Talk:International Phonetic Alphabet

See Wikipedia:Manual of Style (pronunciation) for the style guide on the English Wikipedia regarding the use of IPA symbols.

How with IPA numbers in composed symbols?

When using IPA-numbers, what is the IPA-standard or -preferred way of noting a composed symbol? Example: ʈ͡ʂ = voiceless retroflex affricate, can be:
1. ​105 136​
2. ​105 + 136​
3. ​105 (136)​
4. ​105 + (136)​
On the IPAchart that shows the IPA-numbers both 2. and and 3. are used.

The tie bar has no IPA number I understand; so the brackets are needed. Although, the same sound is described both with and without the tie, if I understand it well. -DePiep (talk) 17:36, 18 January 2011 (UTC)

Are you talking about Unicode values? rʨanaɢ (talk) 23:43, 18 January 2011 (UTC)

No, I mean the "IPA-numbers", as published by IPA. E.g. 105 = ʈ (Unicode nor decimal HTML-entity involved). -DePiep (talk) 01:39, 19 January 2011 (UTC)

new {{infobox IPA}}

{{Infobox IPA/sandbox}} Here is a proposed new infobox IPA. It should replace both old boxes. Here are testcases/examples: {{infobox IPA/testcases}}. -DePiep (talk) 02:44, 28 January 2011 (UTC)

Looks good. — kwami (talk) 08:16, 28 January 2011 (UTC)

I think the IPA symbol is in boldface , and I think it would look clearer in normal font weight. Dan ☺ 14:45, 30 January 2011 (UTC)

You're right, although it was not set to bold. Probably the 5em sizing did it. I have added demonstrations of normals in /testcases#Big symbol: bold or normal (turns out there are only two left, on my browser: normal and bold). Any preference? (Personally, I like the bold one) -DePiep (talk) 15:44, 30 January 2011 (UTC)

I prefer the normal (not bold) version - its what we are normally going to see. —Coroboy (talk) 11:39, 31 January 2011 (UTC)

Changed into normal as requested twice here. Please let me know if there are other strange effects. -DePiep (talk) 12:02, 31 January 2011 (UTC)

That looks better. The diacritics didn't position well on the the bold version. —Coroboy (talk) 12:07, 31 January 2011 (UTC)

Yes, I saw that earlier but didn't know the bolding did it. Two birds, one stone. -DePiep (talk) 12:10, 31 January 2011 (UTC)

Are the examples for semivowels ʏ̯ e̯ o̯ in an IPA template ({{IPA|ʏ̯ e̯ o̯}})? I see box-box-e-box-o-box. Dan ☺ 17:33, 31 January 2011 (UTC)

Indeed they are, and without the {{IPA}} template. Only the big symbol has the {{IPA}} template usage build in. Feel free to experiment (like using that template at input). Especially the two notes are mean to be free text, so I did not frame too much. -DePiep (talk) 17:50, 31 January 2011 (UTC)

Labiodental nasal

"The labiodental nasal [ɱ] is not known to exist as a phoneme in any language.[27]", but in the labiodental nasals article, it is said to be used in many language. Or it that not as a phoneme? Sounds strange.

That is probably the case (that it is there but not phonemic). That is the case, at least, for English. rʨanaɢ (talk) 14:47, 18 February 2011 (UTC)

Like as an allophone of [m]. Dan ☺ 14:49, 18 February 2011 (UTC)

See Labiodental nasal. — kwami (talk) 14:59, 18 February 2011 (UTC)

It is used in English in words like symphony.

Latin?

the article said the vowels a e i o u are based on their pronounciation in latin. but, in latin, e, i, o and u had diffrent sounds. short e was ɛ, short i was ɪ, short o was ɔ, and short u was ʊ. however, a e i o u do represent the pronounciation in Spanish. — Preceding unsigned comment added by Heelsum (talk • contribs) 15:51, 23 April 2011 (UTC)

Actually, to be honest, I think IPA is based most closely on French. I'll have to check and see if I can find a source for that, though. rʨanaɢ (talk) 15:56, 23 April 2011 (UTC)

IPA Lessons on the Computer (eLearning)

Sorry for asking here, I know it's not a forum, and the discussion here is only about improving the article, but I guess it will be more effective to ask here where people are more understanding, then asking in another forum that isn't connected to the subject at all.

I want to learn IPA (in my country the IPA isn't a part of the curriculum, I don't understand why), and I search in the Internet every keyword I could have think off and I haven't found anything.

I need a software that is - first of all, an offline version (downloadable), will demostrate everything, and teach it as it is teached at school, and divided to lessons and includes tasks and tests. Maybe a Flash file, that include everything.

Thanks in advance. Galzigler (talk) 13:33, 27 May 2011 (UTC)

You don't need any special software. Any introductory linguistics or phonetics textbook teaches IPA in detail. rʨanaɢ (talk) 15:08, 27 May 2011 (UTC)

Nomination for deletion of Template:IPAhelp2col

Template:IPAhelp2col has been nominated for deletion. You are invited to comment on the discussion at the template's entry on the Templates for discussion page. DePiep (talk) 09:26, 14 July 2011 (UTC)

Technical Anomaly With Consonant Table

I noticed some strange behavior when copying the table into LibreOffice Calc. While the subheadings ("Bilabial", "Labiodental", etc.) appear to be perfectly fine in the browser, they show up with a solid grey box under some letters in Calc. This isn't apparent anywhere else, but the text is "tainted" as long as it's copied around and not typed manually. I did a hex dump of the copied text and the typed text for the word "Bilabial". Here's the output:

?@?:~$ cat <<< "Bilabial" | xxd -c 256 -ps
42696c616269616c0a
?@?:~$ cat <<< "Bila​bial" | xxd -c 256 -ps
42696c61e2808b6269616c0a

Someone might want to look into this anomaly.

71.247.120.30 (talk) 21:26, 21 July 2011 (UTC)

In the row of this table labelled "manner", the character ZERO WIDTH SPACE (UTF8 E2 80 8B) appears where I have substituted exclamation points below:

↓ Manner   Bila!bial       Labio!dental    Den!tal Alve!olar       Post!alv.       Retro!flex      Pal!a!tal       Ve!lar  Uvu!lar Pha!ryn!geal    Epi!glot!tal    Glot!tal

The solid grey boxes probably indicate that the font that Calc is using doesn't support these characters. They are not in the source, but they are in the generated HTML. Can anyone shed some light on where they come from? —Coroboy (talk) 12:28, 22 July 2011 (UTC)

I remember putting soft hyphens (&shy;) in those locations long ago in a precursor of this table. These served to allow narrow screens. The current version does not scale well. On a narrow screen, part of the table is not visible. Previously, when narrowing the window, the header words would be hyphenated progressively at predefined locations, and the table cells compressed. −Woodstone (talk) 12:57, 22 July 2011 (UTC)

The replacement of soft hyphens with ZWSP marks has no effect on the wordwrapping/non-wordwrapping behaviour. The remark on narrowing seems unrelated to the OP.

re Coroboy: where do the grey boxes come from? -- probably because an editor reproduced the grey boxes in the original chart File:IPA chart 2005 png.svg . Or do I misunderstand your point? -DePiep (talk) 22:47, 22 July 2011 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘Back to the original question: is it a bug at all that the copied text does not end up well that way? Does Wikipedia suggest or promise anything into that? Afaik the rendered result is correct & legal HTML text. -DePiep (talk) 22:47, 22 July 2011 (UTC)

• I can't see how soft hyphens from a "long ago" version of the table could be producing ZWSPs now.
• I was referring to the grey boxes in the subheadings mentioned by 71.247.120.30, at the beginning of this section.

—Coroboy (talk) 00:34, 23 July 2011 (UTC)

I get it. Dunno about the receiving side of this copy-thing, so I'll leave it here (apart from that question: what can and should be expected from Wikipedia in this?)). -DePiep (talk) 07:35, 23 July 2011 (UTC)

I have only seen the gray box it in LibreOffice Calc so far. The text looks fine everywhere else. I never suspected any extra bytes hitching along until the gray boxes showed up. This might have broader implications for accessibility tools, and search engine parsing algorithms. I'm not familiar with Wikipedia's priorities though. (I'm the original poster. I didn't sign in when I added the section.) LCS (talk) 21:13, 23 July 2011 (UTC)

If you are observant, you can also catch yourself having to process an extra invisible character (pressinging the left or right arrows twice to move the cursor to the next visible character or backspace and delete twice to delete a character adjacent to the invisible character). LCS (talk) 21:19, 23 July 2011 (UTC)

Thank you, you clarified (I'm back again ...). May I repeat my main question: the Wikipage is true HTML source (& so like CSS). What expectation may a user have when copying (from screen)? What is the bug? -DePiep (talk) 21:24, 23 July 2011 (UTC)

It’s a matter of separating the content from the style. Though it’s a feature offered by Unicode, I find it more appropriate to encode formatting information like line-break opportunities separate from the text itself. Are there no such facilities provided by HTML or CSS to deal with this? LCS (talk) 21:40, 23 July 2011 (UTC)

I can't answer you, but I can describe the circumstances. WP:Wikitable (like {| ... |- ...|}) uses old HTML (simple table structure). Which is very good to get a table (that is: still content, but structured specilally). CSS (<span>...</span> and <div>...</div>) is CSS style, and produces very good rendering (showing). But they don't combine too well, while we needed that for the IPA chart. So every mix is a compromise.

Now to create the IPA Consonants table, we (active editors at the time) had to compromise with: a strong table (even 4 tables), and well shaped cells (background, border lining, IPA-letter font showing great, ...). We did with HTML/Wikitable code basic cells, and CSS makeup & letter positioning.

In the end: it is HTML code using CSS. If it ends up wrong in 3rd party code, what can we do? To me, it is not a bug. -DePiep (talk) 22:10, 23 July 2011 (UTC)

Actually, the current formatting prescribes fixed cell widths, causing scaling on narrow screens to be blocked. So actually I think it would make no difference if the ZWSPs were removed. However, on the contrary, in my opinion we should go back to a simpler scalable format. The current version is only optimised for wide screen and its styling is very deviant from standard wikitable styling. −Woodstone (talk) 14:55, 25 July 2011 (UTC)

On this scaling thing, I suggest you start a separate thread. Seems unrelated to the OP. -DePiep (talk) 18:08, 25 July 2011 (UTC)

So it looks like the ZWSP character produces the grey boxes (I cannot reproduce in OpenOffice). Afaik, adding ZWSP (or NBSP, a regular space, or RTL) is perfectly legal by the Unicode and HTML. Its rendering & handling is well defined. It is part of the textstream (content). The rendering programme should be able to cope with that (or at least not show unknown invisible characters). Since it is legal Unicode/HTML code, I prefer no further restrictions on usage.
Are we sure it is not a LibreOffice feature/bug (or the used font as Coroboy noted above)? -DePiep (talk) 18:33, 25 July 2011 (UTC)

Any IPA friend with admin rights?

I have posted a "edit protecte template", a request for our (my?) grand {{IPA soundbox}}} system. Is there any IPA-template-knowing admin that supports this editrequest? (I don't want to jump the que. I'm up 2/18 in line now ;-)). -DePiep (talk) 00:17, 27 July 2011 (UTC)

Upstep and downstep

I see this article uses ꜛ and ꜜ (U+A71B and U+A71C) for upstep and downstep.

I’m not absolutely sure but I remember the IPA Handbook (I’ve lost my copy of it) uses ↑ and ↓ (U+2191 and U+2193) instead.

Could some one check it? — Preceding unsigned comment added by Tohuvabohuo (talk • contribs) 08:53, 31 July 2011 (UTC)

Background

• U+A71B ꜛ modifier letter raised up arrow (HTML: ꜛ)
• U+A71C ꜜ modifier letter raised down arrow (HTML: ꜜ)

Both in block Template:Unicode chart Modifier Tone Letters

• U+2191 ↑ upwards arrow (HTML: ↑ ↑)
• U+2193 ↓ downwards arrow (HTML: ↓ ↓)

Both in block Template:Unicode chart Arrows. Still, Unicode does not define IPA symbols, IPA should. Doesn't the IPA chartsay it? File:IPA chart 2005 png.svg-DePiep (talk) 09:07, 31 July 2011 (UTC)

The IPA organisation does not publish Unicode-list. But [:http://www.phon.ucl.ac.uk/home/wells/ipa-unicode.htm this UCL site] does. It states, in table "Arrows",

symbol, dec, hex:

• ↓ 8595 2193 downstep
• ↑ 8593 2191 upstep

which are the Unicode-arrows. Still, the IPA chart clearly shows that the symbols are only half character heigth. -DePiep (talk) 09:25, 31 July 2011 (UTC)

The IPA handbook defines in its character table:

Symbol	Symbol name	Phonetic description	UCS code
↓	Down arrow	Downstep	2193
↑	Up arrow	Upstep	2191

I cannot find codes a71b, a71c in it. −Woodstone (talk) 09:41, 31 July 2011 (UTC)

Codes a71b, a71c did not exist when the Handbook was published. — kwami (talk) 14:05, 31 July 2011 (UTC)

Changed in Suprasegmentals table. -DePiep (talk) 11:54, 31 July 2011 (UTC)

AFAIK, they're supposed to be half-high. The chart is problematic, because when it was created SIL fonts did not support the arrows. Similarly, there have been font problems in the Handbook. (Can't locate my copy right now either.) I've never seen full-height arrows used for this in the lit.

Ah, it's available on Google Books. P. 15 illustrates both, and they are clearly half-height (superscript) arrows. It would seem to be a font problem or coding error.

Yeah, they seem to have manually superscripted it. Remember, the Handbook came out in 1999, which means they were dealing with Unicode 2.1. (3.0 came out in Sept. of that year.) The half arrows did not get Unicode support until 5.1, which came out in 2008, so there was no proper Unicode point to assign to them. The ones they gave would get people by, since they could always manually superscript as the Handbook did.

If you go to the chart on their website, the tone marks have what appear to be manually superscripted arrows as well. They don't line up with the global rise & fall arrows. All the other IPA characters line up with the explanatory text as well, but these two are set above it.

They are also given here as the equivalent of the 'up arrow' and 'down arrow' in Pullum's Phonetic Symbol Guide.

This UManitoba IPA chart has superscript arrows with the IPA numbers, but leaves the Unicode values blank. — kwami (talk) 15:19, 31 July 2011 (UTC)

Adding: global rise, global fall arrows

First four rows are from earlier discussed table. Added: two rows (Global arrows)

IPA Handbook							SIL	Conclusions
Symbol name	Symbol	Phonetic name	IPA number	Unicode	Unicode (PUA)	IPA Handbook	SIL	Unicode	Result
Down arrow	ꜜ	Downstep	517	2193	EEAF 	p. 15, 184	U+F19D  (PUA, deprecated) → U+A71C MODIFIER LETTER RAISED DOWN ARROW Unicode v5.1	U+A71C	Xꜜ
Up arrow	ꜛ	Upstep	518	2191	EEAD 	p. 15, 184	U+F19C  (PUA, deprecated) → U+A71B MODIFIER LETTER RAISED UP ARROW Unicode v5.1	U+A71B	Xꜛ
Down full arrow	↓	Ingressive air flow	661			p. 189		U+2193	X↓
Up full arrow	↑	Egressive air flow	662			p. 189		U+2191	X↑
North east arrow	↗	Global rise	510					U+2197	X↗
South east arrow	↘	Global fall	511					U+2198	X↘

Noise in audio representations

Rather than having just the sound alone, the audio files contain other unnecessary sounds. Could alternative audio files be uploaded without the noise? 173.218.137.22 (talk) 01:14, 1 August 2011 (UTC)

"Not an IPA symbol" - revisited

I'd like to improve our representation of "non-IPA symbols" in the charts. I recall something like this was discussed before, but I cannot find that one.

These are the ones:

Symbol	Phonetic name	IPA number	Considered IPA symbol	Proposed representation in charts	Note
ɽ͡r	retroflex trill	105 101 122	yes	ɽ͡r
я	epiglottal trill		no	я *
ɢ̆	uvular flap	112 505	yes	ɢ̆
ɺ̢	retroflex lateral flap		yes ? yes	ɺ̢	composed with unnamed elements
ᶑ	voiced retroflex implosive		yes ? yes	ᶑ	composed with unnamed elements
ɪ̈	near-close central unrounded vowel	319 415	yes	ɪ̈	(omitted similar ones)

The proposal is build up along this line:

1. Whether a symbol is actually present on the published chart or not, is not relevant. Adding a composed IPA symbol on the right spot is OK, and does not have to be marked as such (example: ɪ̈ and ɽ͡r).
2. If it can be constructed from identified IPA elements, it is an IPA symbol (example: ɢ̆=112 505).
3. If elements are used that are not in IPA at all, it is not an IPA symbol (example: я).
4. If the symbol is constructed with unidentified elements from IPA (not separately named), that is considered to be an IPA symbol, I suggest (Example: ɺ̢).

-DePiep (talk) 16:12, 21 August 2011 (UTC)

I've applied this to the (new) tables {{IPA chart pulmonic consonants}} & {{IPA chart non-pulmonic consonants}}. Only я * is marked non-IPA there. -DePiep (talk) 15:02, 2 September 2011 (UTC)

And in {{IPA consonant chart}}. -DePiep (talk) 08:53, 3 September 2011 (UTC)

I disagree. ɺ̢ is not an IPA symbol. The retroflex tail is not an independent element the way ejective ʼ is. Sure, it's extrapolated from the IPA, whereas я is not, but it's not supported by complete IPA fonts, for example. — kwami (talk) 03:49, 12 September 2011 (UTC)

Mention Android?

I wonder if it would be worth mentioning Android in the closing Technical Note. The Android Browser is an increasingly widely-used current browser, but the default font it comes with lacks the glyphs for the IPA, and so users just see boxes. I don't think there's a way for them to fix this with font downloads at the moment. There is a bug report related to this issue (http://code.google.com/p/android/issues/detail?id=3167 ) which users can "star" if they think it's important, which could encourage Google to act on it. Emertonom (talk) 23:16, 24 August 2011 (UTC)

Merge Lateral release (phonetics) and Lateral consonant?

I've put a Merge-tag on Lateral release (phonetics) and Lateral consonant. But hey, maybe I'm wrong. -DePiep (talk) 00:04, 12 September 2011 (UTC)

I don't see any reason not to. Same goes for nasal release. rʨanaɢ (talk) 00:11, 12 September 2011 (UTC)

(eh, you mean merge nasal release with ... ;-) )-DePiep (talk) 00:30, 12 September 2011 (UTC)

There is: nasal release, Nasal consonant, Nasal vowel and Nasalization (and Nasal (disambiguation). I do not know the diff, please help us out. -DePiep (talk) 00:39, 12 September 2011 (UTC)

About the nasal articles: moved to separate section below. -DePiep (talk) 01:41, 12 September 2011 (UTC)

Actually, now that I look closer, these are all different topics. Lateral release (phonetics) is not about lateral consonants themselves, it's about normal stop consonants that have a bit of "lateralization" tacked on the end, as secondary articulation. (Compare to consonants with palatalization, e.g. Russian [t^j]; these are not actually palatal consonants, they are consonsants with some other place of articulation that have extra palatalization as a secondary feature.) So I'm not sure if the best solution is to merge; a better solution might be to make this distinction clearer in the articles in question. rʨanaɢ (talk) 00:48, 12 September 2011 (UTC)

Merge nasal articles: Nasal release, Nasal consonant, Nasal vowel, Nasalization, Nasal (disambiguation)?

There is: Nasal release, Nasal consonant, Nasal vowel, Nasalization, Nasal (disambiguation). Should we merge? -DePiep (talk) 01:41, 12 September 2011 (UTC)

Probably not, for the same reasons I mentioned above. rʨanaɢ (talk) 02:06, 12 September 2011 (UTC)

Infobox IPA not in Breathy voice &tc?

And editor removed the Infobox IPA from Breathy voice, Strident vowel, Creaky voice and Voicelessness ([1] [2] [3] [4]). The es: remove pointless template. Is that correct? First, the es is too thin, not a motivation. I can imagine the IPA box should be below the phonetics box, all right, but still there. -DePiep (talk) 10:12, 29 November 2011 (UTC)

One-Letter One-Sound

If the IPA operates on a "one-letter one-sound" principle, why do the symbols ɪ and ʊ have two different sounds (stand-alone and diphthong)? Should this be mentioned as an exception in the article? Interchangeable|talk to me 17:29, 30 December 2011 (UTC)

As a learner, I was overwhelmed by the IPA chart displayed in this article and felt that it lacked the interactivity I needed to understand it. I would like to see a link to this interactive version of the same chart next to the plain printable/viewable chart:

http://www.yorku.ca/earmstro/ipa/

Obviously the interactive cannot replace the printable as people still want to print!

)

B — Preceding unsigned comment added by Bencn (talk • contribs) 01:05, 25 January 2012 (UTC)

Interchangeable, all vowels can appear as standalone sounds or as part of a diphthong. I'm not clear how this is an exception to the one-letter-one-sound principle. garik (talk) 01:41, 25 January 2012 (UTC)

Never mind. I thought that they represented one sound on their own but /i/ and /u/ in diphthongs. Interchangeable|talk to me 17:11, 27 January 2012 (UTC)

Ah, I see. No, that's not the case. garik (talk) 21:31, 27 January 2012 (UTC)

Though the diphthong /oʊ/ is in my dialect more like /ou/. Interchangeable|talk to me 19:36, 28 January 2012 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘Sure. There would be nothing wrong with transcribing your pronunciation of "coat" as [kout]. And you could add all sorts of information if you wanted a narrower transcription (such as indicating that the first consonant is aspirated and that the second is unreleased etc.). The idea that the coat-diphthong is pronounced /oʊ/ "in English" is a very broad generalisation, and to some extent a matter of convention. Since we're talking about a phonemic, as opposed to a phonetic, transcription, the precise details don't really matter. This is particularly the case if you bear in mind that the IPA employs only 30-something vowel symbols to mark out points on a continuous space—even for a very narrow transcription, you're just selecting the closest vowel symbol you can. Of course, it also happens to be the case that many people think the second part of the coat-diphthong is more like their /u/ monophthong, when in fact it's closer to their /ʊ/. This was the case for me when I was starting to learn phonetics. garik (talk) 20:42, 28 January 2012 (UTC)

Thanks for the help. Interchangeable|talk to me 00:55, 31 January 2012 (UTC)