join the discussion at Wikipedia:Articles for deletion/Pashyanti. Venkat TL (talk) 10:15, 15 November 2021 (UTC)

The discussion has closed as 'speedy delete' for copyright violation. Cnilep (talk) 01:33, 16 November 2021 (UTC)

Example sentences are all copyright violations and should be removed

In a recent discussion with one of our most active admins in the area of copyvio cleanup, it turned out that she believed that language example sentences (whether glossed or not) are subject to copyright. Given that the owner of the copyright is supposed to be the linguist who first published them, and that the vast majority of such language use we have quoted in our articles comes from linguistics works that aren't going into the public domain any time soon, then this means that we may, in principle, expect all such language examples to eventually get expunged from Wikipedia.

Now, I tried discussing the matter, and that went nowhere. Also, I'm too irked to go on. So I'm wondering, don't we have some sort of community decision on this matter? I vaguely recollect there was a similar discussion in the past (Gosh, I'm even convinced it was with the same person), but I'm unable to locate it. Failing that, is there some authoritative online resource that explains the relevant issues? – Uanfala (talk) 23:06, 3 November 2021 (UTC)

Surely the "threshold of originality" thing is a bit of a grey area. I would love to hear how the argument that a gloss of a "word" isn't subject to copyright whilst a gloss of a "sentence" is works when applied to polysynthetic languages. – filelakeshoe (t / c) 🐱 23:19, 3 November 2021 (UTC)

I am sorry that I don't know of any prior discussion to link to, but agree that it would be very helpful to have something to refer to. I think that Uanfala's comments at the user-page discussion are quite right: (1) copyvio software is likely to flag published linguistic data as plagiarism and/or copyright violation, but (2) most linguists or publishers of linguistic work would not consider such reference to data improper nor illegal. (Copying analysis is another thing, but Uanfala makes that clear in the responses.) And (2b) it would be at least potentially improper to try to paraphrase or otherwise change the data when reusing it. Cnilep (talk) 06:48, 4 November 2021 (UTC)

In case the issue arises again, I want to state my strong agreement with Uanfala. When repeating someone else's example you're not quoting them, you're presenting their data. Botterweg14 (talk) 12:51, 4 November 2021 (UTC)

We should spell this out at WP:C or someplace similar, so that it's part of WP policy, noting especially that paraphrasing linguistic data would be unethical. But I don't know where we should put it. Any ideas? — kwami (talk) 01:46, 16 November 2021 (UTC)

The edit that caused the debate was not just a classical interlinear example, but rather a full table that indeed was a slightly trimmed copypaste. I think we should confront Diannaa with the default way of reproducing interlinearized data, i.e. with (author year:page) in-text following the gloss, and ask if they would object to it when this is good practice among all high-quality publishers. In the case of the original edit, I don't think that a more selective excerpt with visible in-text attribution would have triggered the same reaction. But in any case, it will be safest to set it in stone in WP:C, especially the fact that you cannot fully put a correctly formatted interlinear example into quotation marks. Or rather, you can, but it looks dumb. –Austronesier (talk) 07:38, 16 November 2021 (UTC)

I believe the previous instance (which I've utterly failed to locate) concerned more typical interlinear texts. From the comments of the two copyright editors in the two linked discussions so far, it appears that they consider reproducing such single-sentence examples as inappropriate because of copyright. I strongly suspect that the only reason cases like this don't arise more frequently has to do with the technology: the tools that detect copyvios probably aren't good at picking up sources behind paywalls, in printed books, or in formats they can't parse. All these things can change. – Uanfala (talk) 03:21, 17 November 2021 (UTC)

Yeah, plagiarism-detecting software probably isn't going to have much luck with IPA, which probably won't be detected as words. — kwami (talk) 05:25, 17 November 2021 (UTC)

Kwami has raised the issue at Wikipedia talk:Copyrights#citing linguistic examples. – Uanfala (talk) 02:35, 17 November 2021 (UTC)

Here's a somewhat less borderline example [2] (we're lucky that we can actually see the removed text, in most other cases that's not possible because of the practice of revdelling the relevant bits of the history). You can see that many interlinear glossed texts have been removed, as has been a table with a morphological paradigm. The text concerned isn't completely free from any suspicion: there's some close paraphrasing in the analysis (though how many ways are there to say that indefinite demonstratives can fulfil the role of predicates?) and more examples appear to have been used than is strictly speaking necessary. But still, this looks like a fairly ordinary example of quoting linguistic data, and yet it has been removed on copyright grounds. – Uanfala (talk) 23:09, 21 November 2021 (UTC)

I think it's not a good idea to rip a complete chapter from a grammar with all its examples. Paradigms like pronoun sets, inflections etc. have to be complete, but if a source has five examples for the usage of three existing demonstratives, one example for illustration would suffice (the three demonstratives are listed anyway). We cannot expect an editor specialized in detecting copyvios like in Pukapuka language to trim down the text to a non-plagiarized version; recognizing the essentials naturally requires subject matter expertise. So blanket removal of mass-copied text is IMO a legitimate first aid. –Austronesier (talk) 17:40, 23 November 2021 (UTC)

I agree with that. My concern is the assertion that taking more than half a dozens examples from a source violates COPYVIO even if it doesn't violate any copyright law. In a fully developed article on any language, we're likely to exceed that, especially if the language is poorly attested and we only have one or two good sources to choose from. One response at COPYVIO was that such languages should not have high-quality articles on WP! I don't know of any other field where we supposedly are not allowed to cite enough data to bring an article to FA. — kwami (talk) 07:35, 24 November 2021 (UTC)

Your input is requested regarding an ISO 639 redirect

Your input is requested at Wikipedia:Redirects for discussion/Log/2021 November 23#ISO 639:none. Thanks! —Mx. Granger (talk · contribs) 16:59, 25 November 2021 (UTC)

FAR for Irish phonology

I have nominated Irish phonology for a featured article review here. Please join the discussion on whether this article meets featured article criteria. Articles are typically reviewed for two weeks. If substantial concerns are not addressed during the review period, the article will be moved to the Featured Article Removal Candidates list for a further period, where editors may declare "Keep" or "Delist" the article's featured status. The instructions for the review process are here. Z1720 (talk) 22:33, 27 November 2021 (UTC)

Velar-bilabial affricate

For those who're interested in exotic affricates (or at least have better competence in phonetics than I do), the question at Talk:Affricate#The velar-bilabial affricate kɸ might be of interest. – Uanfala (talk) 20:37, 2 December 2021 (UTC)

Yan (surname 顏)

Hello, I was looking for the Etymology subgroup but the redirect brought me here. Right now, we have an editor who, over the past few years, has added dozens and dozens of names to this page, none of whom are supported by Wikipedia articles. It is a very common name in China so the list of historical figures with the name Yan could possibly be in the hundreds. Is there a policy on who is included in a surname list? I have reverted their edits but they are very focused on this page so I expect to be reverted myself. Thank you for any information (or a second pair of eyes) you can offer. Liz ^{Read! Talk!} 20:15, 15 December 2021 (UTC)

I'm only speaking for myself, but my criterion would be, Are they notable enough to warrant an article, assuming anyone ever got around to creating one? If they don't even have an article on WP-zh, then maybe not.

I've had the experience of going to a dab page, and not being able to find what I'm looking for. If the name is there and is a red link, then I at least know I've hit a dead end with WP-en. But sometimes the red link is enough, because we need to add a little info to dab who is who. If I find a red link for "Yan Xi, 8th-century dentist", that might be enough for me to understand the reference I'm looking up, or to solve my crossword puzzle. Or perhaps all I need are the hanzi for their given name, and there they are! If that entry were deleted just because it didn't have a WP article, then I'd be left drifting.

On the other hand, we wouldn't want 2,000 non-notable red-link entries either. As with anything, we're looking to maximize the utility of WP, and we're not always going to agree on where to draw the line. — kwami (talk) 07:34, 16 December 2021 (UTC)

IPA links for post-stopped nasals

I have noticed that {{IPA link}} redirects the post-stopped nasals mᵇ etc. to pre-stopped consonant because the Module:IPA_symbol/data currently does so in the presence of ⟨ᵇ⟩ etc. Maybe this should be fixed. –Austronesier (talk) 20:34, 21 December 2021 (UTC)

Where should it link instead? Prenasalized consonant? Nardog (talk) 13:49, 22 December 2021 (UTC)

Well, there's the redirect Post-stopped nasal → Prenasalized consonant that brings the reader at least a step closer to an explanation of the phenomenon (see third lede §). I'm not sure if there is sufficient potential in the existing literature for a full article about post-stopped nasals in the near future, but in any case the redirect looks like a good target to me. –Austronesier (talk) 17:53, 22 December 2021 (UTC)

Done. Do we need voiceless ones (⟨mᵖ⟩ etc.) too? Nardog (talk) 08:44, 26 December 2021 (UTC)

Great, thanks! For the languages that I am aware of having (or reported to have) post-stopped nasals (a handful of languages on Sumatra, and some Chinese varieties), only the voiced variants occur, so for the time being, no. –Austronesier (talk) 20:20, 26 December 2021 (UTC)

Vowel templates at TfD

FYI: {{Near-front vowel}} and {{Near-back vowel}} have been nominated for deletion. Nardog (talk) 10:21, 10 February 2022 (UTC)

FAR for Truthiness

I have nominated Truthiness for a featured article review here. Please join the discussion on whether this article meets featured article criteria. Articles are typically reviewed for two weeks. If substantial concerns are not addressed during the review period, the article will be moved to the Featured Article Removal Candidates list for a further period, where editors may declare "Keep" or "Delist" the article's featured status. The instructions for the review process are here. – Kavyansh.Singh (talk) 08:52, 23 February 2022 (UTC)

Scope of project

I am wondering if {{Anthony}}, which is a template of names derived from the Latin Antonius and its Anglicized variant Anthony, is within the scope of this project. I have worked on a few name origin tree templates like this and when I was checking the tags on Antaine, I noticed this project. Should this project be tagged on this type of template.--TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 23:55, 10 March 2022 (UTC)

There's the subproject Wikipedia:WikiProject Linguistics/Etymology which deals with etymologies, but it's so broadly defined that potentially millions of articles could fall under its scope. I don't think the linguistics projects is interested in articles about personal names (any more than it's interested in articles about places). Maybe we should remove the project tag from the half a dozen or so given name articles currently tracked [3]? – Uanfala (talk) 00:05, 11 March 2022 (UTC)

Feel free to remove it in the interest of your project where it is at issue. I was just checking in. Thx for the prompt reply.--TonyTheTiger (T / C / WP:FOUR / WP:CHICAGO / WP:WAWARD) 00:11, 11 March 2022 (UTC)

What is the difference between Performative text and Performative utterance?

I couldn't discern a significant difference, but I don't know a lot about the subject matter.

I wondered if Performative text and Performative utterance should be merged. But I defer to those with expertise. :0) Mark D Worthen PsyD (talk) [he/him] 02:23, 31 March 2022 (UTC)

It looks to me like the topics could be merged. Most of the sources on Performative text (Austin, Searle, Derrida, Foucault, Butler) are studies of the performative as such, which is covered in Performative utterance and Performativity. I don't know the work of Schechner or of Skinner, but Schechner likewise seems to be talking about something like performative utterances. Skinner seems to be the only scholar cited who treats text as such, but even in that case there is reference to Austin. Perhaps others know this work better, but to me it looks like the concepts are closely related and the articles are relatively short (~10 kB each), so that they could be merged. Cnilep (talk) 03:22, 31 March 2022 (UTC)

An academic you might be interested in - Allen Walker Read

Hello WikiProject Linguistics! If you're ever in the mood to write about an engrossing academic personality, check out Allen Walker Read. He was an American linguist and etymologist who, among other things, studied bathroom graffiti and the origin of "ok." He was really more of an etymologist than a straight-up linguist, but contributed a ton to the study of American and British English.

I've been intermittently expanding the article for a couple months and have assembled a bunch of high-quality sources on the talk page here. There's tons of material there that could be used to expand the article. Happy editing! Ganesha811 (talk) 19:13, 5 April 2022 (UTC)

Borrowed words in English

I am looking for opinions on how best to view foreign language words used in English. At what point does a foreign word become an English word of foreign origin? The relevance of this distinction relates to how that word should be treated when used in English - should the word follow English norms of usage, spelling, grammar, pronunciation etc, or should it follow the norms of the lending language. For example, is troika an English word or Russian? Or the phrase 'à la carte', English or French, and should it be spelt with or without the accent? Some might say that if a word is used in English and especially is in an English dictionary, it has become an English word that has begun undergoing assimilation, and that assimilation will vary and be changing depending on many factors. The words troika and eureka would presumably have been written in the Roman script from the start, beginning the assimilation process immediately. This question has come from an extensive debate around the use of Maori words in English. There has been a trend in the last few years to use Maori words more regularly in written and spoken English, such as tīpuna. Some of these words would rarely have been used before, if at all. This therefore goes beyond usage of the well assimilated Maori words that have been used for decades in English, particularly in NZ English, words such as pakeha and kumera. This change is part of the Maori renaissance begun roughly fifty years ago and specifically after legislative and policy changes in the last decade or so. Should the word pakeha, for example, that has been used in NZ English for a couple of hundred years, have its spelling 'corrected' to pākehā. Is it an assimilated English word that does not need to be altered or is it a Maori word that must follow Maori language rules even when used in English, such as its spelling? This NZ example is complicated because nearly all the sources available are subject to legislative changes and official policy directives that require the new Maori spellings to be used and, on national TV and radio, the 'correct' pronunciation. This compromises the independence and hence the reliability of otherwise reliable secondary sources. The implications of using the 'correct' form of a word that has been borrowed has wider implications than just Maori words. For example, is anyone pronouncing the '-t' at the end of 'trait' open to correction? Roger 8 Roger (talk) 21:33, 11 April 2022 (UTC)

Request for comment

		Request for comment on removal of prefix "Islamic" from "Islamic death penalty"
		Contested and attempted removal of the prefix "Islamic" from "Islamic death penalty", which is construct used as a pipe for the wikilink Capital punishment in Islam and as phrase remains unreferenced. Please participate in the discussion at Talk:Page Thanks.--౪ Santa ౪99° 22:04, 17 February 2022 (UTC)

I am pretty sure this is completely unrelated. JanKeso (talk) 00:53, 15 April 2022 (UTC)

Maledictology and Scatolinguistics

Is there a reason to have both Maledictology and Scatolinguistics? They seem to cover the same material...although the articles as written suggest the former is under the purview of psychology? I'm also not sure how notable either of these are qua subdisciplines of linguistics..., like there's certainly research about profanity, but are there enough sources about this as its own field to warrant a Wikipedia article? Umimmak (talk) 07:49, 18 April 2022 (UTC)

Also Maledicta, Reinhold Aman... AnonMoos (talk) 17:00, 18 April 2022 (UTC)

@AnonMoos: I'm not sure I get what you mean when you say Also? Both Maledicta and Reinhold Aman seem notable for Wikipedia? Umimmak (talk) 20:03, 18 April 2022 (UTC)

They're further Wikipedia articles in the same area. I agree that Reinhold Aman is certainly notable. AnonMoos (talk) 21:08, 18 April 2022 (UTC)

Standardise voiced/voiceless over fortis/lenis

Fortis and lenis are both terms that have inconsistent definitions and are rarely used in comparison to voiced/voiceless in every context, it seems, other than Wikipedia pages for phonology; this is needlessly confusing and voiced/voiceless should be enforced over them. There are two different reasons one may argue against this proposition. Firstly, one might say that fortis and lenis mean voiced and voiceless themselves. In this case, fortis and lenis, as the rarer terms, should not be used for the sake of recognition. Otherwise, one may propose an alternative definition for fortis and lenis; it is the strength of a consonant, or the length, possibly something even more exotic. These simultaneous arguments, when viewed together, show why this fails; there is no consistent definition of fortis and lenis! In fact, there exist consonant inventory tables which use fortis and lenis in place of voiced and voiceless, despite the fact that plenty of pages for phonology, including Portuguese phonology and Spanish phonology, are entirely comprehensible without mentioning either word once, even in their many sources. In conclusion, fortis and lenis are dated and ambiguous terms that should be replaced with voiced and voiceless. — Preceding unsigned comment added by 208.185.122.45 (talk) 22:36, 21 April 2022 (UTC)

Fortis and lenis are not synonymous with voiceless and voiced. It is true that they don't have agreed-upon phonetic correlates, but it is also true that there are phonological units, like English /b, d, ɡ/, that contrast with their voiceless counterparts but are themselves also often voiceless. Calling them voiced phonemes would be misleading if not inaccurate. This is why the terms are used in the first place and they remain useful as phonological concepts. We use whatever terms reliable sources use, and not doing so is original research. Nardog (talk) 05:27, 23 April 2022 (UTC)

Regular sound correspondences between Hungarian and other Uralic languages‎

The page Regular sound correspondences between Hungarian and other Uralic languages‎ strikes me as a problem. Leaving aside the weirdly long name, the page is unreferenced and the topic seems unencyclopedic. We generally don't have articles on cognate sets, and nothing about these cognates strikes me as being the subject of major coverage. I considered taking it to AfD, but I'm interested in whether other editors think there's value in keeping this around and how we might be able to improve it. — Wug·a·po·des​ 03:35, 27 April 2022 (UTC)

We have a handful of articles of the type "Phonological history of...", so it might be worth considering to remodel the article along these lines (with sources, of course), as a subpage of History of the Hungarian language. –Austronesier (talk) 14:06, 27 April 2022 (UTC)

User script to detect unreliable sources

Main page: User:Headbomb/unreliable

I have (with the help of others) made a small user script to detect and highlight various links to unreliable sources and predatory journals. Some of you may already be familiar with it, given it is currently the 39th most imported script on Wikipedia. The idea is that it takes something like

• John Smith "Article of things" Deprecated.com. Accessed 2020-02-14. (John Smith "[https://www.deprecated.com/article Article of things]" ''Deprecated.com''. Accessed 2020-02-14.)

and turns it into something like

• John Smith "Article of things" Deprecated.com. Accessed 2020-02-14.

It will work on a variety of links, including those from {{cite web}}, {{cite journal}} and {{doi}}.

The script is mostly based on WP:RSPSOURCES, WP:NPPSG and WP:CITEWATCH and a good dose of common sense. I'm always expanding coverage and tweaking the script's logic, so general feedback and suggestions to expand coverage to other unreliable sources are always welcomed.

Do note that this is not a script to be mindlessly used, and several caveats apply. Details and instructions are available at User:Headbomb/unreliable. Questions, comments and requests can be made at User talk:Headbomb/unreliable.

- Headbomb {t · c · p · b}

This is a one time notice and can't be unsubscribed from. Delivered by: MediaWiki message delivery (talk) 16:01, 29 April 2022 (UTC)