Template talk:IPA

Removal of " font-family /**/:inherit;"

This template didn't work as well in Wiktionary as in Wikipedia. As Template:IPAchar seemed to work I removed the phrase "font-family /**/:inherit;" after "{IPA fonts}};" in this template, to make it more similar to template:IPAchar, and indeed it seems to work better now, as far as I can tell. I don't know what function the removed phrase should have. -- unsigned by 83.248.98.228 on 22 January 2006

The removed part of the template is supposed to force the font only for Internet Explorer since other browsers such as Firefox get the fonts right without the need for any template. — Hippietrail 17:35, 23 January 2006 (UTC)

Then is should be in wikibits.js, in the section for browser-specific tweaks. Having it where it is clogs the JS console in FireFox and Netscape (the CSS is so very bad, it ends up parsed as JS?) If we need browser sniffer code in Monobook.js (or something) then so be it, but having this here is not acceptable. I'll remove this from here soon. --Connel MacKenzie 14:57, 12 September 2006 (UTC)

The removal of " font-family /**/:inherit;" (by 83.248.98.228 on 22 January 2006) was an important change to accomodate a Mediawiki software change. (See subsection "Mediawiki software change" in the Grease pit for description.) It should not be reverted. --teb728 18:53, 8 June 2006 (UTC)

Forced links

Please don't make this template inflexible and less usable by adding a forced link that might just as well be added at the discretion of the individual editor. It needs to be clean so it can be used for in-prose text and such. Take look at how its equivalent at English Wikipedia is used and you'll see what I mean. It's intent is to display IPA properly in any situation. If you feel the need for standardized linkage to a pronunciation guide page, then handle this with a separate template or just good old raw links.

Peter Isotalo 07:32, 30 January 2006 (UTC)

A fair point; however, if you are going to make this change, it must be made for AHD and SAMPA templates too, and you will need to add the "IPA" to all of the pages that use this template. Are you prepared to do this? — Paul G 12:59, 30 January 2006 (UTC)

This one is used under the "pronunciation" headers in numerous entries, so it needs a link. Please use Template:IPAchar for in-text usage. Also I'd recommend having a link to the IPA chart. — Vildricianus 15:22, 30 January 2006 (UTC)

Linking to the over-burdened IPA article at English Wikipedia is of very little value to the average reader. The level of knowledge of phonetics is not nearly high enough for anyone to figure out what that tight mass of symbols means (assuming that their computer can display them properly), and I'm saying this as a very dedicated phonetics nerd, not merely a bewildered outsider. If you want to do it properly, you need to link to a dedicated pronunciation guide that has examples, preferably both written and spoken. This would include separate pages for separate languages as well. Don't you have instructional pages of this kind already?

I don't understand the objections about the AHD and SAMPA templates, though. The links they contain are just as fruitless since they only redirect the reader to articles about AHD and SAMPA respectively, not how to read it. There is very little intuitive info there that actually helps anyone to comprehend the transcription system. And, yes, I'm prepared to add all 50-some links myself. The templates are in vary sparse use, so I don't see this as much of a problem.

Karmosin 16:53, 30 January 2006 (UTC)

Ok, I just noticed that the link hade been changed to a Wiktionary-page. Tons better than linking to Wikipedia, but still in great need of improvement. The tables are a bit arbitrary and can be difficult to read. However, I still don't find it appropriate to use the more compact {{IPA}} just to add a link (that really should be standard with the article templates themselves) and force people to use the longer {{IPAchar}} for display purposes. It also conflicts with how the template is used on other wikis, including the Commons.

Peter Isotalo 17:00, 30 January 2006 (UTC)

I just noticed that there's a page called Wiktionary:Pronunciation key, which is much better suited for than the current link. Since I need the template for other languages and without the automatic link, I'm going to remove it and I recommend that some form of link is established to the pronunciation key page from the pronunciation template, not this one. There's no point in having both create links to separate pronunciation pages.

Peter ^Isotalo 21:58, 15 February 2006 (UTC)

Anyone actually reading my posts? The current link for this template is next to worthless and can only be used for one language. Please stop auto-reverting and discuss the issue. We need links to separate pronunciation pages for each language, not a link to a monster of a table with no possibility of explaining either dialectal differences, inserting sound files and which is almost impossible to read.

Peter ^Isotalo 01:13, 18 February 2006 (UTC)

In this Wiktionary, we use {{IPAchar}} for the purpose you have in mind. We changed {{IPA}} last year to its current form because that’s what we needed. Since then, it has been inserted in numerous pages where the typographic design that we wanted was "IPA:...", using specific fonts. As Paul already told you, if you want to change it back, then you will have to make identical changes to SAMPA and several other templates, and then you will have to find and edit all the thousands of places where we want it to display the way it does now. Your argument is simply too late; it was inserted into too many pages with expectation of the current display. —Stephen 10:55, 18 February 2006 (UTC)

How can it be too late to point out that linking to a page with a huge, unwieldy table without room for examples or comments is bad for any end user? Are you seriously not even going to consider that you should use the far better suited pronunciation key when you already have one? It's easy to link to and it will free up {{IPA}} for what it was intended for (displaying IPA properly, not putting up links) and you won't need to use an extra {{IPAchar}} at all.

The SAMPA template will become superfluous if you use the pronunciation key page instead (linked once per article), since it covers all three standards in use for English in a more self-evident manner and links to general information about both SAMPA and IPA for those who are interested in learning more. I count something like 2-3,000 articles using the IPA template and less than 500 that use the SAMPA template. That's a few days of work, but it's certainly better than sticking to your guns and letting some tens of thousands of articles use the current, much more awkward and unintuitive format.

I must also point out that linking individual transcription systems from each article isn't all that relevant or useful for a dictionary. The point is to instruct people how to pronounce words, not to inform about individual transcription standards. People can look this information up in the wikipedias or through the appendixes.

Peter ^Isotalo 12:38, 18 February 2006 (UTC)

You can change the link (as a matter of fact, we just did that) and you can change the fonts (no change is needed), but not the way the template displays on a page. It was inserted into all of those pages with the current display in mind.

The SAMPA template uses monospaced Courier and is not suitable for IPA.

As far as linking goes, I don’t care about that and I don’t know how often anyone uses it. It might be okay unlinked, but the display must stay as it is. —Stephen 13:15, 18 February 2006 (UTC)

I'm not particularly fond of unlinking it. Which page it links to is not my concern either for now, as long as it is one which clarifies in some way what the IPA characters stand for. I understand from your posts that you are keen to make that page (whichever it is) better; please do so. I agree that the tables in Wiktionary Appendix:IPA Examples are somewhat messy, but Wiktionary:Pronunciation is too incomplete. A blend of both into one page would be excellent. — Vildricianus 12:29, 19 February 2006 (UTC)

It's hardly incomplete, Vildricianus. It has all the relevant information on English that the table has plus SAMPA and AHD. A blend of both on one page would make it one truly messy pronunciation guide. The point is that trying to keep the pronunciation guides for all languages on one page is going to make them too big or too complicated. Just trying to edit those huge tables, for example, is quite difficult.
I am, by the way, rather puzzled by this "who cares if anyone reads it"-attitude. Why would you insist on keeping any particular layout or link at all if you don't even care whether it's useful or not? Are your trying to help readers or to satisfy consensus decisions? There are better ways of fixing your current pronunciation guide layout, which has a lot of redundancy. Here's an example:

Pronunciation

UK

(file)

• IPA: /ˈdɪkʃən(ə)ɹɪ/
• Template:X-SAMPA

US

(file)

• enPR: dĭk'shə-nĕr"ē
• IPA: /ˈdɪkʃənˌɛɹɪ/
• Template:X-SAMPA

The SAMPA- and AHD-links (not templates) could be ditched and the main link doesn't need to be the section header, and there are probably even better constellations, but what I don't see is why we need so much repetitive linkage and layout in the current format. It can all be summarized in just one link and one IPA-template that's more practical and universal in usage.

Peter ^Isotalo 17:58, 20 February 2006 (UTC)

Doesn't look too bad at all. One note: what I meant with "incomplete" is that it currently lists only the English system and just one link to Swedish. If we can work out similar pages for every language which are linked from this page, then in overall this is a good idea. — Vildricianus 18:38, 20 February 2006 (UTC)

groan, whine So much work... I'll get cracking on Japanese and Chinese (yes, it'll have to be Standard Mandarin). I'll ask for assistance with the other major languages in the Beer Parlour.

Peter ^Isotalo 23:38, 20 February 2006 (UTC)

See also

• Template:Unicode
• Template:IPAchar
• Template:ARchar
• Template:FAchar
• Template:RUchar
• Template:THchar
• Template:SAMPA
• Template:AHD

CSS guru

Could someone identify and fix the css error being generated by this template?

Expected ':' but found ''. Declaration dropped. Line 0.

Thanks in advance. --Connel MacKenzie 16:34, 25 August 2006 (UTC)

because "Font names containing … or whitespace should be quoted".

style="font-family: Arial Unicode MS,DejaVu Sans,Segoe UI,Lucida Grande,Charis SIL,Doulos SIL,Code2000,Gentium,Gentium Alternative,TITUS Cyberbit Basic,Lucida Sans Unicode;

It comes from a nested template used for the fonts. Yuck! Why isn't that in the stylesheet, with nothing but a class assigned to the span here?! Especially since the size will vary based on the actual font chosen.

—Długosz 06:57, 26 December 2006 (UTC)

WP link

Why does it link to IPA chart for English? It is not only used on English words Smiddle / TC@ 21:16, 23 January 2007 (UTC)

If there exists any IPA tables for other languages, one could add some (optional) parameter to specify that language, I believe. \Mike 14:13, 17 March 2007 (UTC)

I added one Smiddle / TC@ 08:51, 22 March 2007 (UTC)

Nice :) \Mike 08:52, 22 March 2007 (UTC)

The switch statement on the langcode could be problemetic if it grows too large and if there are many pronunciations on a page. That's not the case yet, but in the future this may need to be revised. Hopefully we'll have an extension to convert language code to language name by that time. DAVilla 10:58, 1 July 2007 (UTC)

Style

There should be no need for the style attribute here. The style hack is present in MediaWiki:Common.css; putting the style attribute here changes the font for all browsers, making it gratuitously look different, but it should only need to be different in IE6. I was about to make this edit myself but minutes before I could, it got protected. Hairy Dude 15:06, 26 June 2007 (UTC)

Yes, it got protected so you wouldn't send the job queue through the roof for a change that would be rapidly reverted. This was discussed on WT:GP, and everyone seems to like it because the default presentation in most browsers is nearly unreadable. It is needed in Firefox on Win* just as much. Robert Ullmann 15:13, 26 June 2007 (UTC)

Where exactly? I can't see any reference to readability of IPA on WT:GP. Hairy Dude 15:20, 26 June 2007 (UTC)

Wiktionary:Grease pit archive/2006/December. It is reasonable to use the CSS font list (if it is checked/updated), but the template has to keep the size 110% (because we don't want that to apply to every use of IPA automatically). Robert Ullmann 15:33, 26 June 2007 (UTC)

The given example is actually less readable with the template here (Firefox on Ubuntu Linux: ɛ̃ without, ɛ̃ with template) - the tilde looks like a macron. Hairy Dude 17:11, 26 June 2007 (UTC)

I use Firefox in WinXP, and it looks perfect for me, appearing in Arial Unicode MS. I suspect that it isn’t Firefox or Linux, but the fonts that you have (or don’t have). The IPA template uses these fonts: Gentium, GentiumAlt, Arial Unicode MS, Code2000, DejaVu Sans, Segoe UI, Lucida Grande, Charis SIL, Doulos SIL, TITUS Cyberbit Basic, Lucida Sans Unicode. —Stephen 17:28, 26 June 2007 (UTC)

Can't we just remove forced fonts and let users customize their User:YourUserNameHere/monobook.css? I set a style for my .IPA class, but it's being overridden by this template, making its .IPA class reference meaningless. And for {{IPAchar}} too. I use FireFox 2.0.0.9. - Gilgamesh 07:28, 14 November 2007 (UTC)

A lot of us (me included) don’t know how to edit Monobook.css, and we rely on the template to give us something we can see. You might discuss it with User:Robert Ullmann. Perhaps you just need to add another font to {{IPA fonts}}. User:Robert Ullmann should have some ideas. —Stephen 16:04, 14 November 2007 (UTC)

Okay, I have successfully edited my own User:Gilgamesh/monobook.css to override the fonts in this template. See my code to see how it can be done. - Gilgamesh 15:32, 15 November 2007 (UTC)

template:style/IPA T

Dear friends, some time ago I have created template:style/IPA T which was deleted, see

special:Log|type=delete&page={{URLENCODE:template:style/IPA}} .

I assume that the template template:IPA T is not an equivalent of yi:template:style/IPA because template:IPA can neither be used inside the

[[foo bar|{{IPA|bla bla}}]]

nor inside the

[foo url {{IPA|bar}}]

syntax.

example

{{IPA|vɛʁ ɛs legt zɪx mɪt ɑ hʊnt ʃlɔfn̩, ʃtet ɔjf mɪt fle.|lang=yi}}<br />
[http://www.omegawiki.org/index.php?title=DefinedMeaning:dog_%28966%29 {{IPA|vɛʁ ɛs legt zɪx mɪt ɑ hʊnt ʃlɔfn̩, ʃtet ɔjf mɪt fle.|lang=yi
}}]<br />

generates:‎
IPA^(key): vɛʁ ɛs legt zɪx mɪt ɑ hʊnt ʃlɔfn̩, ʃtet ɔjf mɪt fle.
IPA^(key): vɛʁ ɛs legt zɪx mɪt ɑ hʊnt ʃlɔfn̩, ʃtet ɔjf mɪt fle.

comments

As you can see the second line is broken. Please recreate the template by copying the actual version of yi:template:style/IPA. Thanks in advance! Best regards
‫·‏לערי ריינהארט‏·‏T‏·‏m‏:‏Th‏·‏T‏·‏email me‏·‏‬ 03:24, 20 December 2007 (UTC)

What is the reason for ever including an IPA template inside a wikilink? --EncycloPetey 03:41, 20 December 2007 (UTC)

What is the reason to be not allowed to include IPA characters inside a wikilink?

‫·‏לערי ריינהארט‏·‏T‏·‏m‏:‏Th‏·‏T‏·‏email me‏·‏‬ 03:56, 20 December 2007 (UTC)

Because it is needless clutter of the Template space. We do not need more useless templates. We are already clearing out many that have been abandoned. You seem to want the template solely for display on your user page, which is just about the only thing you've edited here besides the templates. --EncycloPetey 04:17, 20 December 2007 (UTC)

See a meaningfull example at yi:user:לערי ריינהארט/!#OmegaWiki. You will need to expand the « OmegaWiki » stuff. Just click on « װײַזן · vayzn · vɑyzn̩ ».

"We do not need more useless templates." I do not understand who is « we » . Is this « pluralis majestatis ^{pluralis majestatis} » ? Regards

‫·‏לערי ריינהארט‏·‏T‏·‏m‏:‏Th‏·‏T‏·‏email me‏·‏‬ 08:20, 20 December 2007 (UTC)

Template {{IPA}} is for Pronunciation sections. You want the script template {{IPAchar}}. And use 1= for the parameter if there is an evaluation problem. (the only reason the one on yi: works is that it uses IPA=)

[[foo| some text bar some other text]]

vɛʁ ɛs legt zɪx mɪt ɑ hʊnt ʃlɔfn̩, ʃtet ɔjf mɪt fle.

See? Please do not create more templates trying to fix things that are not problems. They will be deleted on sight. Robert Ullmann 09:07, 20 December 2007 (UTC)

Neither template:IPA nor another template from category:Script templates is an equivalent of the examples from m:wikt:yi:category:style templates. There only named parameters are used and additional parameters are passed to support bidirectional issues.

Another solution then deteling the template would be to use either a redirect (but this will break) or an inclusion where parameters are passed / initialized and renamed / anonymized.

Most of the names specified at template talk:IPAchar|oldid=3446496#See_also are redirects as well. Are you going to delete all these redirects? Regards

‫·‏לערי ריינהארט‏·‏T‏·‏m‏:‏Th‏·‏T‏·‏email me‏·‏‬ 18:25, 21 December 2007 (UTC)

the way the template language works, "1" is a perfectly good name for a positional parameter. We do not have bidi issues here with IPA, because the text is always LTR, as is IPA. So use 1= as shown; and just forget about the bidi support because it is inappropriate here. Got it? (oh, and you would do much better to just refer people to your user page on yi.wikt? Robert Ullmann 13:36, 22 December 2007 (UTC)

The redirects you refer to are the old pattern of script template naming, we now use ISO 15924 codes; they are used on a lot of pages; they will slowly be deprecated; and they are utterly irrelevant to this issue. And this has nothing to do with Template:IPA, which is a formatting template for Pronunciation sections, not a script template. Robert Ullmann 13:40, 22 December 2007 (UTC)

lang=

What does the lang parameter do? --Keene 17:09, 12 January 2008 (UTC)

It adds a link to the corresponding language's phonology article at Wikipedia, if one exists. Angr 20:33, 1 April 2008 (UTC)

Icelandic

Can someone with admin privileges add Icelandic to the language list (WP has w:Icelandic phonology). – Krun 02:02, 14 February 2008 (UTC)

Welsh

Please add a parameter allowing lang=cy to point to w:Welsh phonology. 85.178.10.92 14:17, 1 June 2008 (UTC)

I've done this, I think correctly. Note here if it hasn't worked. Thryduulf 13:15, 2 June 2008 (UTC)

It seems to work; thanks! Is there any particular reason this template is fully protected rather than semi-protected? Angr 13:42, 22 June 2008 (UTC)

The relevant line from the log is "14:58, 26 June 2007 Robert Ullmann (Talk | contribs | block) changed protection level for "Template:IPA" ‎ (UI, used in very large number of pages [edit=sysop:move=sysop])"

"UI" is user interface, but I don't know what a "user interface template" means or why (or how) they are special (if they are). "very large number of pages" means 51,225 transclisions (excluding any through redirects) as of 23:15, 22 June 2008 (see Special:MostLinkedTemplates). See MediaWiki:Manual:Job queue for how changes to templates are processed and why changes to heavily transcluded pages should be minimised.

In context, {{IPA}} is the 22nd most transcluded template, {{isValidPageName}} is top with 280,523 transclusions, followed by {{form of}} ‎with 206,492 (w:template:! on en.Wikipedia has a massive 2,749,441 inclusions!). Thryduulf 00:20, 24 June 2008 (UTC)

WP link moved

The Wikipedia chart is now at w:IPA chart for English dialects. —Michael Z. 2008-08-07 19:44 z

Navajo

Navajo phonology is missing. Maro 20:43, 11 September 2008 (UTC)

I've added it so it should now work. Thryduulf 11:13, 13 September 2008 (UTC)

Greek

There has been a work here about Greek Pronunciation (see Wiktionary:About Greek/Pronunciation). I'd prefer that this template have a link to this page rather than a Wikipedia article. --flyax 08:32, 10 October 2008 (UTC)

Taos

You should add

twf=Taos

to the list. Ishwar 04:52, 2 January 2009 (UTC)

Done. Nadando 04:58, 2 January 2009 (UTC)

Lithuanian

The English wikipedia doesn't have a full article on Lithuanian phonology, but could lt= be added to the list to link at least to w:Lithuanian alphabet#Phonology? — [ ric ] opiaterein — 15:41, 13 March 2009 (UTC)

Phonology links

Why are we sending our readers to Wikipedia articles with this template? Why not keep the pronunciation guides within Wiktionary instead? When it comes to languages other than English, readers are referred to articles on phonology that are supposed to serve as explanatory articles on linguistics, not pronunciation guides.

Peter ^Isotalo 06:58, 24 March 2009 (UTC)

Good point. Even at Wikipedia, some languages have IPA guides that would be more helpful than the phonology articles (e.g. w:WP:IPA for Irish would be a better place to send people than w:Irish phonology). Better yet, we could copy w:WP:IPA for Irish here and link to it intra-Wiktionarily. Angr 11:57, 24 March 2009 (UTC)

I aggree. See above (#Greek) --flyax 10:00, 11 April 2009 (UTC)

We can use #switch to send people to our own or WP's pages based on language with the current link as default. Not sure how much server load that will generate, though. Just checked: we already do use #switch. This is easy, then. I'm for it.—msh210℠ 17:05, 26 March 2009 (UTC)

Since there seems to be no opposition to this suggestion, could we start by switching the Wikipedia links to Latin, Swedish and Spanish to our own pronunciation guides?

Peter ^Isotalo 11:28, 6 April 2009 (UTC)

English dialects

Hi. Can we change the English Wikipedia link to "IPA chart for English" as it's now been moved to w:IPA chart for English dialects - thanks! - Alison ^❤ 05:41, 28 April 2009 (UTC)

Change link

I'm sure we've discussed this before, and I don't see why it hasn't happened yet. Would an admin please change the link to w:IPA chart for English so it points to Wiktionary:English pronunciation key instead? Thanks! —Angr 16:14, 14 September 2009 (UTC)

I think it will be better to revert to the previous version. Currently, for languages with no pronunciation guide, link redirects to non-existent page on Wikipedia. See зова for example. Maro 21:06, 14 September 2009 (UTC)

The current version is much better. For cases like зова all we need is to make redirects at Wikipedia. —Angr 21:53, 14 September 2009 (UTC)

I suggest changing it to sth like this:

<pseudocode>

no "lang=" parameter or "lang=en" – [[Wiktionary:English pronunciation key]]
 #if:{{{lang|}}}:
 {{#switch:
 |el|no|gd|sv|cy| = [[Appendix:{{language|{{{lang}}}}} pronunciation]]
 |ab|ang|ar|arn|bn|ca|cau|cs|da|de|es|fa|fi|fr|ga|haw|he|hu|is|it|iu|ja|ko|la|nl|no|nv|oj|pl|pt|ro|ru|so|twf|uk|vi|yi=[[w:{{language|{{{lang}}}}} phonology]]
 |[[w:International Phonetic Alphabet]]
 }}

</pseudocode> Maro 20:44, 15 September 2009 (UTC)

The problem is that it would then have to be edited every time a new "XXX phonology" article is created at Wikipedia, unless the #ifexist switch has some way of looking into Wikipedia's space and seeing if the article exists there. —Angr 05:53, 16 September 2009 (UTC)

IPA isn't only used for English

The IPA: should't redirect to w:IPA chart for English, because IPA isn't only used to transcribe English. Instead, it should redirect to w:Wikipedia:IPA. --Mahmudmasri 04:52, 17 September 2010 (UTC)

It only links to w:IPA chart for English for English pronunciations. In, for example, Hebrew pronunciations, it links to w:Hebrew phonology (IPA^(key): ...). --Yair rand (talk) 05:13, 17 September 2010 (UTC)

What about using it here Appendix talk:Egyptian Arabic Swadesh list#Broad Transcription?? It should link to w:Wikipedia:IPA in that case. --Mahmudmasri 09:06, 17 September 2010 (UTC)

In the main namespace there's a bot that will add the correct language if none is specified. For appendices, you have to add it yourself using lang=. Mglovesfun (talk) 09:09, 17 September 2010 (UTC)

Examples? I don't get you. --Mahmudmasri 09:20, 17 September 2010 (UTC)

• I tried that {{IPA|lang=arz|/r/}}: IPA^(key): /r/, but, Egyptian Arabic doesn't have its own page on Wikipedia. --Mahmudmasri 09:25, 17 September 2010 (UTC)
• Also in the main page (Appendix:Egyptian Arabic Swadesh list), it redirects to w:IPA chart for English. --Mahmudmasri 09:27, 17 September 2010 (UTC)

Fixed, thanks. (See [1], [2].) —Ruakh_TALK 12:23, 17 September 2010 (UTC)

Plural?

Would it be possible to add an optional parameter for the pronunciation of the plural version of the word? I ask because words in Aramaic often have the same basic spelling in the singular and plural (so having two separate articles is impossible), but their pronunciations differ. Adding a couple of different subheaders underneath just to have "Plural form of X" would be tiresome. The German Wiktionary regularly has the pronunciation of both the singular and plural forms (even when the plural form is spelled differently and has its own article), see here as an example. Preferably, it would look something like that (the IPA entry with the singular, then a comma followed by "plural:" and the IPA entry for the plural). Also, it might be a good idea to add multiple parameters for different plural pronunciations. --334a 02:24, 20 October 2010 (UTC)

Take a look at entries like amaverimus and palma, where there are multiple Pronunciation sections. Would this accomplish what you need? Such issues are common in Latin, where there are identically spelled forms of a given word that are pronounced differently. --EncycloPetey 02:31, 20 October 2010 (UTC)

Hmm, good point. That Latin examples also reminded me of other non-plural forms of Aramaic words that are identical in spelling but different in pronunciation, and you would have to solve those problems using another method from the one I explained above. Alrighty then, I guess I'll just have to overcome my laziness and do it the established way. :) --334a 16:26, 23 October 2010 (UTC)

Just add the IPA to the plural entry? Mglovesfun (talk) 16:32, 23 October 2010 (UTC)

link within site if lang=en

Could the template link to Wiktionary:IPA pronunciation key (especially instead of w:English phonology when lang=en)? Perhaps it should also be more prominent that clicking it will lead to a pronunciation guide. — lexicógrafa | háblame — 14:34, 18 October 2011 (UTC)

w:template:IPAc-en

I'm just wondering, why is this template, which is (in my opinion) the most descriptive template for pronunciation, not on this wiki? I was just reading ruthless, wondering if Google Translate Speech had its pronunciation correct, and I had to look up the correct pronunciation that matched u: in a quite large table with a lot of weird characters, not knowing if the : was part of the preceding or succeeding character. I think it would be a very useful thing to have that template where you can just hover over the letter.Joeytje50 (talk) 20:15, 1 October 2012 (UTC)

HTML lang attribute

If there is a language set, then the lang attribute should appear in the HTML, with the appropriate variant subtag for IPA. For example:

<span lang="en-fonipa"> . . . </span>

If there is no language set, then either the empty string or the code und (undetermined language) indicates this, but the code is necessary to add the subtag:

<span lang="und-fonipa"> . . . </span>

Can someone familiar with the template add this?

References:

• W3C, Language tags in HTML and XML
• W3C, Tagging Text with No Language
• IANA, Language Subtag Registry

 —Michael Z. 2013-02-09 18:55 z

Oppose until we resolve the already-known issues with this. —Ruakh_TALK 19:35, 9 February 2013 (UTC)

Spacing

There should be a space after the colon, so it’s “IPA: […]”, and not “IPA:[…]”. 〜britannic124 (talk) 19:25, 9 February 2015 (UTC)

I brought this up a few days ago at the Grease Pit, but no one's responded so far. —Aɴɢʀ (talk) 20:17, 9 February 2015 (UTC)

|ni= params

@CodeCat Unless I am crazy (which is a distinct possibility), it does not seem like this method works correctly:

* IPA(key): /hɪˈdʒɑːb/, /hɪˈdʒæb/
===References===

Thoughts? —JohnC5 07:55, 3 June 2015 (UTC)

Why, what's wrong with it? —CodeCat 12:30, 3 June 2015 (UTC)

It generates the literal text "<ref>" and "</ref>" instead of actual ref tags. —Aɴɢʀ (talk) 13:43, 3 June 2015 (UTC)

That's strange, it should just leave those. —CodeCat 14:21, 3 June 2015 (UTC)

How do we fix this problem? 20:18, 3 June 2015 (UTC)

I think I see it now. The ref tags are inserted by the module, they're not being given as part of the parameters. As far as I know, Lua doesn't process HTML tags in the output, just like it passes { } through without processing. I'm not sure how to get around it though, I'll have to see. —CodeCat 21:20, 3 June 2015 (UTC)

What if, as I have shown above, you just make sure users add the <ref> tag in the param? I know it is not super elegant, but it may be the best you can get. —JohnC5 21:28, 3 June 2015 (UTC)

That's just a workaround and it doesn't solve the real problem. —CodeCat 22:22, 3 June 2015 (UTC)

It at least allows you to cite each transcription specifically but it does not obviate the need for <ref> tags. Otherwise I have no idea how to fix this unless the preprocessor treats {{#tag:ref}} differently than <ref>. —JohnC5 22:29, 3 June 2015 (UTC)

It's fixed, see my latest change to Module:IPA. But I have no idea why it has to be done that way, it's not exactly very intuitive... —CodeCat 00:09, 4 June 2015 (UTC)

Yeah, that's weird. Glad it's fixed though. 00:13, 4 June 2015 (UTC)

Apparently, just like templates, we can't use wiki tags like that (of course we can use HTML tags). Did you comment that part in your last change by mistake? The purpose of it looks to be adding a cleanup category to entries that have used ref tag in input. --Z 06:21, 4 June 2015 (UTC)

Nocat parameter?

For pages such as Zuch. DTLHS (talk) 23:40, 17 October 2015 (UTC)

This template should only be used to list pronunciations in bullets in pronunciation sections. In running text, use {{IPAchar}}. —CodeCat 23:41, 17 October 2015 (UTC)

Lang or 1st parameter

@Erutuon Many of our existing templates now accept the language code as either lang= or as the first parameter. Can this feature be added to {{IPA}} as well? See {{compound}} and Module:compound/templates for how it might be implemented. —Rua (mew) 17:31, 21 September 2017 (UTC)

@CodeCat: That could certainly be done, but a discussion should be started in the Beer Parlour first. I would rather not do that myself at the moment. I mean, both options would be supported, but there might be practical issues that need to be discussed, and people should be made aware. — Eru·tuon 00:53, 22 September 2017 (UTC)

Dzongkha

So, I've noticed that entries with the parameter lang=dz don't link to the existing Appendix:Dzongkha pronunciation, but rather to the nonexistent w:Dzongkha phonology. Is there any reason in particular why this is happening, or is it just that it takes a bit to update itself, and could someone please fix it if it happens to be the former? Thanks! -/ut͡ʃxʎørnɛja ☭/ ^{(탁ᷞ, кон-, ឯឌឹត្ស, 𐎛𐎓𐎄𐎛𐎚𐎒)}. 01:03, 3 June 2019 (UTC).

@Учхљёная: Just add the language code for Dzongkha to the langcodes table in Module:IPA/data to tell Module:IPA that it has a pronunciation appendix on Wiktionary. (I might try making this automatic though.) — Eru·tuon 03:44, 3 June 2019 (UTC)

@Erutuon: Ah, thanks! -/ut͡ʃxʎørnɛja ☭/ ^{(탁ᷞ, кон-, ឯឌឹត្ស, 𐎛𐎓𐎄𐎛𐎚𐎒)}. 12:03, 3 June 2019 (UTC).

@Учхљёная: Sorry about deleting Dzongkha. I guess the Dzongkha page had been added after the last dump came out so my search didn't find it. — Eru·tuon 18:59, 8 June 2019 (UTC)

@Ertutuon: Nah, it's fine. I only created it a few days ago and don't know how to add it to the dump, I guess. -/ut͡ʃxʎørnɛja ☭/ ^{(탁ᷞ, кон-, ឯឌឹត្ស, 𐎛𐎓𐎄𐎛𐎚𐎒)}. 19:03, 8 June 2019 (UTC).

@Учхљёная: Don't worry, there's nothing you need to do. Everything on Wiktionary is automatically included. It's just that I was working from the May 20th dump, generated before you'd created Appendix:Dzongkha pronunciation; now that the all-titles file for the June 1st dump has come out (it takes a few days for the dump to be generated), the page is listed in it. — Eru·tuon 22:58, 8 June 2019 (UTC)

add <ƞ>?

<ƞ> was removed from the IPA because it didn't have a specific phonetic value, but was only used as a phonemic symbol. But it's quite handy for languages where a nasal is homorganic. E.g. the Sanskrit prefix /səƞ/ 'half', which is either realized with a nasal vowel as [sə̃], or as homorganic [səm sən səɳ səŋ] depending on the following consonant. More economical to write it /səƞ/. kwami (talk) 23:38, 20 August 2019 (UTC)

@Kwamikagami: I have no objection to using ⟨ƞ⟩ for the placeless nasal, except that it's easy to confuse with ⟨ŋ⟩. Currently I'm using ⟨ɴ⟩ for the placeless nasal of Burmese (see Appendix:Burmese pronunciation), which has met with some complaints since the one thing it never is is uvular. ⟨ƞ⟩ doesn't have an XSAMPA equivalent, but we could use n\, which doesn't exist in XSAMPA. —Mahāgaja · talk 09:44, 21 August 2019 (UTC)

Yeah, the similarity with ⟨ŋ⟩ can be a problem. ⟨ɴ⟩ could be even more of a problem, though, if people take it at face value. The similarity to ⟨ŋ⟩ isn't a problem with Burmese, is it? Otherwise I think ⟨N⟩ would probably be best, since that's the wildcard for any nasal. But of course this template won't accept that either. Anyway, I mostly just wanted ⟨ƞ⟩ available, not that I was particularly planning on using it. It might be worthwhile for Hadza, but I've just been using ⟨n⟩. kwami (talk) 10:31, 21 August 2019 (UTC)

{{my-IPA}} doesn't mark syllable boundaries, so there is possible confusion between ⟨àƞà⟩ = [àɴ.à] and ⟨àŋà⟩ = [à.ŋà], but we could always program it to mark syllable boundaries. Actually, come to think of it, the former would be written ⟨àƞʔà⟩ anyway, so it is still unambiguous. The template doesn't accept ⟨N⟩ at the moment, but it's easy enough to add it to the list of acceptable characters if there's consensus to do so. The closest things strict IPA allows for a placeless nasal are probably /ɰ̃/ and /ə̯̃/, but both of those would give me a headache if I had to use them in a phonemic transcription of Burmese or Sanskrit or Japanese. —Mahāgaja · talk 13:15, 21 August 2019 (UTC)

which non-IPA characters does the template accept?

I've put in older transcriptions and had them rejected because it's no longer valid IPA. But the template accepts a lot of things that never were IPA, such as digits. E.g., at same#Norwegian_Nynorsk, there's a superscript digit 2, and that doesn't trigger the error. Do we have a list of such non-IPA characters accepted by the template? kwami (talk) 19:40, 20 August 2019 (UTC)

Module:IPA/data/symbols DTLHS (talk) 19:42, 20 August 2019 (UTC)

Thanks.

I made a note in the documentation. Some are useful extensions, and may even be in the extIPA or VOQS. But there are a few I'm dubious about.

What are these used for? I don't think I've ever seen them.

𝆏 𝆑 ⁻

And the digits,

¹ ² ³ ⁴ ⁵

are undesirable because they're ambiguous, varying in value from language to language and author to author. They should probably be moved to the data["suggestions"] section. The problem is that e.g. ¹ and ² may be either high tone or low, ³ may be mid or high, etc. And sometimes they don't indicate tone at all, but the number of syllables that are accented. So unlike the IPA and extensions, they don't have any fixed meaning, which defeats the whole purpose of the IPA. kwami (talk) 20:19, 20 August 2019 (UTC)

We have three Muong entries that use other digits: dã and mỗng use ⁶, and mốch uses ⁸. I have no idea what they're intended to represent. I think we use ² in Norwegian because the actual realization of that tone varies from dialect to dialect, so using ² instead allows us a cross-dialectal transcription. +—Mahāgaja · talk 20:56, 20 August 2019‎ (UTC)

𝆏 and 𝆑 aren't used in {{IPA}} or {{IPAchar}} in the main, Reconstruction, or Appendix namespaces. Searching for occurrences in the Module and Template namespaces yields no results outside of this module. ⁻ is used between superscript numerals in sím-mih (for instance, {{IPA|lang=nan|/ɕim⁵³⁻⁴⁴ mĩʔ³²/}}), and in various Chinese pronunciation modules used by {{zh-pron}} (to find them, search module: insource:/⁻/).

Since 𝆏 𝆑 are musical symbols (MUSICAL SYMBOL PIANO, MUSICAL SYMBOL FORTE), I'm not sure what they'd be doing in IPA transcriptions. They first appeared in Module:IPA in this edit by ZxxZxxZ (before I moved them to Module:IPA/data/symbols). Perhaps ZxxZxxZ could explain? — Eru·tuon 21:47, 20 August 2019 (UTC)

@Erutuon: Yes, that would be good, and since they're not used, we should probably just remove them. They might've been intended for their VOQS values (that is, as musical notation), but AFAIK such usage is not superscripted. kwami (talk) 22:50, 20 August 2019 (UTC)

Hm. The Muong examples with '6' look like they're supposed to be tone levels, which only go from 1 to 5. I thought maybe Muong was like some of the Omotic languages, which have been claimed to have 6 tone levels (and therefore cannot be transcribed accurately in the IPA), but 4 levels seem to be sufficient for Muong, so there's no reason to use '6'. BTW, I see the error message 'invalid IPA characters (⁶)' in page preview, but not in the regular display. kwami (talk) 04:55, 21 August 2019 (UTC)

Mahagaja, yes, that's what the Scandinavian tone numbers are for. The ⁶ and ⁸ should be similar -- at least in other languages of the region, they'd be tone numbers 6 and 8. I don't have much problem with such usage. It doesn't tell you how to pronounce the words, but it does identify the tones in non-standardized languages in a way that the actual phonetic values would not, like using "UR" for the NURSE vowel in English. But since they're not pronunciations, personally I don't think they should be formatted as IPA or placed in the pronunciation section. Better in the etymology section, I'd think, or alternative transcription. For the pronunciation section, I think the solution would be the same as we use for English, e.g. the Oslo and Bergen (or wherever) pronunciations in actual IPA.

But IPA^(key): /ɕim⁵³⁻⁴⁴ mĩʔ³²/ is a problem. What is effectively a superscript en dash separates the underlying tone from the phonemic tone in that compound word. (That is, on its own the first element is IPA^(key): /ɕim⁵³/, but here it's IPA^(key): /ɕim⁴⁴mĩʔ³²/.) There are extended IPA tone letters to handle that -- they face the other direction, so the two are placed back-to-back. See Wikipedia:Tone_sandhi#Transcription for an example. In this case, it would be IPA^(key): /ɕim˥˧꜓mĩʔ˧˨/, or else IPA^(key): /ɕim˦mĩʔ˧˨/ with IPA^(key): /ɕim˥˧/ in the etymology section.

The problem with the digits is that different linguists follow different conventions. In Southeast Asia, 5 is high and 1 is low. Actually, often 55 is simple high and 11 low, with 5 and 1 meaning that there's a plosive at the end, or doubled as 55 and 11 to avoid confusion with tone numbers 5 and 1, which would be something else entirely (like those Muong tones 6 and 8). In West Africa, 1 is high and 5 is low. In some Mesoamerican languages I've seen, 3 is high and 1 is low. But high might be 2 or 4 depending on the number of tone registers in the language, and in Africa 2, 3 or 4 may be low. So all you can really tell is that sequential numbers go in a certain direction, but you don't know which direction that is, or what the gamut is (that is, if 1 is high/low, then whether low/high is 2, 3, 4 or 5). So the numbers are essentially useless to anyone who doesn't know the language and the convention, which is no better than using Webster or Random House phonetic notation for English.

There are other non-IPA phonetic symbols used in SE Asia, that trigger an error and have a suggested replacement. I suggest that, if we wish the IPA to actually be IPA as we claim, that we should do the same with the digits, with the suggestions following the SE Asian convention (55 high and 11 low). Since these are being used as phonetic symbols, we don't have the pan-dialectal issue we would with Scandinavian 1 and 2. A bot could replace the existing examples, language by language, and then the digits could be moved to the suggested-replacement section. Otherwise we're setting an example for people to use other common non-IPA conventions like IPA^(key): /stīl/ for English style and IPA^(key): /bŭt/ for but.

Meanwhile I think we should add ⟨꜒ ꜓ ꜔ ꜕ ꜖⟩ to the list of accepted characters, so people have those IPA extensions as an option. (Hm. When I first used <꜓> in the example, it triggered a non-valid character warning. Now it isn't. Have they just been added?) kwami (talk) 22:48, 20 August 2019 (UTC)

The tone letters have been in the list of accepted characters, and the list hasn't changed for quite a while; I'm not sure how a warning was generated for one of them. — Eru·tuon 00:14, 21 August 2019 (UTC)

⟨꜒ ꜓ ꜔ ꜕ ꜖⟩ are not included at Module:IPA/data/symbols. kwami (talk) 00:45, 21 August 2019 (UTC)

Oh, I see, I didn't notice that ꜒꜓꜔꜕꜖ are facing a different direction from ˥˦˧˨˩, the letters in the module. I've added the "left-stem" tone letters. — Eru·tuon 01:46, 21 August 2019 (UTC)

Thanks. Yeah, the 'left-stem' letters are used for sandhi effects. They aren't official IPA, but are an intuitive extension of it (like ᶑ, barred ɪ ʊ, etc.). I suppose the minus sign is intuitive as well, it's just not the convention I've seen. kwami (talk) 02:31, 21 August 2019 (UTC)

All of the Chinese modules have I think been changed over to IPA tone letters now. I'll look around for what else uses the digits. kwami (talk) 04:25, 21 August 2019 (UTC)

@Kwamikagami: I'm wondering if we could be using ⟨꜈ ꜉ ꜊ ꜋ ꜌⟩ for neutral tone (qingsheng). It's again not official IPA, but it would be useful to distinguish the neutral tone from other tones. (For Hokkien, it might even be useful to use ⟨꜍ ꜎ ꜏ ꜏ ꜑⟩.) — justin(r)leung _{{ (t...) | c=› }} 22:36, 21 August 2019 (UTC)

@Justinrleung: @Erutuon: Yes, those could be quite handy. We'd only want to use them for phonetic transcription, of course. We only have minimal coverage at WP-en, so we'd want to expand that in the main articles on tone so people don't have to hunt to figure out what they mean. Meanwhile, I agree that support should be added here.

Justin, I don't suppose you have any idea what Muong tones '46' and '8' are supposed to be (above)? Muong only has 5 or 6 tones, depending on dialect, so it doesn't make sense for them to be tone numbers, and 4 pitch levels is enough to describe the contours, so it doesn't make sense to use an expanded 1-to-6 musical scale. kwami (talk) 23:08, 21 August 2019 (UTC)

Added ꜈꜉꜊꜋꜌꜍꜎꜏꜐꜑ because I figure it doesn't hurt. They don't seem to be used anywhere yet. — Eru·tuon 23:42, 21 August 2019 (UTC)

non-IPA or Sinological notation?

(thread break)

@Kwamikagami, Erutuon: The recent changes to the Chinese modules have been reverted by @Wyang. — justin(r)leung _{{ (t...) | c=› }} 05:25, 25 August 2019 (UTC)

The number of Chinese books using ꜖: < 10 (generous estimate)

The number of Chinese books using "fake IPA": at least thousands

These changes are not an improvement. Ill-advised, alien, confusing, and ignore consensus practice. Wyang (talk) 05:36, 25 August 2019 (UTC)

@Wyang: Sorry, but the IPA is consensus practice. If you don't know the IPA, then you should learn it. If you wish to transcribe words in Sinological phonetic notation (with consonants ȡ ȶ ȵ ȴ , vowels ɿ ʅ ʮ ʯ ᴀ ᴇ etc., and using numbers for tones in a way that contradicts much of the rest of the world), that's fine, but then you can't call it the IPA, any more than it would be appropriate to do the same with Americanist or Uralicist notation. If you label something the *international* phonetic alphabet, it needs to be international. Everyone accustomed to regional conventions needs to give up some of those preferences for that to happen. I still get mixed up sometimes going back and forth with transcriptions that use <y> for [j], but that doesn't mean I get to use <y> and claim the result is IPA. kwami (talk) 05:54, 25 August 2019 (UTC)

Then change the note to Sinological PA. "IPA" as you described is not the consensus in Chinese linguistics. And notations such as ꜖ are fringe confusing notations at best. Changing ²¹³⁻¹³, ²¹³⁻²⁴, ²¹³⁻²¹ to ˨˩˧꜖꜔, ˨˩˧꜕꜓, ˨˩˧꜕꜖ is hardly helpful to readers. Wyang (talk) 06:00, 25 August 2019 (UTC)

@Wyang: IMO Wiktionary should not be biased towards Chinese linguistics, but based on internationally recognized standards like IPA. While using numbers to indicate tone is common among Chinese linguists, it is not the only acceptable way of indicating tone. Important publications like 現代漢語方言大詞典 and 方言 (journal) use the usual IPA notation for tone. We've already gotten rid of Sinologist notations like ȵ (→ n̠ʲ) and ᴇ (→ e̞), so it would be inconsistent. — justin(r)leung _{{ (t...) | c=› }} 06:14, 25 August 2019 (UTC)

What book uses this ꜖ notation? Wyang (talk) 06:17, 25 August 2019 (UTC)

Template:R:zhx-teo:Chaozhou Fangyan Cihui 1991 is one. —Suzukaze-c◇◇ 06:19, 25 August 2019 (UTC)

Only one probably. Wyang (talk) 06:21, 25 August 2019 (UTC)

If it's too confusing, then we can remove all sandhi transcription, and have only phonemic notation. We would then need to give two transcriptions -- the form in isolation and the form in context. That would be perfectly correct, if not as concise. But using non-IPA letters such as digits in an IPA template is quite confusing to anyone who doesn't read the same books you do. kwami (talk) 06:25, 25 August 2019 (UTC)

@Wyang: I don't know if you're misunderstanding what I said, but I said 現代漢語方言大詞典 and the journal 方言 use the ꜖ notation for sandhi/tone change. — justin(r)leung _{{ (t...) | c=› }} 06:34, 25 August 2019 (UTC)

2 then. Wyang (talk) 06:39, 25 August 2019 (UTC)

They're Chao tone letters! Could we get any more basic than that? Really, they're a quite standard -- and routine -- notation. This is like me criticizing you if you used British spelling for English because I don't recognize it and therefore insist that it's "wrong". kwami (talk) 06:46, 25 August 2019 (UTC)

@Wyang: 1 + 2 = 3. If you want more books, here: 漳平方言研究, 福清方言研究, 上海市区方言志, 南京方言志, 永春方言志, 湖南省志方言志, 海丰方言, 黎川方言研究, 洛阳方言研究, 长沙方言研究, 沅陵乡话研究 ... — justin(r)leung _{{ (t...) | c=› }} 06:52, 25 August 2019 (UTC)

Fine. And the books using "fake IPA"? Wyang (talk) 06:58, 25 August 2019 (UTC)

@Wyang: It's true that many other books use numbers instead of Chao letters, but that's besides the point. Wiktionary is not only for Sinologists / Chinese linguists. IPA is internationally recognized, so unless IPA can't represent something, we should be following its standards. — justin(r)leung _{{ (t...) | c=› }} 07:03, 25 August 2019 (UTC)

Bear in mind though, that neither is ꜖ an IPA letter. Which is why numbers were used in the first place -- due to IPA being unable to represent tone sandhi clearly. (Talk:五) Wyang (talk) 07:12, 25 August 2019 (UTC)

Actually, ꜖ is an IPA letter. It was accepted at the 1989 Kiel Convention. It may be unfamiliar because it (like the majority of contours) does not appear on the Chart, but it is official IPA. kwami (talk) 22:06, 14 June 2020 (UTC)

Indeed. Wyang, you're edit-warring with three other editors. That can get you blocked for disruption.

I have no problem with using Sinological notation, and if you'd changed the wording so that's what it said, then we could simply create parallel IPA transcriptions. (Though perhaps for discussion as to how many transcription systems we want.) But you didn't do that. Your edits do not produce an honest presentation for our readers. Even when you were reverted, you insisted on an untruthful presentation. kwami (talk) 06:22, 25 August 2019 (UTC)

If you have no problem with using Sinological notation -- then just change the wording. I will do that for you. Wyang (talk) 06:28, 25 August 2019 (UTC)

That's just me -- there are two others you're edit-warring against. Also, if you're going to do that, at least do a responsible job of it. Fix all the templates, and don't link them to the WP-en article on the IPA. You need to link to a key that explains what the symbols mean. That's the whole point in having a key! kwami (talk) 06:49, 25 August 2019 (UTC)

@Wyang: If you're going to call it "Sinological", then we'd have to change /n̠ʲ/ to /ȵ/, /e̞/ to /ᴇ/, /pʰ/ to /p‘/ for consistency. I would be against it. — justin(r)leung _{{ (t...) | c=› }} 06:59, 25 August 2019 (UTC)

Those are inconsequential equivalents. However, using tone letters is an aesthetic and frequency concern. Tone sandhi written as ˨˩˧꜖꜔, ˨˩˧꜕꜓, ˨˩˧꜕꜖ becomes less legible and understandable, and becomes a disservice to the readers. Wyang (talk) 07:12, 25 August 2019 (UTC)

So instead you opt for ambiguity? If sandhi notation is too difficult for your intended audience, then you could remove all sandhi notation from Wiktionary. I would advise consensus before such a change, however. But if you say that something is tone [3], what does that mean? High tone? Low tone? Mid tone? Because it could be any of those. You at least need a key to explain what those arbitrary digits mean. kwami (talk) 07:26, 25 August 2019 (UTC)

@Wyang: It is true that reverse Chao letters are not standard IPA either, but there is still that ambiguity problem. I believe I mentioned to you a while back that we should be considering splitting into phonological/lexical/citation and phonetic pronunciations. This would allow us to use Chao letters and would also solve another problem of misusing slashes. — justin(r)leung _{{ (t...) | c=› }} 07:33, 25 August 2019 (UTC)

If anyone would like to do that, then great. I just don't want to see ˨˩˧꜖꜔. Wyang (talk) 07:37, 25 August 2019 (UTC)

But that is standard notation, and has been in use for what, nearly a century now? Neither IDONTLIKEIT nor ignorance of common conventions is a legitimate argument. Digits should never be used unless they're defined, and they should never be misrepresented as IPA. And the IPA should be used for all entries, albeit alongside other conventions if they're well established. Intuitive extensions of the IPA are acceptable for things like morphophonemic distinctions that the IPA intentionally doesn't cover, but not notation contrary to the very concept of the IPA. Though I agree with Justin about the misuse of slashes -- we should have e.g. IPA^(key): //˨˩˧꜖꜔// for morphophonemic sandhi transcription. But there will occasionally need to be some extension of the IPA for phonemic notation, as the IPA has purged itself of anything (except for that silly Swedish letter) that's not phonetic. kwami (talk) 08:05, 25 August 2019 (UTC)

Where is ꜖꜔ defined as standard? Wyang (talk) 08:07, 25 August 2019 (UTC)

Justin should answer that. He knows the lit better than I do. But I'm willing to wager they've been around for longer than either of us has been alive. Although I'm not sure 'where' is the appropriate question. Which sources would you even accept, since you try to deny the ones that Justin does provide? kwami (talk) 08:11, 25 August 2019 (UTC)

Can you provide one standard at least? Or are you just going to quibble? Wyang (talk) 08:15, 25 August 2019 (UTC)

Sure. Here's SIL describing the tone systems they cover - [3] There's also the following English-language authors who use the same standard:

• Y.R. Chao. 1965, 1968. A Grammar of Spoken Chinese. Berkeley. (dot letters)
• Jin Sunde. 1986. Shanghai morphophonemics. Bloomington. (left-stem)
• Xiao-nan Susan Shen. 1989. The Prosody of Mandarin Chinese. Berkeley. (left-stem & left-stem dot)

Although, as you note, these have not been accepted by the IPA -- as the IPA only accepts phonetic transcription, and these are not phonetic -- they are the convention used for tone sandhi by authors who use the IPA. That is, the IPA only adopted part of the Chao system -- the phonetic part -- but people who use it may use the entire thing. They don't mix and match, not that I've seen. Though, if you know of authors who use the IPA-accepted part of Chao and some other convention for sandhi, I'd be quite interested in seeing it.

So, the question is, given that the international standard is IPA, and uses the basic Chao tone letters, what do we do for sandhi when transcribing in IPA?

I just came across this explanatory note in Huang & Hui (2012) New Horizons in Chinese Linguistics:

(Note that in Pike's notation used here, the numbers have inverse values to those of Chao's tone letters [sic], so that 1 = high, 4 = low, etc.)

You also say above that the other differences between IPA and sinological notation "are inconsequential equivalents." But then so is using music staves vs digits -- and music staves are much clearer and intuitive, the superior system, since they have international definitions and the digits do not. Much as IPA <ʰ> is superior to the sinological apostrophe because the latter can be misunderstood to mean 'ejective'. (In fact, I've seen claims that Armenian has ejectives because someone misread an apostrophe = aspiration transcription.) You also say "using tone letters is an aesthetic concern." But that's an opinion, not a factual argument. For me, they're an aesthetic eyesore. You know how pinyin looks like Morse code if you're used to hanzi? Digits look like Morse code if you're used to tone letters. The digits are much harder to read. Anyway, personal opinions about aesthetics are trumped by universal standards.

BTW, in Europe there's a Danish dialectological notation, a German dialectological notation and a Greek dialectological notation in addition to the Uralic Phonetic Alphabet. There are presumably others. It would be unreasonable to expect our readers to learn them all. People came up with an international phonetic alphabet for a reason.

kwami (talk) 18:32, 25 August 2019 (UTC)

All in all, as you have admitted, the symbols you used are nonstandard IPA symbols, as are tone contour numbers. As such, they are nonsuperior to tone contour numbers, which are the consensus practice in Chinese linguistics and the longstanding practice on Wiktionary. Not only are they nonstandard, they are also much less commonly used, making them inferior instead of just nonsuperior. Additionally, forward and reverse tone letters are also confusing and less legible graphically. I see no reason to switch from a more common, non-IPA system to another much less common, equally non-IPA system. Unless you are able to prove that they are standard IPA, or would like to remove the use of reverse letters without loss of information, I will not support the switch. Wyang (talk) 07:48, 26 August 2019 (UTC)

I'm siding with Wyang on this one. To implement a change with such a huge impact, please start a WT:Beer Parlour discussion first to seek opinion from other editors. Alternatively, consider adding the IPA notation next to the sinological notation without any loss of information.

While the change to IPA tone symbols was being implemented, I had trouble reading Min Nan. If sinological notation was used, I could compare the tone levels relative to that of Mandarin or Cantonese. There are various dialects within Min Nan itself and using sinological notation makes it easier to compare the various Min Nan readings. KevinUp (talk) 10:23, 26 August 2019 (UTC)

Okay, I'll start a Beer Parlour discussion when I get the chance, maybe add sinological alongside the IPA as well (for modules where there's any sandhi -- Wyang doesn't object to the basic IPA itself, so any reason not to use just straight IPA for those?). But as for comparison between dialects, that's just as easy with IPA as it is with digits, which are nothing but Chao's typewriter substitution for tone letters -- easier, really, since you can take them in at a glance. kwami (talk) 05:25, 27 August 2019 (UTC)

I made an error of fact above. Wyang has been saying that ⟨꜒ ꜓ ꜔ ꜕ ꜖⟩ are nonstandard IPA letters, and I'd assumed they were correct. But it turns out they are IPA: they were introduced in the 1989 Kiel Convention for broad transcription, and can be found in the Report. They are not on the chart, but there are other things that are not on the chart, including superscript letters and most contour tones. There is thus no reason not to use them for tone sandhi in a pure IPA transcription. kwami (talk) 21:32, 14 June 2020 (UTC)

g

why so fussy about the /g/ phoneme? It is a pain to have to copypaste just because a word has a /g/ in it ... IPA doesnt distinguish between two /g/'s, it just uses a particular letter shape for the one it has. —Soap— 01:17, 20 June 2020 (UTC)

You don't have to copypaste. You can use {{subst:x2i}}, {{subst:x2ipachar}}, and {{subst:x2IPA}} to convert any string of XSAMPA (all in easy-to-type ASCII characters) into proper IPA characters. That makes it easy to type not only ɡ but all the other IPA characters that aren't available on a keyboard (ɑ, ð, ə, ɛ, ɪ, ŋ, ɔ, ɹ, ʃ, θ, ʊ, ʌ, ʍ, ʒ etc.). —Mahāgaja · talk 07:35, 20 June 2020 (UTC)

Has there been any more work on this? I assume peope want the IPA template to stay as it is, but i find it very annoying. I tried using {{x2IPA}}, but it seems very fussy and I havent been able to see any pattern as to whether it will or will not convert a given symbol, even "g", into IPA. If anything its actually easier to just copy-paste from the error message than to put in the x2IPA template, preview, copy the resulting {{IPA}} template, paste it, preview, check if it worked, and then fix the problem by copypasting from the error message like I could have done in the f irst place. e.g. test it with {{x2IPA|/zu:g/|/zʊg/|/ʃu:g/}}. This is proper IPA except for the g on all three entries. —Soap— 02:23, 12 December 2022 (UTC)

@Soap: Actually, that isn't correct IPA, because you used the punctuation colon (:) instead of the IPA colon (ː). But the "x2" template series will fix that too. Don't forget to use subst: and the language code. {{x2IPA|/zu:g/|/zʊg/|/ʃu:g/}} may not work correctly, but {{subst:x2IPA|de|/zu:g/|/zʊg/|/ʃu:g/}} will. And once you're using {{subst:x2IPA}}, typing can be even easier because you can write {{subst:x2IPA|de|/zUg/}} with an easily accessible capital U rather than the IPA ʊ, which probably requires copy-pasting or clicking from the character selection box, just like IPA ɡ does. In fact, I don't understand why anyone would consider using IPA ɡ to be more of a pain than using ʊ or ɑ or ə. —Mahāgaja · talk 07:55, 12 December 2022 (UTC)

Okay, thank you. So I guess the x2IPA template doesnt convert the first parameter because it's supposed to be the language code. I had thought it would be better not to put the language code in the template at all, but I guess I just wasn't seeing something. As for why I keep bringing up g, I have a rich keyboard with lots of IPA symbols on it, mostly the ones used in official orthographies. So typing symbols like ʒ ə etc is no problem for me, but it would be a bit silly to have two g keys that look exactly the same when typed. And yes, likewise for the colon. Anyway, clearly if nobody else is complaining, the templates are fine the way they are, and I will probably just use the templates some of the time and copy-paste from the IPA chart other times, depending on how many symbols I need. Thank you, —Soap— 12:57, 12 December 2022 (UTC)

@Soap: If it's a situation where you don't want the "IPA^(key): " prefix, you can use {{subst:x2ipachar}}, which generates the {{IPAchar}} template. But generally we do want the prefix in entries, and it needs the language code to link to the right key. — Eru·tuon 14:23, 12 December 2022 (UTC)

If you just want the IPA characters without any other template, use {{subst:x2i}}. If you want the transcription enclosed within the {{IPAchar}} template, use {{subst:x2ipachar}} (with annoyingly differing capitalization). Those two don't require a language code. But {{subst:x2IPA}} puts the transcription in {{IPA}}, which does require the language code and outputs the "IPA^(key):" prefix. —Mahāgaja · talk 16:24, 12 December 2022 (UTC)

^(Key) for gsw

IPA^(key): // It currently links to Alemannic German (by a redirect) which isn't very helpful. The best page to link to is: Help:IPA/Alemannic_German Fytcha (talk) 02:22, 1 November 2021 (UTC)

I'm not sure how to fix that, but if you create Appendix:Alemannic German pronunciation, even if it is initially just a mirror or w:Help:IPA/Alemannic German, then we can link to that instead. —Mahāgaja · talk 07:47, 1 November 2021 (UTC)

Would a simple redirect to the correct Wikipedia page suffice for now? Fytcha (talk) 10:49, 1 November 2021 (UTC)

I don't think that's possible. For any language name Langname, the only things Module:IPA/data will link to are [[Appendix:Langname pronunciation]] and [[w:Langname phonology]]. You'd have to go to Wikipedia and redirect [[w:Alemannic German phonology]] to Help:IPA/Alemannic_German, but Wikipedia frowns on redirects across different namespaces. —Mahāgaja · talk 13:21, 1 November 2021 (UTC)

fenomeen

Sorry to come here twice in one week, but i just added the pronunciation to fenomeen because I felt it was important to indicate that the word has final stress. But it turned two of my vowels red, and they show as red even when I look at it on another device, so I know it isn't just something i set in CSS somewhere. Presumably this is a nonverbal error message warning editors and readers alike that the pronunciation must be wrong, but what is wrong with it? If it is saying that Dutch has no short /e/ or /o/, than it's at odds with our own pronunciation guide for Dutch, which indicates clearly that short /e/ and /o/ do exist. This is at Appendix:Dutch pronunciation.

I copied the pronunciation from the Dutch Wiktionary, except that they seem not to recognize long /e:/, whereas we do, as can be seen by the difference in pronunciation between their entry and ours for words like geen. If I make the third /e/ short as well, I see all three vowels turning red.

I will add pronunciations very sparingly from now on, but I still would like to know what's going on here. Thank you, —Soap— 23:26, 14 December 2022 (UTC)

The relevant code is at Module:IPA/data#L-275, where /eː/ and /oː/ are given as valid phonemes, but /e/ and /o/ are not. I can't comment as to whether this is accurate or not. 98.170.164.88 23:31, 14 December 2022 (UTC)

Even Dutch Wiktionary gives the pronunciation as /fenoˈmen/, and they ought to know. I'm adding /e o/ to the list of Dutch phonemes. —Mahāgaja · talk 07:27, 15 December 2022 (UTC)

@Mahagaja: That solves the immediate error, but there is still a broader discrepancy left to resolve, as other entries on nl.wiktionary use /e/ and /o/ where we use /eː/ and /oː/ (a few basic examples: hallo, geen, weet). Who is right? 98.170.164.88 19:57, 15 December 2022 (UTC)

For that, you'd need someone who knows more about Dutch phonology than I do. —Mahāgaja · talk 20:32, 15 December 2022 (UTC)

Similar in Wikipedia

w:Template:IPAc-en BoldLuis (talk) 21:31, 14 February 2023 (UTC)

Missing letters

Please add extIPA 𝼈 as a valid letter. If they're not already there, ꞎ 𝼅 𝼆 𝼄 ᶑ 𝼉 𝼊 should be included as well, since all occur in non-disordered language. kwami (talk) 08:56, 27 May 2023 (UTC)