Wikipedia talk:Manual of Style/Pronunciation/Archive 10

Archive 5

←

Archive 8

Archive 9

Archive 10

Archive 11

With respect to placement

Pronunciation should not be put in the first sentence of the article IMO. We are not a dictionary. A number of projects have been moving it to the infobox such as iron and metoclopramide were it fits well and IMO is sufficient for the lead. Should we change this MOS to reflect this practice? Doc James (talk · contribs · email) 03:59, 5 September 2015 (UTC)

I think it depends on the word. If the pronunciation is unexpected, then it should be indicated in or close to the first sentence. How you pronounce a term is a highly important fact, just as much as its spelling (people communicate primarily by speaking). Readers will often make their own natural assumptions about how a word is pronounced rather than look it up or seek it in infoboxes or wherever, so if those assumptions are likely to be wrong, then we do them a great service by drawing their attention to that fact. If explaining the pronunciation is a complex matter then it can be done in a footnote or somewhere else, but there should at least be a link to it in the first sentence IMO. W. P. Uzer (talk) 05:34, 5 September 2015 (UTC)

Pronunciation of English words of foreign origin

Hallo, there is actually an issue about the pronunciation of English words of foreign origin: often at the beginning of the article's lead there are the English and then the native pronunciation. A fellow wikipedian is removing the latter, leaving in place only the English one. I would like to know if the Manual of Style has a guideline about that. I found here a guideline about foreign names which support both pronunciations, but maybe this is limited to the proper names. Thanks, Alex2006 (talk) 18:05, 15 October 2015 (UTC)

I think it depends on how "naturalized" the word is. Pretty much all English words are "of foreign origin" if you go back far enough. I remember a student from Italy who tried to convince me that the English pronunciation of "plus", /plʌs/, was wrong, and we should be saying /plu:s/, because it's a Latin word. Obviously we're not going to do that. However, for words that are sort of obviously "still foreign", to the extent that you'd be tempted to write them in italics, then sure, I favor keeping the original pronunciation as well. --Trovatore (talk) 18:37, 15 October 2015 (UTC)

I think foreign pronunciations should only be detailed for foreign words that are used only in specific cultural contexts, such as yakuza, jihad, hajj. By contrast, words like "pizza" and "cliché" have become naturalized and routine. "Pizza" is now an English word that just so happens to be spelled and more-or-less pronounced the same as the Italian word that it copied.Kurzon (talk) 18:54, 15 October 2015 (UTC)

Maybe there are some examples we can agree on. I agree with you on "pizza" (I don't think it should have any pronunciation guide, even an English one). On the other hand, I'm pretty sure we should have the Italian pronunciation for zabaglione. I'd call gelato an intermediate case — I'm not sure the Italian pronunciation is really necessary, but it's not implausible either. --Trovatore (talk) 19:00, 15 October 2015 (UTC)

If there's ambiguity about whether the foreign/native pronunciation is relevant, keep both. Obvious cases where the non-English pronunciation isn't relevant means it should be taken from the lede, but putting it elsewhere in the article is always a viable option. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 20:28, 15 October 2015 (UTC)

What do you mean by relevant? Back to the food examples, they describe Italian foods which became popular also in the English-speaking countries. Is it relevant for the description of these food to add the way how the Italians spell it, or not? Sorry for my insistence, but I am trying to find a kind of general guideline. Thanks, Alex2006 (talk) 05:13, 16 October 2015 (UTC)

I mean it in the broadest sense. I hesitate to be more concrete than that because I don't think it's my place to offer a hard line on the matter. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 06:23, 16 October 2015 (UTC)

I don't think it's really possible for the MoS to offer a rigid criterion here. Maybe it could say something about the general principles, but the details are going to have to be worked out article-by-article according to normal consensus-building procedures. --Trovatore (talk) 07:00, 16 October 2015 (UTC)

A week ago I got into an edit war with an unregistered editor who wanted to insert IPAs for "mafia" for British English, American English, and Italian into the lede of the Mafia article. I thought it was absurd to have IPAs at all, let alone three. Accents aside, "mafia" is pronounced more or less the same in English as it is in Italian.Kurzon (talk) 07:44, 16 October 2015 (UTC)

This is an issue that has to be confronted many times by editors of dictionaries of English pronunciation. If you add an “authentic” pronunciation to the conventional anglicized one, you need to be clear whether this is recommended for use by native speakers of English as an alternative. For example, for ‘Bolognese’ a standard dictionary pronunciation (English) would be /ˌbɒləˈneɪz/, and an alternative pronunciation (slightly closer to the original Italian) would be /ˌbɒləˈnjeɪz/. An attempt at a more Italian-like pronunciation might be /ˌbɒləˈnjeɪzeɪ/, though you wouldn't recommend an English speaker to pronounce the word in this way in 'Spaghetti B.'. However, this is still not the same as an authentic Italian pronunciation; what is given by WP at Bolognaise sauce is / bboloɲˈɲeːse/ (I must admit to having doubts about this transcription, especially the double /bb/ and the use of /s/ rather than /z/, but I’m not an expert on Italian). In practice WP normally works this out fine by giving the usual English pronunciation and then where relevant and useful giving the “authentic” pronunciation indicating that this is the pronunciation in the (named) original language. I can’t see why anyone should object to this. RoachPeter (talk) 14:54, 16 October 2015 (UTC)

/b/ in ragù bolognese (but not in ragù alla bolognese) is geminated because ragù is accented on the last syllable - see syntactic gemination. Peter238 (talk) 18:01, 16 October 2015 (UTC)

Really? I would definitely recommend to any English speaker that bolognese should be pronounced with four syllables, even when speaking English. Hearing it with three syllables is really grating. --Trovatore (talk) 19:52, 16 October 2015 (UTC)

It's not up to you, though. Nearly everyone says it with three syllables (at least, where I come from), and I suspect that's the first or only pronunciation given in modern dictionaries. I certainly think the original pronunciation and the (perhaps multiple) standard English pronunciation(s) ought to be given for any word like this which is an obvious borrowing; this information doesn't necessarily belong in the first sentence though. W. P. Uzer (talk) 07:07, 17 October 2015 (UTC)

Just a parenthesis about ragù bolognese: /bboloɲˈɲeːse/ is NOT the pronunciation that should be indicated, which should be /boloɲˈɲeːze/ if this Wiki has got to report the actual pronunciation og the words, because /bboloɲˈɲeːse/ is only written in old orthoepic manuals and pronounced by southern-Italian people, even in Tuscany (where the language spoken should be the most similar to standard Italian) young people tend to abandon this old way to pronounce words, even Italian dubbers (which have studied Italian diction) always say /boloɲˈɲeːze/. All I have just said was to say this: /bboloɲˈɲeːse/ is felt like a dialectal variant, new manuals of pronunciations indicate it as 2nd choice and new generations prefere the 1st utterance, also in TV is heard only this one while the other one is heard only when a person with a strong regional accent is speaking, certainly before the end of this century all dictionaries will report /boloɲˈɲeːze/ as correct and most used form. But there are still some people who insist that this is the only correct pronunciation and provide as source guides which consider a language like something dead and static instead of a living and permanently changing think; they are not the majority, just more noisy and extremist. Alas, at this moment they rule... — Preceding unsigned comment added by 5.101.99.101 (talk) 09:46, 2 November 2015 (UTC)

Need a guideline about pronunciation of foreign words

Good morning, since some time there are repeated edit wars about pronunciation of Italian words. First, I would like to know if here there is a guideline about this subject. Second, I ask if it is correct to remove these pronunciations if unsourced: my opinion is that basically is WP:OR to insert a pronunciation without a reliable source. Thanks, Alex2006 (talk) 07:35, 6 November 2015 (UTC)

Help:IPA for Italian offers guidelines on the pronunciation of Italian specifically. If there is dispute about a particular word, you would need a citation in the article(s) in question. If there is a dispute about Italian pronunciation in general or how best to transcribe Italian, the issue should be discussed at Help talk:IPA for Italian. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 07:45, 6 November 2015 (UTC)

Thanks! Alex2006 (talk) 07:56, 6 November 2015 (UTC)

Time to archive?

This page has gotten so long that it is unwieldy to scroll down it and find information. I've archived my own talk page several times, but never a talk page in another space. I'd like to suggest it be done to this one soon, if someone feels confident enough and has the authority to do so. Thanks! KDS4444^Talk 19:56, 13 November 2015 (UTC)

 Done — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 00:19, 14 November 2015 (UTC)

IPA for words whose pronunciation is commonly understood

I am following a thread I noticed in Archive 9 under the heading "Propose: Where multiple pronunciation of a name exist..." and the subheading "Threaded discussion" and the subsubheading, "No need of IPA for ordinary English". I had asked the question, does anyone know if this has this ever been formalized into a policy or Manual-of-style guideline somewhere? If not, it seems like it should be and I would be glad to put together a proposal; if so, could someone please point me to where? Thanks! but the question got archived before I could get any answers so am reposting it here with crossed fingers. KDS4444^Talk 04:01, 14 November 2015 (UTC)

I understand that "this" means not including pronunciations of arbitrary words? I doubt that there's anything "formal" written about it except what appears in this guideline, but others may know better. W. P. Uzer (talk) 09:07, 14 November 2015 (UTC)

Use of ɵ for broad IPA transcriptions

There are no sources for the use of ⟨ɵ⟩ in broad IPA transcriptions outside of Wikipedia. See Wikipedia:No original research/Noticeboard#Is it OK for pronunciation symbols to be Original Research? — Refactored to talk page on behalf of J. 'mach' wust 18:18, 9 November 2015 (UTC), by me.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  12:27, 12 November 2015 (UTC), from incorrect use of mainspace {{citation needed}} template to post this into the guideline itself.

Clearly the ruling clique doesn't like it to be pointed out that what this guideline recommends is blatantly contrary to Wikipedia policy, as it results in the inclusion in articles of assertions that cannot possibly be supported by reliable sources. W. P. Uzer (talk) 12:42, 12 November 2015 (UTC)

Please see WP:ARBATC#All parties reminded; implying that people who disagree with you are part of conspiratorial cabal is clearly within the scope of that general admonition.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  13:39, 12 November 2015 (UTC)

This MOS page recommends an original research symbol to be used in the mainspace. We should not recommend any original research to be used in the mainspace. Arguing that we can recommend original research to be used in the mainspace because the MOS is blind to original research thwarts WP:OR and is a case of wikilawyering.

I suggest we change the wording of this MOS page so it no longer recommends any potential original research symbols, e.g. by changing the sample word “pronunciation /prɵˌnʌnsiˈeɪʃən/” to “Wikipedia /ˌwɪkiˈpiːdiə/”. --mach 🙈🙉🙊 12:44, 12 November 2015 (UTC)

I'm not sure that's a great example either, unless we have that transcription from a reliable source. Would be better to use a well-established word for which there are plenty of sources and no ambiguity as to which IPA symbols to use. (I've chosen England for now.) W. P. Uzer (talk) 12:53, 12 November 2015 (UTC)

Transcriptions in our own (two) WP-specific pronunciation indication systems, like the formatting used in our internal citation styles, are WP-original metadata, and are not subject to WP:NOR as to their orthography (even if one of those transcription systems is a modified form of IPA; the other one is a totally synthetic one based loosely on [mostly American] dictionary-style pronunciations [pro-NUM-see-ay-shunz]). It's physically impossible for any external source to be cited for a transcription that is in WP's own internal transcription systems, unless and until some off-WP sources adopts our transcription systems. Let's ground this back in reality.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  13:39, 12 November 2015 (UTC)

Well yes, this is exactly the problem. There is nothing to exempt pronunciations (or any other pieces of information that we provide in articles) from the requirements of WP:NOR, and nor should there be. We should absolutely not be using a system that causes articles to contain statements that are not verifiable using reliable sources. W. P. Uzer (talk) 14:40, 12 November 2015 (UTC)

Exactly. The only reason why “[i]t's physically impossible for any external source to be cited for a transcription that is in WP's own internal transcription systems” is because the current WP transcription system is original research. If we loose the original research, then we can very easily cite external sources. --mach 🙈🙉🙊 15:09, 12 November 2015 (UTC)

There are already two open discussions about this, the first already mentioned, the second at Help talk:IPA for English#The Diaphonemic System. No issue's discussion should be forked onto three pages at once. It will probably close without resolution at the NOR noticeboard, so it should be resolved at the IPA for English talk page, where it's already been under multi-thread discussion for some time. No changes should be made to this stable guideline until an actual consensus emerges from the already-ongoing discussions about about approach to IPA. (FWIW, I sympathize with the view that IPA should be used correctly, not in modified form, but arguing this case on as many pages as possible until one gets the answer one wants is the wrong way to go about resolving the issue.)  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  13:39, 12 November 2015 (UTC)

Let me agree with both points made above. There is a sound case to be argued that it wasn't helpful to readers who know the IPA (other readers aren't likely to be helped in any way by its use) for the English Wikipedia to make up its own uses for some of the symbols. It certainly confused/confuses me (I used to teach British university students a version of the IPA). But this has nothing to do with WP:OR; any publication is perfectly entitled to use its own unique internal symbols, so long as these are explained. Please focus on the issue: is there a better system which will work for a international encyclopedia allowing a range of ENGVARs? Peter coxhead (talk) 17:26, 12 November 2015 (UTC)

So you think it would be perfectly OK under WP:OR to say that black is white, as long as white were linked to a page that "explains" that on Wikipedia white means something else than it does in the real world? Of course not, yet this is exactly what's being done here. If there was any reason to think that all readers would reach the explanation, it wouldn't bother me so much. But just as readers who see "white" won't bother clicking the link if they think they already know more or less what "white" means, so readers who see IPA symbols will not click the link if they think they already know what those symbols mean (or conversely, if they know the pronunciation, they will assume the symbols they see are the correct standard IPA for representing that pronunciation - thus mislearning IPA instead). Of course there's a better (though not perfect, obviously) system - give pronunciations in the standard dictionary way, not attempt to synthesize except in cases where the sources synthesize, and avoid any perceived excessive clutter by the use of footnotes or links to other parts of the article. This really isn't difficult. W. P. Uzer (talk) 17:55, 12 November 2015 (UTC)

@W. P. Uzer: we're going over the same ground here, but here's one last response. Your example is way off-beam. The IPA is not used in exactly the same way in all sources (I've just checked three on my shelf). There is no "correct standard IPA" – there are phonetic and phonemic uses, there are broad and narrow transcriptions, etc. There is no "standard dictionary way" – there are notations used in different dictionaries, most notably in different countries trying to represent different dialects of English.

Notations aren't "right" or "wrong", but they can be better or worse. A small part of the scheme used in the English Wikipedia is somewhat odd to people used to the IPA (which doesn't make it "wrong"). So the question that needs to be discussed is whether there a better way to represent the broad phonemic transcription needed in an encyclopedia that allows the use of different variants of English in different articles. I don't know the answer to this question, and I wish we could discuss it instead of wasting time on irrelevant and straw men arguments about OR, sourcing, or whether "white" can mean black. Peter coxhead (talk) 11:13, 13 November 2015 (UTC)

These arguments are not at all irrelevant; they go to the crux of the matter. Of course not all sources use IPA in exactly the same way, just as not all sources use the word "white" in exactly the same way, but that doesn't give Wikipedia the right to just assign these terms or symbols whatever meaning might come into some editors' heads. We need to work carefully within the range of meanings that sources ascribe; to make up meanings of our own is plainly ridiculous - just as ridiculous, and for exactly the same reason, in the case of IPA symbols as in the case of words like black or white. W. P. Uzer (talk) 11:44, 13 November 2015 (UTC)

I have restored the earlier version of the MoS (keeping the benign removal of the redundant link to IPA for English). It is clear that users object to inline citation requests in the MoS; their purpose is to indicate that material is challenged and this challenge has been thoroughly established in three places now. Further attempts to restore these citation requests would only have value in the ways that disruption elicits discussion.

Changing the example to an "uncontroversial" one is problematic because, as long as the diaphonemes are part of our transcription system, an example word with diaphonemic symbols is not only appropriate, but more thoroughly illustrative. As long as ⟨ɵ⟩ is part of the system, it should be perfectly fine to provide a transcription here that includes it. If the symbol ⟨ɵ⟩ is to be removed, the discussion should occur at Help:IPA for English

I also reverted changes to the introductory paragraph, as they inherently challenge some agreed-upon conventions and goals that the editor who made the changes objects to. Namely that

• Transcriptions should not depend on the reader's accent
• IPA transcriptions are strongly preferred over other systems

Again, before these changes to the MoS can be made, an agreement on changing these goals should be made at Help:IPA for English. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 17:50, 13 November 2015 (UTC)

That IPA transcriptions are strongly preferred is surely something that was reinforced by the changed text. And that "transcriptions should not depend on the reader's accent" is a weird thing to say - readers don't make the transcriptions - but what the guideline says (that the interpretation should not depend on the accent) is at total variance with the system that appears to be in place (a system designed - however flawedly - to make the interpretation depend on the accent), so I don't understand why you (as an apparent supporter of the system) would want to make the guideline state the opposite. And to deliberately choose a controversial example when we could choose any of hundreds of uncontroversial ones seems to go quite against the spirit of consensus. W. P. Uzer (talk) 07:42, 14 November 2015 (UTC)

I'll say it again. As long as ⟨ɵ⟩ is part of the transcription system, it should be perfectly fine to provide a transcription here that includes it. If it is controversial, the remedy is to discuss the issue at Help:IPA for English and come to an agreement that mitigates this, not hide it from the MoS like we're ashamed of it.

The changes I was referring to include the replacing of this:

"broad transcriptions of the International Phonetic Alphabet (IPA) should be used"

with this:

"Pronunciation in Wikipedia is usually transcribed using the International Phonetic Alphabet."

I don't think a reasonable reader would see this change as strengthening the preference for IPA over other systems. One is prescriptive and the other is descriptive.

Similarly, the removal of "should be transcribed in such a way that its interpretation does not depend on the reader's accent" without replacing it with something clearer seems designed as a way of removing mention of the primary motivation for adopting a diaphonemic system (dialect neutrality). There could be a clearer way of wording this, but removing it outright certainly isn't the way to go. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 08:07, 14 November 2015 (UTC)

Well yes, there is a clearer way of saying something than saying the exact opposite of it. And "usually" seems to be correct, since there are examples in the guideline of instances where IPA is not considered to be the best method. W. P. Uzer (talk) 08:18, 14 November 2015 (UTC)

Perhaps "should generally be used" is what we're looking for, then. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 08:24, 14 November 2015 (UTC)

Maybe; I just tried again (before I saw this suggestion), tinker with it if you like. Another reason for not using "pronunciation" as the example is that this is not a word for which we give a transcription in its article, and is probably not one that we would normally want to give one for. W. P. Uzer (talk) 08:41, 14 November 2015 (UTC)

I like the changes you made. I think that gets at both of our concerns. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 18:07, 14 November 2015 (UTC)

I understand that using an example with a diaphonemic symbol can be useful. I would suggest, though, we switch to a sample with a less controversial diaphonemic symbol. --mach 🙈🙉🙊 08:49, 14 November 2015 (UTC)

Which one would not be controversial? Maybe /i/ for the happy vowel (if that's intended to be diaphonemic). W. P. Uzer (talk) 09:03, 14 November 2015 (UTC)

Yes, that one, or the diaphoneme that can either correspond to /ə/ or to /ɪ/ as in the words “wanted, luggage, buses” (CGEL. “Preliminaries”, p. 15s.). --mach 🙈🙉🙊 09:14, 14 November 2015 (UTC)

Has that symbol been adopted into the Wikipedia system now? I thought they were using a different (and thus controversial) barred i symbol for that. W. P. Uzer (talk) 09:27, 14 November 2015 (UTC)

I don't like the reasoning provided thus far behind changing the example. If we are to change it, we should have a good reason to do it. It seems more like an effort to further the opposition to the symbol used, which is still part of the system. Because this is the primary motivation and because it's part of a current discussion, let's hold off on changing it until that discussion is concluded. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 18:07, 14 November 2015 (UTC)

Leaving aside the issue about using diaphonemic transcription, I think it should be said that the word 'pronunciation' is a poor example to give for a case where a vowel may be either [ə] or [oʊ]. To use [oʊ] in the first syllable of this word would sound comically archaic. The example given at Help:IPA for English is 'omission' (transcribed with //ɵ// in the first syllable) which seems fine to me. RoachPeter (talk) 10:50, 15 November 2015 (UTC)

Primarily, I think the example needs to be one from an article where we actually do give the pronunciation (and where there would be no doubt that it is one where we ought to give the pronunciation). There seems no need to choose one which uses this or any other particular symbol; indeed, it would be better to give an entirely uncontroversial one. Given that we now have at least three essentially different objections to the current example, I'm going to change it again. I'll again use the apparently uncontroversial "England"; if someone has any better ideas, please substitute. W. P. Uzer (talk) 07:30, 16 November 2015 (UTC)

Qualifications

I suspect that most of the pronunciations being added to articles are OR, i.e. what people know or think they know, and to be realistic I doubt we can escape that (except maybe make a ref an absolute requirement, but that's not likely to happen).

So that brings me to my point, which is that however brilliant a person is at the IPA, they should not be adding IPA (or any other) pronunciations for any language of which they are not either a native speaker or have a native-like proficiency. Is there any way this could be articulated as a guideline?

Awien (talk) 17:17, 27 November 2015 (UTC)

Another problem is that even when pronunciations can be sourced, they use different notation. Changing the source notation to the English Wikipedia notation is again likely to involve OR. Thus gardening books in both the UK and the US often use respelling methods to show the pronunciation of scientific names, but converting these to any form of IPA requires knowledge of the intended dialect. Peter coxhead (talk) 22:49, 27 November 2015 (UTC)

Awien, your proposed solution doesn't have a problem attached to it. What is that supposed to solve?

Peter, converting from one transcription system to another is not original research. If a source uses centimeters, we can use inches without worrying about it being original research. If a source uses a transcription system that is insufficiently clear, then it is not a reliable source for pronunciation. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 00:09, 28 November 2015 (UTC)

(Please bear with my lack of IPA), Aeusoes, but we saw the problem when the IPA buffs didn't know that quatre is often, very often, not pronounced kat ruh but kat; an earlier version had the impossible katr shu voh; I just found Diderot's name given as a disyllable instead of a trisyllable; Beauharnais was given as ending with a pronounced s till I deleted it; and so on. The people who added those incorrect pronunciations are simply not qualified to do so because their French isn't good enough, i.e. native or native-like (which mine is).

Also, the conversions between imperial and metric can be done automatically and are absolute, whereas conversion from spelling to IPA simply can't - people are making it up themselves and making mistakes.

We do have a problem.

Awien (talk) 00:51, 28 November 2015 (UTC)

Plus, converting centimeters to inches might not be a problem, but converting centimeters to newly invented *wikimeters might. Converting an IPA symbol to another IPA symbol with a newly invented use is basically the same thing. Some of us think this would be a clear case of OR. --mach 🙈🙉🙊 00:59, 28 November 2015 (UTC)

Currently, the way it works is if a pronunciation is challenged, a source is required. Otherwise, the unverified pronunciation persists with de facto approval. In the case of quatre, sources are what solved the problem.

Rather than bar non-proficient speakers, the better solution would be additional scrutiny by knowledgeable editors. I am by no means a speaker of Russian, but I am knowledgeable about Russian phonology and periodically go through pages that link to Help:IPA for Russian to fix transcriptions. In your proposed solution, I would not be allowed to do this. The same for Help:IPA for Spanish.

As a native speaker, you can use your expertise to fix transcriptions. Barring non-native speakers wouldn't accomplish anything and opens up a whole can of worms regarding the spirit of Wikipedia. Also, sometimes native speakers are simply wrong about things about their language, including phonetic transcriptions. Placing some sort of convention that gives native speakers preferential treatment I think would be counterproductive.

You might also want to take a look at the explanatory notes at some of the IPA for X pages. If you think an explanatory note could help other editors make better transcriptions, then it is warranted.

And it is false that one can't convert from a respelling system to IPA automatically. We don't do that, but we could if we wanted to. That sort of a conversion is a feature of dictionary.com entries. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 02:09, 28 November 2015 (UTC)

How do you mean we don't? "We" enforce a "rule" that insists that IPA (or Wikipedia warped pseudo-IPA, in the case of English) must be used in nearly all cases, even if we have a respelling pronunciation given. So if the source gives the pronunciation as a respelling, we must (if we are to give the pronunciation at all) convert from that respelling system to IPA. Is your point something to do with this process not being automatic? W. P. Uzer (talk) 11:45, 29 November 2015 (UTC)

Yes. I was responding to the claim that conversion from a spelling system to IPA couldn't be done automatically. We don't have such a system in place for English. We do for Polish (see {{IPAc-pl}}). We could possibly do something similar to dictionary.com that allows readers to switch between them. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 19:40, 29 November 2015 (UTC)

I actually never said anything about converting from a respelling system to IPA. I was talking about people going directly from written French to IPA when their command of the language is insufficient. It's clear that's happening, and those people should be discouraged because too many mistakes are being made that could remain uncorrected indefinitely. It may sound harsh, but maybe we do need to strictly enforce the no original research rule by requiring a proper source for all pronunciations given.

Also, if W.P. Uzer is right in saying that a non-standard version of the IPA is being used, that is simply unacceptable. The whole point of the IPA is for it to be universal, and WP can't go making up its own version. We have serious issues here. Awien (talk) 22:02, 29 November 2015 (UTC)

I think you're generalizing too strongly from the one example. On top of that, the amount of effort it would take to strictly enforce VER on all French transcriptions would take more effort than for those proficient in French to go through and check the transcriptions for accuracy.

Regarding English transcriptions, W.P. is one of a handful of users who disagree with the IPA transcription system as it is used at Wikipedia. I (and many others) contend that it is neither OR nor "non-standard" (whatever that can even mean with the IPA). Take a look at the explanation of how the system works at Help:IPA for English and, if you dare, the archives where criticism of it was addressed. If you feel you have something new to say, you are welcome to do so at Help talk:IPA for English. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 00:10, 30 November 2015 (UTC)

It's not really fair to characterize those who disagree with you as "a handful" and those who agree with you as "many". Awien and others can take a look, among many other places, a little way up this page to see just some of the criticism that hasn't been and can't possibly be "addressed", because the system unquestionably is both OR (how can it not be, when it causes editors to make statements about Anglosphere-wide pronunciation that no reliable source has ever got close to making) and non-standard (I don't know how anyone could possibly dispute that - it takes IPA symbols, or in some cases made-up IPA-like symbols, and gives them arbitrary meanings that no-one can possibly deduce without inside knowledge). I genuinely can't understand why you of all people - an apparently knowledgeable, clever and experienced linguist and Wikipedia editor - continue to defend this obviously absurd system of misinforming readers about pronunciations, in spite of all the overwhelming arguments that have been reiterated time and time again against it. W. P. Uzer (talk) 07:46, 30 November 2015 (UTC)

A minority is a minority, no matter how right you think you are. What I see here, at IPA for English, and even NOR/N is a very vocal minority arguing against everyone else. The particular subset of this minority opposing on the grounds of OR is even smaller. A number so small you can count them on one hand. It certainly isn't the other way around. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 05:22, 1 December 2015 (UTC)

Is it really a minority? My impression is that opinions are divided evenly – and that it is only your perception that makes you think your own POV is the majority POV. --mach 🙈🙉🙊 09:55, 1 December 2015 (UTC)

I'm not sure how you're counting, but in the recent discussions I'm seeing a 2:1 ratio between people who are overt supporters of the system and those who disagree with it. I'm not counting Petercoxhead or Darkfrog as either. Those who cite OR concerns specifically are pretty much just you, Bazj (who hasn't really participated in the conversation), and WP Uzer. I might be counting differently than you because some comments have been specifically about opposing certain parts of the system, not the system as a whole. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 17:46, 1 December 2015 (UTC)

The numbers are barely relevant - has anyone actually addressed the OR concerns? For me they are simply unanswerable - the current system leads to statements that can have no reflection whatever in any reliable sources we know of (for the various reasons given above). The only attempted excuse I've heard is WP:IAR, on the grounds that the present system has some kind of "practical" value, but it's been shown time and time again that far from having practical value, it simply provides people with wrong or incomprehensible information. A sensible respelling-based system would probably be of far greater practical value. W. P. Uzer (talk) 10:43, 3 December 2015 (UTC)

You have a right to your opinion, but it's clear that the community doesn't agree with your take on this. There's no need to rehash what has already been discussed if nothing new will come of it. Indeed, it may even be considered a form of disruption. As can ballooning discussion of changes to Help:IPA for English in multiple places. As can hijacking threads to make them about something completely unrelated. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 21:15, 6 December 2015 (UTC)

There is no evidence whatsoever that “the community doesn’t agree” with W. P. Uzer’s (or my) take on the issue. That is just your POV. Your claim that there is a “2:1 ratio” in favour of your POV is unfounded because you have only counted one side, and only selectively. But even if it were true (I believe it is not), then a 2:1 ratio would not automatically invalidate the original research concerns that are the topic of this discussion. The original research concerns should be addressed directly nonetheless. And please stop accusing other users of disruptive behaviour. --mach 🙈🙉🙊 23:19, 6 December 2015 (UTC)

You two are talking out of both sides of your mouth. First you quibble about the number of editors who support your stance and then say that these numbers don't matter. Now you're making up whimsical accusations about how I'm categorizing the stance of other editors (counting "only one side" doesn't even make any sense in this context) and then saying that it doesn't matter what other editors think when they disagree with your interpretation of NOR.

I politely decline your request regarding calling out disruptive behavior. I would rather gently remind people of proper talk page etiquette than make a case for formal sanctions. Ignoring these reminders is your choice. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 01:40, 7 December 2015 (UTC)

I am not “talking out of both sides of [my] mout” – what I am doing is pointing out constantly that you misrepresent your POV as if it were an accepted majority view. I am not “making up whimsical accusations” – you have indeed only counted one side, and selectively [1]. Your unbased claims and accusations are tiresome. You are yet again shifting the discussion away from its topic (original research) to an unrelated sidetrack (in this case, the behaviour of other editors). Please return to addressing the original poster’s original research concern. If you do not want to return to the topic, I invite you to continue this exchange on my talk page. --mach 🙈🙉🙊 07:00, 7 December 2015 (UTC)

@W. P. Uzer: we would have a more useful conversation if, as I've pointed out several times before, the issue of OR was laid to rest. Choosing a notation within a publication and using it consistently is not "original research", however many times people claim that it is. The important issue is whether the system used in the English Wikipedia is as helpful as it could be. My view at present is that it probably isn't, because although there is not one consistent use for IPA symbols, there are ranges of use expected by IPA users, and a small part of the English Wikipedia system does depart from these. What we need to discuss is whether there is a better system, remembering the inherent limitations of an international encyclopedia that allows different English variants to be used. But where are the specific proposals for a different system, which still meets the required aims? Peter coxhead (talk) 11:23, 30 November 2015 (UTC)

Sorry, it is definitely OR, not (only) because of the idiosyncratic use of notation, but because of the synthesis and unsourced generalization involved (no source gives pronunciations that are intended to cover all varieties of English, so how on earth can we attempt to do so?) But it is the idiosyncratic use of notation that seems the major practical defect in the system. As to alternatives - why do you think there has to be a system? We don't systematically give pronunciations for every headword, like a dictionary does. The method that works best may differ depending on the word at hand. In most cases the answer the present system gives is probably perfectly fine. But in the cases where we are currently using symbols with made-up meanings that cannot possibly do anything else than mislead readers (in the multiple ways that have been pointed out many times), almost any rational alternative is better. In some cases a respelling will work best. In other cases IPA will be necessary - but it should be indicated what variety(ies) of English is being represented (footnotes can always be used for more extensive information). We don't need one size fits all. W. P. Uzer (talk) 14:14, 30 November 2015 (UTC)s

Well, we're not going to agree over OR. [W]hy do you think there has to be a system? I don't think a priori that there has to be a single system (for English), but a "good enough" one has considerable advantages over providing a separate representation for every ENGVAR. "The best is the enemy of the good" applies here. As soon as you label a pronunciation with a national origin, experience shows that nationalists from elsewhere will want their versions to be included. Who is going to accept their variant being consigned to a footnote while someone else's appears in the lead section? I can't see that going down this route will be helpful to readers. Peter coxhead (talk) 15:56, 30 November 2015 (UTC)

Perhaps we could adopt the principle that the (first) pronunciation shown is the one that corresponds to the variety of English used in the article (per ENGVAR)? Certainly it seems absurd (on top of all the other absurdities we have already identified in relation to the present system) to give, for example an American pronunciation for Chester and an RP one for Boston, as we currently do. W. P. Uzer (talk) 10:36, 3 December 2015 (UTC)

We don't do either of those things. Both are simply English pronunciations. /r/ doesn't make s.t. GA, and /ɒ/ doesn't make it RP. You're basically claiming that the UK needs to change the spelling of "Chester" to "Chesta" and that the US needs to change "Boston" to "Bawstn". — kwami (talk) 21:23, 7 December 2015 (UTC)

I don't think WP is arguing that we do do that. He's saying that, instead of the diaphonemic system, we could adopt a system where we put both RP and GA pronunciations except for instances that are clearly more relevant to one variety or another (such as geographical locations). 00:46, 8 December 2015 (UTC)

Yes, except that for the latter type of instance, we can also give both RP and GAm (and any others, such as regional and local dialects), provided we have references for them; the less "relevant" pronunciations would then either come after the most relevant one(s), or be given in a footnote, if clutter is felt to be excessive. W. P. Uzer (talk) 13:01, 9 December 2015 (UTC)

/r/ doesn't make s.t. GA – or does it? We want to be pan-dialectal. However, others who want to be pan-dialectal find a way for comforting both rhotic and non-rhotic pronunciations. The CGEL, for instance, uses the symbol ⟨ʳ⟩ ([2], p. 16), or Daniel Jones (phonetician)’s English Pronouncing Dictionary provides variants. Rigidly insisting on /r/ is a one-sided solution – just as one-sided as if we were only providing non-rhotic pronunciations and then adding a note that says: “In American pronunciation, every ‘r’ of the ordinary spelling is retained” (cf. OALD, inside back cover).

We should provide a system that is flexible enough. Otherwise, editors will not use it. I have already encountered several non-rhotic placenames without an /r/. I see no reason why we should impose a rigid standard. We do not impose a rigid standard for normal spelling either. --mach 🙈🙉🙊 16:50, 9 December 2015 (UTC)

Possible POV and a fallacious justification in this talk page’s introduction

The introduction to this talk page includes a list of perennial topics. Among them is the following:

That the current "pan-dialectal" English convention at Wikipedia:IPA for English is arbitrary/unreferenced/original research, and is therefore invalid. However, this is a help key, not an article, and no more needs to follow Wikipedia's no original research policy than the pronunciation guide of any other encyclopedia or dictionary, nearly all of which use in-house conventions.

— kwami, [3]

I have several concerns about this introduction text:

1. Does this really represent a consensus? I have found no evidence that it has ever been explicitly discussed, and after kwami introduced it, it has remained unchanged (except for one linkfix by kwami [4]).
2. Even if it turns out to represent a consensus, the justification is fallacious. While it is obvious that MOS pages are not articles and therefore do not have to be verified, the point of our pronunciation symbol schemes is that they be used in the article namespace – and whatever is used in an article has to be verified.
3. Why don’t we use the Template:Round in circles?

I would not normally care much about the introduction to a talk page, but it has recently been quoted in order to show that this question is “well-settled” [5]. I wonder, is it really? Subsequently, Ƶ§œš¹ has tried to argue that it really represents a consensus (see [6], [7]), but his evidence has been indirect at best (see [8], [9]). That is why I am now asking here directly: Is this really based on consensus?

Please do not jump on the original research issue. The introduction text does not only talk about the pronunciation schemes being “original research”, but also about them being “unreferenced”. My impression is that very few editors would advocate for unreferenced pronunciation symbols, whether or not they think unreferenced pronunciation symbols are original research (some strongly think it is not, while others strongly think it is – we have been over this time and again). --mach 🙈🙉🙊 08:35, 7 December 2015 (UTC)

Yes, we have been over this time and time again, and I see no consensus supporting your view that very few editors would advocate for unreferenced pronunciation symbols. The consensus view is, and has been for many years, as stated at the head of this article, namely that there is no need for referencing in the Manual of Style. SMcCandlish has explained this in great detail at Wikipedia talk:Manual of Style#Wikipedia:Manual of Style/Supports. (This comment does not imply support for the presently used notation.) Peter coxhead (talk) 14:10, 7 December 2015 (UTC)

Right. This is more rehash, the third overlapping attempt to WP:FORUMSHOP this, using exactly the same arguments that have already been refuted. {{Round in circles}} applies when a discussion is going nowhere because consensus is not emerging, which is not the case here; consensus already exists. WP:IDHT and WP:1AM are what apply when one editor refuses to get it and consensus is not changing to suit that person's demands. Our Help:IPA for English system does need work, but it's because it's potentially confusing in an incoming way to people already familiar with actual IPA, and in an outgoing way to people who think they're learning actual IPA from WP's internal adaptation of it. It's a usability problem, not a WP:CCPOL problem; CCPOL does not apply to internal documentation. CCPOL does not apply to formatting templates, it only applies to content. WP's pronunciations keys (there are at least two of them, one not based on IPA) are not content; the pronunciation conveyed by following either of the keys is the content. We've been over this several times already with the same editor. Ignoring refutations and simply restarting back from one's original position as if no one addressed it is what is fallacious here (namely "proof by assertion").  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  20:04, 7 December 2015 (UTC)

OK, so you are saying more or less that there can be consensus by force of habit, where nobody can point to any particular place where this consensus has ever been formed because people only noticed after some time that the consensus was there.

Why don’t you appreciate that I am not alone in questioning this consensus? I have demonstrated that I am not alone [10]. Please don’t pretend I haven’t.

Ignoring refutations and simply restarting back from one's original position as if no one addressed it is what is fallacious here. – I take exception to that accusation. I have not ignored any refutations. It rather seems to me that you, SMcCandlish, are ignoring my pointing out the weaknesses and contradictions in your alleged refutations, see Wikipedia:No original research/Noticeboard#Wrong venue; internal documentation is not subject to WP:NOR.

The consensus view is, and has been for many years, as stated at the head of this article, namely that there is no need for referencing in the Manual of Style. – Well, the time-proven consensus on Help:IPA for English seems to be a different one since there is indeed referencing on that page.

@SMcCandlish ☺: I really do not get your POV. You have said “it’s a terrible, terrible idea to warp a globally accepted standard like IPA” – but now it seems you are saying we should not care for sources? Have I misunderstood you? I see no difference between referencing to sources and caring for accepted standards. It seems to me that you are de facto sticking to WP:CORE by saying pronunciation symbols should be based on what is being used outside of Wikipedia (unless I have misunderstood you), but you very strongly refuse mentioning WP:CORE. --mach 🙈🙉🙊 00:47, 8 December 2015 (UTC)

It certainly seems bizarre that on a topic where there is obviously a very serious lack of consensus, the talk page should announce what are claimed to be the arguments in favor of one position, ignoring all counter-arguments. (I've tried altering or removing some of the obvious falsehoods and absurdities, but changes just get reverted.) I'm not claiming my experience is all-encompassing, but I don't think I've ever seen anything like this anywhere else on Wikipedia. Similarly I've never seen any part of the substantial information found in WP articles asserted to be exempt from sourcing requirements (or requiring an unsourced Help page to enable readers - even theoretically - to make sense of it). This is all very weird - but I hope we are moving on to reaching a more sensible solution. The pronunciations given for the vast majority of articles, I suspect, are perfectly fine as they are; we just need to find a way to deal with the minority that clearly are not. W. P. Uzer (talk) 13:09, 9 December 2015 (UTC)

To repeat again what seems obvious to me and other editors: the pronunciations themselves need sourcing; the notation used does not (but should be as helpful as possible to readers). Peter coxhead (talk) 16:49, 9 December 2015 (UTC)

And to repeat again what seems obvious to me and other editors: The notation should have sources just like everything else in the main article namespace – there is no reason for having a special exception. Let’s agree to disagree and move on to something more productive. --mach 🙈🙉🙊 17:03, 9 December 2015 (UTC)

(replying to Peter after edit conflict) Yes, but the alleged pronunciations are "diaphonemic" for English as a whole and are therefore almost certainly not sourced or sourceable; and the notation used (at least in a certain small number of its aspects) is not helpful to readers, for the reasons that have been stated on many occasions and I think are basically accepted. The placing of one side's arguments permanently at the top of a talk page, without allowing anyone to amend them or refute them or indicate that they are challenged, is, to put it mildly, not what one would expect. W. P. Uzer (talk) 17:15, 9 December 2015 (UTC)

RfC: Should we continue recommending the sign ⟨ɵ⟩?

There is an RfC at Help talk:IPA for English#RfC: Should we continue recommending the sign ⟨ɵ⟩?:

We are currently recommending that the sign ⟨ɵ⟩ be used for a reduced vowel diaphoneme that can correspond either to the phoneme /oʊ/ or the phoneme /ə/, for example in the word omission. Should we continue to do so?

In the IPA, the sign ⟨ɵ⟩ represents the close-mid central rounded vowel. Our use of ⟨ɵ⟩ is based on Bolinger, Dwight (1986), Intonation and Its Parts, Stanford University Press, pp. 347–360. Bolinger proposed not to analyze the reduced vowels as mere versions of the full vowels, but as a special set consisting of three vowels: The “fronted” Willie vowel /ɨ/, the “central” Willa vowel /ə/, and the “backed” willow vowel /ɵ/. Bolinger’s use is slightly different from ours (1) because Bolinger’s /ɵ/ is not a diaphoneme, but a phoneme; (2) because there are words such as willow or lasso or the second part of the MOUTH diphthong where Bolinger would use /ɵ/, but we would not; and (3) because Bolinger’s analysis allows for an alternation between /ɵ/ and /ə/ in words such as canopy, Indonesia, allophonic, or composition.

The English Wikipedia is the only major dictionary that has adopted Bolinger’s sign ⟨ɵ⟩ in a broad phonemic IPA transcription scheme (for an overview of other dictionaries’ broad phonemic IPA transcription schemes, see Help:IPA conventions for English#Reduced vowels).

If you are interested in the edit history of ⟨ɵ⟩ on Help:IPA for English, you may want to check out User:J. 'mach' wust/sandbox#Edit history of ⟨ɵ⟩ (with diff links etc.).

Thanks for your comments. --mach 🙈🙉🙊 12:44, 16 December 2015 (UTC)

Rules about inclusion of pronunciation?

Are there any rules as to which pages should/can have pronunciation guides and which shouldn't? I'm having a very weird episode over on the Hogmanay page with someone who seems to have a pet hate of pronunciation while I'm arguing that Hogmanay is hardly a clear-cut case of pronunciation, especially for non-Scots. Akerbeltz (talk) 12:21, 3 January 2017 (UTC)

Akerbeltz I don't know of any, but I would argue any non-english word (excluding common loan words), or any rarely used english word (inclusive of loanwords) would have it be pronounced, as in my opinion, words sound weird only if they aren't heard a lot. Iazyges Consermonor Opus meum 00:25, 5 January 2017 (UTC)

Delete {{USdict}} from this manual page

Resolved

 – Done, since the template doesn't even exist any longer.

This manual page mentions the respelling template {{USdict}}, but only very hesitantly. It says that such respellings “are widely understood only in North America”, and it continues with a sizeable quote from Wells that strongly opposes them.

I think we better delete all references to {{USdict}} on this manual page instead of hesitantly mentioning it. --mach 🙈🙉🙊 11:56, 21 January 2017 (UTC)

RfC: remove Respell (pronunciation) from Infobox element

• An RfC is open at Infobox element. Proposed is to remove {{Respell}} wholesale, leaving IPA. -DePiep (talk) 23:13, 7 November 2017 (UTC)

Other transcription systems - homonym

The examples of using words for pronouncing are not consistent using italics or quotes, both in the MOS and in the articles, i.e.:

• MOS vs article
• Peter Coke "cook"
  • Peter John Coke (/ˈkʊk/ "cook"; 3 April 1913 – 30 July 2008)
• RAZR (pronounced as razor)
  • The Motorola Razr (styled RAZR, pronounced "razor")
• KIOTI coyote
  • KIOTI Tractor (/kaɪˈoʊti/ "coyote")
• C++ as "pronounced cee-plus-plus"
  • C++ (pronounced cee plus plus /ˈsiː plʌs plʌs/)

So do we agree on the preferred:

1. style: double quotes vs italic: say italic pronounced as razor, (which is also consistent with italics created by {{Respell}})
2. location before/after IPA (say after IPA)
3. use quotes in the MOS to show this style for clarity Widefox; talk 14:56, 7 November 2017 (UTC)

Other article examples that were double quotes but I've changed to italics:

• javac javac (pronounced java-see)
• SPDY SPDY (pronounced "speedy")

Widefox; talk 21:48, 7 November 2017 (UTC)

Do we care? MOS:WAW says either can be used; I think the default is italics, but it depends on the context; if the material is already italics-heavy for other reasons, prefer quotation marks.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  13:26, 8 November 2017 (UTC)

That's my understanding too. The examples are already in the MOS, so seems that we do.

Does anyone object to clarifying the MOS for the current examples and their articles to consistently use italics, with exception examples of double quoted examples for italic heavy context? Widefox; talk 12:39, 9 November 2017 (UTC)

I don't in theory, I guess it would depend on the exact edit. Could you take the quoted version and do a mockup of the changes?  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  14:36, 9 November 2017 (UTC)

I took the bold step of clarifying and fixing the current examples in the MOS directly, and making the articles in those examples consistent. There's a couple of unresolved inconsistencies that I html commented. I don't have an example for using double quotes, and agree that should be worked out here. Widefox; talk 14:07, 10 November 2017 (UTC)

Seems reasonable enough. I reformatted it a bit, code wise, e.g. to stop monospacing the output examples. Use of the "indent to get <pre>...</pre> effect" trick is not a universal tool.  :-)  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  17:50, 10 November 2017 (UTC)

Yes that looks better. I'm more used to the example style in WP:MOSDAB which is very consistent. In this MOS, the example style could be made more consistent which would help. Regards Widefox; talk 22:58, 10 November 2017 (UTC)

Many of these pages need considerable work in multiple ways.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  05:08, 11 November 2017 (UTC)

Italic makes the word hard to distinguish from a respelling. It should be quotes to indicate it's a word as a word and not some artificial representation of a pronunciation. Nardog (talk) 11:27, 8 December 2017 (UTC)

I'll buy that.  — SMcCandlish ☏ ¢ >^ʌⱷ҅_ᴥⱷ^ʌ<  12:57, 8 December 2017 (UTC)

Italian dialects help

I currently have a draft ready for a page regarding Italian dialects (where {{IPA-itdia}} is going to point): what am I supposed to name the page? Thank you :) イヴァンスクルージ九十八_（会話） 21:03, 6 August 2018 (UTC)

Never mind, I placed it at H:IPAITDIA. イヴァンスクルージ九十八_（会話） 09:06, 9 August 2018 (UTC)

Show tooltips for all IPA characters instead of useless "Help:IPA/<foreign language>"?

There is currently a tooltip system that allows for hovering over IPA letters to get an example pronunciation of the IPA symbol with an English word -- but only if the word was originally from English. The vast majority of the time that you need pronunciation assistance though, it's when the word is from a foreign language!

Can we change this so that ALL the IPA pronunciations show the English example word tooltips when you mouse over each character instead of showing the useless tooltip "Help:IPA/<foreign language>"? I think this would be a major improvement for usability of IPA on Wikipedia.

Unfortunately, I am not well-versed enough in the plumbing of Wikipedia templates to figure out how to make this technical change myself. I would appreciate it if someone else could make this change.

172.58.27.32 (talk) 01:36, 6 May 2019 (UTC)

I don't see what you mean. Useful to which user group? I mean, what does a useful tooltip for the IPA symbol [ʕ] look like? "voiced pharyngeal fricative or approximant"? Or maybe "the consonant in the middle of the Arabic given name Sa‘id"? (BTW, I am only fluent in half a dozen languages, so not all the language examples are of use to me. But I'm fine with articulatory descriptions of sounds, so I'd opt for that.) Love —LiliCharlie (talk) 04:30, 6 May 2019 (UTC)

I think I understand what the anon is asking for and we shouldn't use English examples for non-English words, even when a foreign word has close approximations to English sounds. I suspect that having a phonetic descriptor like "voiced pharyngeal fricative or approximant" would have limited utility. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 21:05, 6 May 2019 (UTC)

The French Wikipedia has a module that essentially achieves this, but I think it seriously misrepresents what the IPA is and we shouldn't have one here (nor there). When we write [ʁ] linking to Help:IPA/French, we do not mean a voiced uvular fricative but the phoneme /r/ of French that can have any realization, be it a voiced uvular fricative, a voiceless velar fricative, a uvular trill, an alveolar trill, or what have you. As the second clause of the current Principles of the IPA succinctly states, each IPA character is a shorthand for a phonetic value that can have vastly differing degrees of specificity and meaning. Nardog (talk) 21:21, 6 May 2019 (UTC)

I'm the previous anon replying again (sorry for delay in response -- was stuck on cellphone tethering for internet for the past week, and Wikipedia apparently HATES T-Mobile). Useful to which user group? Useful to non-experts who know English trying to find out what something is in the first place! (i.e. the primary audience of the English language Wikipedia.) If you see a term and don't know what it is, looking it up on Wikipedia is the natural first thing to do for a great many people! It would be really helpful if looking up the term also allowed the reader to find out how to pronounce it. The pronunciations *are* given for many words, but most readers of Wikipedia -- myself included -- are not fluent in reading IPA. For English language words, tooltips are given that are very, very helpful for quickly finding out what an unfamiliar IPA symbol means. Even though I've already memorized some of the symbols through repetition, there are inevitably symbols I've forgotten, and I have to go consult a table to find the symbol I don't know then go back to the pronunciation then usually go back to the table and so on and so forth... instead of just mousing over the ones I don't know and getting an example. What I'd like is an English language example *wherever it is possible*. If [ə] shows up in a foreign word, for example, there is no reason it couldn't tell the reader that it is pronounced like the "a" in "above" (which is what is used on Help:IPA). This is already done for English words, just not foreign words -- where you most commonly need pronunciation assistance! (The other common case where you really need pronunciation assistance, of course, is for names.) For symbols where there is no good way to give an English word with the sound (like your [ʕ] example), a tooltip with an explanation similar to what's on Help:IPA and/or an example from a foreign language like the Sa'id explanation would be best, I think. (Perhaps for that one, "A light sound deep in the throat. Does not occur in English. The consonant in the middle of the Arabic given name Sa‘id." would be good -- though that might be a bit long.) It might take a while to come up with good, concise tips for sounds that have no equivalent in English, but there are already explanations that could be used as a starting point on Help:IPA. To give a real-world example, I was looking at the entry for Spaghetti aglio e olio and trying to decipher the pronunciation [spaˈɡetti ˈaʎʎo e ˈɔːljo] before I made my request here. If you mouse over it, it annoyingly gives "Help:IPA/Italian" instead of examples for symbols like ʎ which I had forgotten. If you look at the help page, ʎ shows an example indicating that it is like the "lli" in the middle of the English word "billion" -- and pretty much every IPA example for Italian *does* have an English equivalent there. For cases like that, there's no reason a tooltip with the example couldn't be shown -- it's just that no one has implemented it yet. I would greatly appreciate it if someone could implement this, and I think it really would help a lot of people out, not just me. 76.167.106.227 (talk) 02:47, 13 May 2019 (UTC)

CfD on categories for pages including recorded pronunciations

There is a discussion on categories named "Pages including recorded pronunciations" or "Articles...", which editors here may be interested in. Nardog (talk) 17:48, 27 January 2020 (UTC)

References

A recent edit added the text "A citation is not required for a pronunciation, since citations are not always available. (In some cases, the converse is true: A pronunciation, while not intuitive, is well known and easily verifiable.) However, if the pronunciation is contested by a user, or if it is subject to debate (for example, in written or recorded sources), then a citation should be provided, and the pronunciation may be removed if a credible source is not provided."

Words whose pronunciations are genuinely "well known" don't require it to be given here. Given the variety of dialects of English and the variation both between and within English-speaking countries, words that do require pronunciations in the text need citations. The argument that citations are not required because they are not always available is clearly against the fundamental principle of verification. Peter coxhead (talk) 05:23, 13 May 2020 (UTC)

Which pronunciations are "well known" is a rather subjective issue. For me as a non-native speaker I certainly appreciate a generic dia-phonemic pronunciation, added by a well educated native speaker, even when no sources can readily be found. English orthography is a disaster for deriving pronunciations. −Woodstone (talk) 10:39, 13 May 2020 (UTC)

Yeah, if the pronunciation of a word is "well known" then it most likely doesn't need to be spelled out in the first place (see § Appropriate use), and if a source is not available then it should not be added at all. And that unsourced material should be removed if challenged goes without saying as it's already covered by WP:Verifiability. Nardog (talk) 13:53, 13 May 2020 (UTC)

Completely agree: pronunciations are not exempt from WP:VER (and, in my opinion, neither is the use of pronunciation symbols). --mach 🙈🙉🙊 18:04, 13 May 2020 (UTC)

Get rid of these stupid IPA pronunciations entirely

Split from above, this is a different conversation entirely. Nardog (talk) 22:18, 23 May 2020 (UTC)

I have a better idea: get rid of these stupid IPA pronunciations entirely, because no one but phoneticists understands them anyway. Use a pronunciation system normal people can make sense of. EEng 15:46, 23 May 2020 (UTC)

• Sounds good, but there isn't one that works for all dialects of English. How do you represent the different pronunciations of the vowel in dog in English dialects other than in the IPA? Respelling doesn't work for those (e.g. most Americans) who don't have the phone I use (as I'm from southern England) in their dialect. Peter coxhead (talk) 20:18, 23 May 2020 (UTC)

  Then have two or three versions, each targeted at a different group. I'm serious about this. If you want people to get anything out of this at all it needs to made accessible; otherwise just forget it. Nobody knows what all those upside down thingamajigs and heavy-metal umlauts and runes mean. And it's not like they're 90% comprehensible and people might pick up the rest with time; the whole system is impenetrable. I understand it has its function for specialists but that's not our audience here. EEng 23:08, 23 May 2020 (UTC)

English spelling is impenetrable. We ought to rewrite all articles in IPA and banish English colonial orthography to a footnote where it won't annoy and confuse people. Love —LiliCharlie (talk) 00:26, 24 May 2020 (UTC)

• The above discussion is about whether to put pronunciations in parentheses, not whether to have them at all. You can advocate against them all you want separately, but don't hijack a discussion and turn it into something whose outcome has far greater consequences. Nardog (talk) 22:18, 23 May 2020 (UTC)

  Whether or not to put in parens something utterly unhelpful to anyone but specialists seems quite inconsequential. EEng 23:08, 23 May 2020 (UTC)

  I'm sorry, EEng, I need those. As a Wikipedia reader, I frequently check here for things like which ü, which sibilant, whether there's a secondary stress, and whether the US pronunciation differs from the British, and don't ask me to use aural media instead, I'm increasingly deaf. My beef with them is that we are insufficiently descriptive, for instance including the purely theoretical <r> in some English placenames. (As a geek, I also like to use the pronunciations to see what ways IPA has evolved to represent some of the more challenging features.) As a speaker of a godsawful Mischsprache myself, I realize there's a problem with multiple regional and class dialects, but I do think accurate representation of the major pronunciations in an internationally accessible way should be part of our encyclopedic aim. IPA must be murder on screenreaders but in every other respect (including those of us who can't hear and the sounds that simply can't be clearly shown by respellings, like all the non-English ones and the "th" pair) it's hands-down best. Yngvadottir (talk) 23:30, 23 May 2020 (UTC)

  Careful with the talk about murder or T&S will get you. EEng 03:33, 24 May 2020 (UTC)

• The notion that nobody but linguists uses the IPA seems to be an American misconception. British dictionaries widely use the IPA, and I think educated readers who are familiar with using such dictionaries can reasonably be expected to understand them. --Paul_012 (talk) 02:56, 24 May 2020 (UTC)

  No wonder you guys have a different dialect every 10 miles. EEng 03:33, 24 May 2020 (UTC)

• I move that we close the discussion, per WP:SNOWBALL, WP:FORUM, and WP:TPG#YES. Snippy comments from all parties and the absolutely clear nature of the consensus to keep the status quo shows that furthering it will be counterproductive as it will erode parties' good will to each other and waste everyone's time. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 15:15, 24 May 2020 (UTC)

  Calm down. Let the issue be explored. For all you know someone may come up with a happy solution. EEng 16:16, 24 May 2020 (UTC)

cluttering with double brackets

I notice that a lot of pronunciations are double-bracketed. Sometimes that's necessary, when the pronunciation info is complicated, but for simple entries it's just clutter. The IPA templates already provide brackets (virgules or square brackets, depending), and so don't need parentheses as well.

So, instead of

FGH (/fɡh/) is the blah blah blah ...

or

FGH (/fɡh/ ) is the blah blah blah ...

we should avoid the redundancy and just format it

FGH /fɡh/ is the blah blah blah ...

or

FGH /fɡh/ is the blah blah blah ...

That's the purpose of the IPA delimiters, after all. The extra parentheses aren't meaningful. Respellings should probably use parentheses, though, since the template doesn't supply delimiters.

FGH /fɡh/ (ef-gee-AITCH) is the blah blah blah ...

That way it's clear that the respelling is an explanation of the IPA, functioning something like a sound file.

— kwami (talk) 01:14, 23 May 2020 (UTC)

Pronunciations in the lead should be kept in parentheses because they are parenthetical, i.e. non-essential. In other words, they are not meant to be read separately, only to provide auxiliary information related to (the word that signifies) the subject.

A pronunciation notation should not be taken out of parentheses unless the pronunciation itself is being discussed, as in The word X is pronounced /ɛks/ or The pronunciation /ɛks/ is used. In comparison, e.g. An X /ɛks/ is an instance of... is semantically inferior because the pronunciation is not the subject being discussed. Nardog (talk) 14:40, 23 May 2020 (UTC)

I agree with Nardog's reading. Unlike some other encyclopedias, Wikipedia's style is to have the beginning of an entry read as a full sentence; plugging in a pronunciation without parentheses would disrupt the flow and break it. It might be okay if we changed the style to not have the opening be a sentence, e.g. "FGH /fɡh/ : The blah blah blah ...," but that is probably never going to happen. --Paul_012 (talk) 03:07, 24 May 2020 (UTC)

They aren't meant to be read separately regardless of whether we use parentheses, so they don't disrupt the flow either way. They're equivalent to furigana for difficult words in Japanese, and no-one would put those in parentheses. When read out loud, the following are identical:

• FGH is the blah blah blah ...
• FGH /fɡh/ is the blah blah blah ...
• FGH (/fɡh/) is the blah blah blah ...

Given that, what do the parentheses add, other than clutter?

This is a bit like leads that read,

• The Kabba, also known as the Kaba, Cabba or Caba, are...

Read that out loud, or use a screen reader, so the word is repeated three times, and it sounds idiotic. The same with the IPA -- you wouldn't actually say the word twice, regardless of parentheses, and the IPA delimiters already mark it as pronunciation, so I don't see how the parentheses have any function. — kwami (talk) 08:51, 25 May 2020 (UTC)

Those alternate names are still meant to be read, albeit perhaps not out loud. A pronunciation is not, in that it's not part of the prose.

In reality, though, I think I would oppose codifying one or the other in the MOS except to imply through examples, given WP:CREEP and the outcome of Wikipedia talk:Manual of Style/Lead section/Archive 19#Request for comment on parenthetical information in first sentence, which decided against micro-management. Nardog (talk) 17:42, 25 May 2020 (UTC)

Banish them to a footnote where they won't annoy and confuse people. EEng 23:09, 23 May 2020 (UTC)

We already do that (or should) with complicated pronunciation guides that would disrupt the lead or cause confusion. But sometimes we just need a simple guide for an ambiguous word. We can't do much about readers who were never taught how to read a dictionary. We can only dumb things down so much before the exercise becomes counter-productive. — kwami (talk) 08:58, 25 May 2020 (UTC)

You can't seriously be suggesting that anyone but specialists (and, if truth be told, almost none of them either) know that ɻ is Mandarin's "Voiced retroflex approximant", much less what to do with that information. EEng 19:57, 25 May 2020 (UTC)

Maybe you can point to a specific page's lede to elaborate on the point you're making, EEng. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 18:17, 26 May 2020 (UTC)

@Peter coxhead, LiliCharlie, and Aeusoes1: What say you? Nardog (talk) 17:42, 25 May 2020 (UTC)

• In this specific context, the extra bracketing is largely redundant, but removing it doesn't do anything to enhance the flow of the first sentence IMHO. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 19:01, 25 May 2020 (UTC)
• The "brackets" aren't just there to perform the usual function of parentheses in a sentence. The /../ and [..] mark the difference between phonemic and phonetic transcriptions. So there is a logic in also having parentheses. When I look at the opening part of Phoneme right now, which has:

A phoneme /ˈfoʊniːm/ is a unit of sound that distinguishes one word from another in a particular language ... the sound patterns /sɪn/ (sin) and /sɪŋ/ (sing) are two separate words

it does look wrong to me. The first phonemic transcription, "/ˈfoʊniːm/", is an aside, and is not part of the flow of the sentence, whereas the second two are an intrinsic part of the phrase "the sound patterns /sɪn/ ... and /sɪŋ/". To me it reads as though we are saying that "/ˈfoʊniːm/" is a phoneme, just as you would write "the phoneme /ŋ/ is part of most English dialects". So I would put parentheses in "A phoneme (/ˈfoʊniːm/) is a unit of sound". However, (disclosure) I've taught linguistics, so I pay more attention to the transcriptions than the average reader and am more aware that the /../ and [..] do not mark parenthetical parts of a sentence. Peter coxhead (talk) 07:49, 26 May 2020 (UTC)

Pronunciations should be moved to footnotes. Kaldari (talk) 03:09, 8 December 2020 (UTC)

Automating Template:Respell better

Is there a reason that an automatic invocation of {{respell}} can't be incorporated into {{IPAc-en}}? Since the conversion has to be done according to the respelling key, I'd think it'd be a simple enough matter to just have {{IPAc-en|ə|ˈ|k|ɪ|l|iː|z|respell=yes}} than to have to do {{IPAc-en|ə|ˈ|k|ɪ|l|iː|z}} {{respell|ə|KIL|eez}}. This would help us avoid repeating ourselves, making it easier if we ever decide to e.g. change our pronunciation key. Would anyone be interested in coding the |respell= parameter? {{u|Sdkb}} ^talk 01:44, 8 December 2020 (UTC)

There is no point in having respelling at all if it's just a translation of IPA. The whole point of using respelling is that you shouldn't need a key to understand it; it should use standard English pronunciation written in unambiguous ways. (For example, if we needed a respelling of "poke", it should just be "poke", and not for example "pohk", which could also be read as intending /pɑːk/.) --Trovatore (talk) 02:45, 8 December 2020 (UTC)

Trovatore, the first paragraph of the documentation at {{Respell}} states Notation must be spelled accordingly to the respelling key at Help:Pronunciation respelling key. I'm not too familiar with this area—is that consensus less settled than the documentation would make it seem? {{u|Sdkb}} ^talk 04:18, 8 December 2020 (UTC)

That is precisely what this guide proscribes. Nardog (talk) 17:06, 8 December 2020 (UTC)

It's wrong and should be changed. --Trovatore (talk) 17:58, 8 December 2020 (UTC)

While deriving the IPA from a respelling is easy, the reverse is not, because syllabification is only optional in IPA. This means that converting IPA to respelling would entail figuring out phonotactics and preparing for all the possible onsets, which I doubt is feasible as there must be outliers in non-native words, which many IPAc-en transcriptions we have are of. Nardog (talk) 17:06, 8 December 2020 (UTC)

Subscripts and superscripts

FYI, based on comments by Nardog and the tools, charts, and articles referenced by this page, I'm adding a note to Wikipedia:Manual of Style/Superscripts and subscripts that IPA templates should be used with Unicode subscripts and superscripts. This is contrary to the practice in almost all other cases (which is to use <sup>...</sup> and <sub>...</sub> or templates that make those) so I just wanted to make sure everyone is OK with that. (I have a bunch of edits to undo if so.) -- Beland (talk) 07:55, 5 March 2021 (UTC)

I'll also note this means changing some HTML superscripts to Unicode superscripts, for example on Voiced alveolo-palatal affricate. -- Beland (talk) 23:52, 9 March 2021 (UTC)

UPA subscripts markup

@Nardog, Uanfala, and 1234qwer1234qwer4: Continuing the conversation on my talk page, I've started auditing superscripts and subscripts in linguistics contexts. The first question I have is how to handle {{UPA}}. I see only one instance of HTML tags, namely in Uralic Phonetic Alphabet which says:

pᴞ·ń₍ᴅ́ᴢ̌́ö̭

marked up as

''{{UPA|pᴞ·ń₍ᴅ́ᴢ̌́ö̭}}''

Is that preferred or should it be converted to Unicode? -- Beland (talk) 03:02, 10 April 2021 (UTC)

In line with IPA, I would suggest Unicode. This is the reason why some of the subscripts like this were included in Unicode in the first place. 𝟙𝟤𝟯𝟺𝐪𝑤𝒆𝓇𝟷𝟮𝟥𝟜𝓺𝔴𝕖𝖗𝟰 (𝗍𝗮𝘭𝙠) 14:02, 15 April 2021 (UTC)

@Beland Do you have any updates concerning fixing markup like this? ~~~~
User:1234qwer1234qwer4 (talk) 14:00, 4 June 2021 (UTC)

@1234qwer1234qwer4: I changed the article Uralic Phonetic Alphabet to use a Unicode subscript, as requested. As for the changes we were considering to {{IPA}} and {{lang}} text and similar templates, it appears not all of the characters are available as Unicode superscripts. For example, I wasn't able to find a superscript version of "ʢ". After pondering it for a while, I think the proposed mix of Unicode and HTML superscripts is going to 1.) be a pain to figure out which are desired, and 2.) look messy, so I've decided to just leave that markup in the inconsistent state that it has always been in. When the next database dump arrives, my focus will simply be on making sure problematic text is inside one of these templates, so it isn't parsed as English text for spell-checking purposes. -- Beland (talk) 18:00, 4 June 2021 (UTC)

@Beland The character you are talking about will be added to Unicode with version 14. Maybe it is better to wait until September for now and then try to fix these up. ~~~~
User:1234qwer1234qwer4 (talk) 18:08, 4 June 2021 (UTC)

@1234qwer1234qwer4: That doesn't address the other reasons I'm not motivated to do this. If you would like to do so yourself, I'm happy to provide reports on which articles have which markup that you may want to look at. -- Beland (talk) 18:17, 4 June 2021 (UTC)

These characters are added to Unicode exactly to use them in phonetic transcriptions, so I do think that we should use them. I would wait until September though as I said, which would both provide more recent data and allow for replacement of a wider range of characters. ~~~~
User:1234qwer1234qwer4 (talk) 18:27, 4 June 2021 (UTC)

Double Chao tone letters

A common convention with level tones is for a single tone letter with a checked syllable, and a double letter with an unchecked syllable (e.g. [a˧˧] but [ak˧]. Completely redundant, because any long vowel will be marked for length. And anyway double tone letters occur with short vowels -- and how do we consistently handle long vowels in checked syllables? and the convention isn't possible with contour tones, which are similarly shortened in checked syllables. Plus some fonts that support contour tones don't render double tone letters.

Is this something we should fix, so we don't make a spurious distinction? It's easy enough to automate, and there's no loss of information, as long as we avoid contour tones that are partially level (like [a˧˧˥]). — kwami (talk) 15:21, 3 September 2021 (UTC)

Do we need to specify that current pronunciations are relevant?

Over at Talk:Joule, there's a discussion about which pronunciation to indicate. In short, the argument is that since some dictionaries published in 1901 list JOWL as pronunciation, that pronunciation is to be considered standard and the consensus pronunciation JOOL that has been established since should be listed as non-standard.

I thought it would help to point to the relevant section of this page which states that current pronunciations are to be preferred, and need current sources to back them up, but alas, that section does not exist.

Should we add it?

IpseCustos (talk) 12:15, 25 June 2022 (UTC)

I would be opposed to saying that more recent pronunciations are "preferred". In the case of a conflict, give both, and if sourceable, the trend can also be indicated. --Trovatore (talk) 17:39, 25 June 2022 (UTC)

I said nothing about "more recent", and of course I agree that a recently-introduced pronunciation isn't automatically to be preferred over an established one. According to most current dictionaries, JOWL is no longer a current pronunciation, and the article misrepresents that. IpseCustos (talk) 18:18, 25 June 2022 (UTC)

What counts as “tonal language” here?

This page prescribes a difference depending on whether a piece of content belongs to a “tonal language”. But it fails to define which languages count as such. Here, “tonal language” is not linked, but a link to our general article would be of little use, anyway. Searching for “tonal language”, one gets redirected to Tone (linguistics), which defines “tonal language” nicely theoretically, but not very usefully for this purpose.

Take for example this discussion about Japanese: The place name discussed there does not meet that definition per se. But while nobody there claimed that Japanese was not a tonal language, the consensus seemed to be that it be treated as if it didn't fall under the relevant clause of this MOS.

So, can we agree on a list of “tonal languages” for the purpose of this MOS? It doesn't have to be exhaustive; we can assume that editors, at least as a community, will be able to see analogies. (I hope we can agree not to need a higher level of granularity, such as for cases in which Japanese does distinguish words.) ◅ Sebastian 11:55, 27 March 2023 (UTC)