Help talk:IPA/English/Archive 12

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 5

←

Archive 10

Foal in the GOAT vowels

Could we have some note or link to English-language_vowel_changes_before_historic_l#Wholly-holy_split if we're sticking foal in the GOAT vowels? Because pronouncing it with that vowel would be very marked to me (in south-east England), suggesting a very old-fashioned/conservative version of RP. I'm not sure how you'd usually do this, given that I don't think the actual vowel common for foal here is in the table, so maybe just a link? 86.162.71.235 (talk) 23:03, 8 April 2011 (UTC)

Many accents have weird allophones of some vowels before some consonants; that's the whole point of having examples covering different environments. (And this is also the reason why I don't like having a tooltip for each individual phoneme; having one for each onset and rhyme as in /'kwɑrk/ would be much better IMO.) OTOH, [ɒʊ] has become phonemic for some speakers, so it might be mentioned. ― A. di M._plé^dréachtaí 21:53, 10 April 2011 (UTC)

jɑː vowel sound

If we're going to list juː as a vowel sound then shouldn't jɑː also be listed as a vowel sound?
Thatotherperson (talk/contribs) 13:07, 26 April 2011 (UTC)

ju: can appear after another consonant, which jɑː cannot. Although I don't think we should list ju: separately myself. Grover cleveland (talk) 14:58, 26 April 2011 (UTC)

What we represent as /juː/ is traditionally and dialectally considered a diphthong. /jɑː/, on the other hand, isn't. — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 03:04, 27 April 2011 (UTC)

IPA spellings for English

Why is it ok on Wikipedia for IPA English to ignore certain things such as aspiration? Another thing, concerning English words with 'ia' at the end. These are often transcripted into IPA as /iə/ instead of /ijə/ (for example see Lithuania, Albania). This presents a confusion when IPA for another language appears side by side with English IPA. In some cases, the other language actually pronounces it like a diphthong. While, I'm not sure if any form of form of English dialect actually pronounces it like a real diphthong. I'm not really familiar with all of IPA, is there like a specific symbol to use for diphthong vowel sequences? --Dara (talk) 03:09, 11 May 2011 (UTC)

We ignore aspiration in English because it's never phonemic. Aspiration occurs automatically, so we don't need to tell people when to do it. We can use /p/ for both the p in pit and the p in spit because speakers "know" (subconsciously) when to use which. The same is true with the /j/ between the /i/ and the /ə/ in words like Lithuania and Albania: it occurs spontaneously, so we don't need to tell people to do it. In both cases, Wikipedia is not alone: if you look at any dictionary that gives pronunciations, they'll use the same /p/ symbol for both aspirated and unaspirated p, and they won't indicate /j/ in the middle of -ia. This page is just a pronunciation guide, not a thorough analysis of English phonology. —Angr (talk) 07:19, 11 May 2011 (UTC)

Actually, Merriam Webster's Learner's Dictionary does "indicate /j/ in the middle of -ia.": /ˌlɪθəˈweɪnijə/ (http://www.learnersdictionary.com/search/Lithuania). If anything, it makes transcriptions easier to read. --TheAmericanizator (talk) 13:52, 12 May 2011 (UTC)

Cool, I didn't know that. I'm also surprised the M-W Learner's Dictionary uses IPA. —Angr (talk) 13:56, 12 May 2011 (UTC)

That, to me, seems to kind of defeat the purpose of using IPA. It makes me ponder why bother transcripting English words into IPA if the the speaker is automatically assumed to know how English words are supposed to be pronounced? This puts IPA like it's a rendition of the English alphabet with it's spelling/pronunciation inconsistencies. Someone not too familiar with IPA might read the IPA transcription of another language based on the English IPA transcription (not realizing what /iə/ should actually sounds like in some cases). I hope the special omissions for English are noted somewhere in the IPA article. It also seems to totally ignore that there is a vowel difference in the word 'ear'. Shouldn't we at Wikipedia, as an encyclopedia that includes many other languages also, consider using IPA 'correctly' as to not cause confusion? --Dara (talk) 04:59, 14 May 2011 (UTC)

I don't see that using a broad rather than a narrow transcription "defeats the purpose of using IPA", nor do I see that we are using IPA "incorrectly" at present. There's some information people have to be told if they're going to pronounce a word correctly, e.g. what phonemes to use, where the stress falls, etc.; in short, everything that isn't predictable. Information that is predictable, such as aspirating voiceless stops at the beginning of a word or stressed syllable, inserting glides after high vowels and diphthongs in the usual places, etc., doesn't have to be mentioned explicitly – and would clutter up the transcriptions, making them harder to read and interpret – since English speakers apply these rules automatically without thinking about it. The correct place for discussing such phonetic details is English phonology. —Angr (talk) 06:24, 14 May 2011 (UTC)

Yes, we're using IPA here to let readers know how to pronounce words that might not have intuitive spellings. To do this, we only need to give a broad, phoneme-level transcription; whether the [j] present between vowels is phonemic or not depends on dialect and on analysis (the issue is the same with the pre-rhotic [ə]), though that there is no [i.V] [i.jV] contrast makes it seem like it's not phonemic IMHO. We're not trying to give readers phonetics lessons, which is what indicating aspiration etc. would do. — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 19:16, 14 May 2011 (UTC)

The vowel in song

Song is listed as an example of /ɒ/, which is used for lot, pod, and doll. That's strikingly incorrect for most Americans (and I don't know whether it's correct for anyone else). For most Americans, lot, pod, and doll have [ɑ], which is legitimately represented by /ɒ/, but song has /ɔː/. For me, in the northeastern US, the vowels in other words ending in -ng are unpredictable (I think): strong, wrong, long, belong have [ɔː], while throng, tong(s), gong, pong, bong, dong, Hong Kong have [ɑ]. For this reason, for purposes of illustrating the symbols in this article, it's best to avoid -ng words. I think strong is the only one, and it should be removed. Linguistatlunch (talk) 15:55, 11 March 2011 (UTC)

Hmmm, it may be that the merging of /ɒ/ and /ɑː/ (mentioned at Phonological history of English low back vowels) by most Americans is not as clear-cut, somewhat like English-language vowel changes before historic r#Historic "short o" before intervocalic r describes. It could be that the cot-caught merger is spreading. — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 16:55, 11 March 2011 (UTC)

Song is generally considered a CLOTH word, i.e. it has the same phoneme as lot in present-day mainstream standard southern British English and the same phoneme as thought in in present-day mainstream standard midwestern American English. --A. di M. (talk) 18:40, 11 March 2011 (UTC)

No it wouldn't be "strikingly incorrect" for most Americans. According to Labov's latest NPR inteview, about half of Americans are caught-cot merged. However, since before /ŋ/ and /ɫ/ the /ɑ/ is often rounded, being closer to /ɒ/, I doubt it would be noticed by even most non-merged Americans.

Furthermore, many of the words in your list that supposedly have /ɑ/ can have /ɒ:/ (a.k.a. what you wrongly write as /ɔː/) too. --TheAmericanizator (talk) 16:46, 14 March 2011 (UTC)

It's not “wrong” to write /ɔː/ for the THOUGHT vowel in a phonemic transcription, even less in a diaphonemic one. (For young people in southern England that vowel can be as close as [oː], and anyway if the point was to be phonetically accurate we'd have to use e.g. ʉː rather than uː. --A. di M. (talk) 16:56, 14 March 2011 (UTC)

Before a velar, there is wide variety between LOT and CLOTH in non-cot-caught-merged AmE accents. Probably better to avoid them in giving examples of LOT words. Grover cleveland (talk) 15:00, 26 April 2011 (UTC)

According to MW's learners dictionary SONG is pronounced as /ˈsɑ:ŋ/ http://www.learnersdictionary.com/search/song — Preceding unsigned comment added by 93.136.150.142 (talk) 21:12, 29 May 2011 (UTC)

Yep. And if you listen the audio associated with that, it sounds like [sɔŋ], not like [sɑːŋ]. —Angr (talk) 21:18, 29 May 2011 (UTC)

Long 'a' in 'bade'

I like the hover-over popups that we now have on (many) IPA spellings to provide real-world examples of how those glyphs are supposed to be pronounced. (I previously complained here about it being too hard for the average person to find a simple key for IPA pronunciation.)

But I notice that the popup for /eɪ/ says, "long 'a' in 'bade'." As far as I know, most of the world -- including myself, an American -- pronounces "bade" with a short a, exactly like "bad"; the long-vowel pronunciation is considered nonstandard. Dictionary.com shows the IPA as /bæd/. Can somebody who knows where the relevant sourcetext for these popups is located please correct this? I'd suggest "long 'a' in 'face'," in keeping with the first example given for the glyph on this project page. Glenfarclas (talk) 18:46, 9 May 2011 (UTC)

I agree; bade is traditionally pronounced /bæd/ if it's uttered at all, but it's such a rare word that /beɪd/ as a spelling pronunciation is not unheard-of. Still, the example word should be less problematic. Face is good, or if we want to stick to the pattern of putting the vowel between a labial stop and an alveolar stop, we can use bait, paid, or pate (though the last word is also rather rare and runs the risk of being mistaken for pâté). —Angr (talk) 20:26, 9 May 2011 (UTC)

Actually, I'm changing it to base, since that's what Template:IPAc-en/doc already says. —Angr (talk) 20:32, 9 May 2011 (UTC)

Excellent, thanks! Glenfarclas (talk) 20:40, 9 May 2011 (UTC)

I took a glance at Angr's user contributions to find exactly which template contains these texts, so now, looking at the whole list, I have a couple of other remarks:

h = 'h' in 'high': confusing, because "high" contains two aitches, pronounced differently. Could this say "'h' in 'happy'"?
w = 'w' in 'wind': this gives me pause, because the homographs "wind [a watch]" and "wind[s of change]" are pronounced differently. I guess the w is the same in both cases, but it seems to present a cause for unnecessary puzzlement and mental stress. Could this say "'w' in 'wine'"?
ən = 'on' in 'button': I, for one, pronounce button "butt'n", with essentially no vowel at all. It that what this is example is actually supposed to indicate? Because I can also imagine pronouncing it with a vowel (though it'd strike me as a bit precious, like pronouncing all four syllables in veg-e-ta-ble), so which is right?
ʊ = schwa 'u' in 'beautiful': confusing, because "beautiful" contains two u's, pronounced differently. Could this say "schwa 'u' in 'thankful'"?

Glenfarclas (talk) 21:13, 9 May 2011 (UTC)

These were all Kwami's examples, so I'm going to hope he notices this thread and responds before I make more unilateral to the rather well-hidden Template:H:IPA. —Angr (talk) 21:24, 9 May 2011 (UTC)

I was trying for minimal sets AMAP, but that may not be possible. All the changes seem fine except for button: that's precisely the point of that example. You may pronounce it as a syllabic en, or with a schwa, but we don't distinguish because AFAIK the diff is not phonemic and in any case it would be impractical. — kwami (talk) 01:30, 10 May 2011 (UTC)

Thanks for the reply. As a non-linguist, I don't know what a "minimal set AMAP" is, but I certainly noticed that attempts were made to keep the words very similar, like having consonants followed by a long i. Having given it a bit of thought, though, I'm not sure if this is really the best idea for these popups. The net result is that you wind up with quite a few uncommon words and even proper names, such as 'nigh' and 'Zion', or words with the key letter followed by an unusual letter combination, like 'guy'. As I see it, the popup examples are not meant for linguists, but for ordinary people who want a quick way to know how an IPA glyph is pronounced. It strikes me, then, that it would be better to shoot for very common, ordinary words: 'n' in 'no', 'z' in 'zip', etc. The stranger the example word, the more, I think, people will be left trying to puzzle out if there is something unique about the pronunciation of the letter in that specific word that necessitated choosing it. Imagine telling someone, "Well, I guess you could say it's pronounced like the 'n' in 'nectarivorous'" or "like the 'z' in 'zorotypus'." I wonder if greater clarity and easier comprehension could be achieved by jettisoning examples like 'nigh', 'Zion', 'rye', 'guy', 'vie', and 'baud' in favor of elementary-school words and concrete nouns of the sort that might appear in an alphabet book, like 'no', 'zip', 'rat', 'go', 'van', and 'cause'. I assume that a linguist would perceive some corresponding drawback, too? Thoughts? Glenfarclas (talk) 21:34, 10 May 2011 (UTC)

"AMAP" = "as much as possible". A minimal set is a group of minimal pairs, which is a pair of words that differ from each other in only one phoneme; for example night and right differ only in that night starts with the /n/ sound and right with the /r/ sound. A minimal set would then be something like {tight; kite; bite; fight; sight; shite; height; might; night; light; right; wight} where each word can form a minimal pair with each other word. There are advantages to using them to illustrate phonemes, but I agree that we can eliminate some rare words like "nigh", "vie", "Zion", and "baud" in favor of more common words at the (relatively low) cost of sacrificing some purity in our minimal set. I think "rye" and "guy" are OK, though; every fluent English speaker knows how those words are pronounced. —Angr (talk) 22:00, 10 May 2011 (UTC)

You and the previous makers of this template severely underestimate the difficulty of reading English and overestimate the abilities of very many non-native (and even some native speakers!) of English when you suggest that "rye" and "guy" are OK. I have adult English students, some even CEOs of major Finnish companies, who panic, for example, when they see "pie" and even "buy"/"pay" "sale/sell". They have so much trouble trying to remember how to pronounce the insane logogram-like, analphabetic vowel symbols in these and many other English words that they don't have any energy left to notice, for example, whether they're saying a voiced or voiceless consonant or rolling their Rs. They have less or no problems saying "red" or recognizing the vowel sounds expected of them in "bad" or "boy", so there's a chance they will be able to pay enough attention to the consonant that the word is trying to illustrate so that the popover will actually help instead of confusing them. You can see why minimal pairs have no intrinsic value whatsoever and can produce lots of problems for most non-native speakers. I tried to replace some of the difficult, formal, old-fashioned, and rare examples in IPAc-en (and UK/US errors/omissions) with words known to most or at least more users, but it apparently has to be done by an admin here. There are others too that need to be fixed - "moral" makes Finns and many other non-native speakers think so much more than many other simpler words that the popover hurts more than it helps. --Espoo (talk) 01:24, 3 June 2011 (UTC)

/ən/ means either [ən] or [n̩]; almost nobody distinguishes between the two, but lots of people distinguish them from non-syllabic [n] /n/ (lightening from lighten vs lightning the noun), so it makes more sense to transcribe [n̩] as /ən/ than as /n/ (if we did the latter, you couldn't tell whether /iːvnɪŋ/ is two syllables or three). ― A. di M._plé^dréachtaí 01:56, 10 May 2011 (UTC)

On the vowel of "bath"

In the sentence: "My fAther always wears a hAt in the bAth", the three capitalised As are distict phonemes: call them A1, A2, A3. A3 is practically confined to environments where the next sound is an unvoiced fricative, pass, bath, laugh, though it also occurs in few place names: FrAnce, AlabAma. In a few words (elastic, plastic) A2 is heard in an environment where A3 might be expected. This shows that A3 is a distinct phoneme.

We can distinguish three dialect types: 1 (e.g. British RP) which merge A3 with A1, 2 (e.g. that of Northern England) which merge A3 with A2 and 3 (e.g. Southern US) distinguish all three.

Wikipedia IPA has symbols for A1 and A2 but nne for A3. It assumes that speakers will merge A3 with A1 or with A2. Thus BA2th and BA1TH are given as alternative pronunciations of "Bath", the town in Somerset. Wikipedia gives A2 for the third syllable of Alabama but this does not agree either with people who live there (who speak type 3) or type 1 speakers, who use A1.

If we use Wikipedia to find out how a word is pronounced we want to know how it is pronoumced in a particular dialect, usually, but not always, our own. A symbol for A3 would inform readers about Alabama and it would help native type 2 speakers who want to learn type 1 to distinguish between cases like "pass" and "plastic".

Bukovets (talk) 12:28, 23 May 2011 (UTC)

Even if it's true that Southern U.S. English distinguishes all three (and I'm not convinced that's true, but I know what you're getting at), the purpose of this page is not to accommodate every conceivable phonemic distinction in every accent of English. We can really only handle the accents that are well represented in dictionaries and the like, which means primarily RP and GenAm, as well as accents that aren't too divergent from like Southern Hemisphere English (which is phonetically very distant from RP but phonemically fairly close to it) and Canadian English (which is both phonetically and phonemically close to GenAm). Trying to accommodate more divergent accents such as Scottish English and Southern U.S. English would lead to a proliferation of symbols (and frankly we already use a pretty staggering quantity of symbols here) that would just make our transcriptions harder to read. And since there are no or few pronouncing dictionaries of such accents, including them would also require a fair amount of original research, which we should try to avoid. —Angr (talk) 13:00, 23 May 2011 (UTC)

But distinguishing these three vowels is something that we can do using dictionaries. This isn't the first time someone's brought this up (see the last time). The most difficult issue would be choosing the symbol. — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 14:47, 23 May 2011 (UTC)

/a/ would be an obvious choice for the BATH words (see e.g. Speak with Distinction). However, I'm more interested in the original claim. Does Bukovets have FATHER=BOTHER? I thought all U.S. southern accents did. In which case, what is the phonetic difference between AlabAma, (A3) hAmmer, (A2?) and bOmber (A1?)? Grover cleveland (talk) 03:17, 3 June 2011 (UTC)

Both the /æ/~/ɑː/ and /ɒ/~/ɔː/ alternations have been suggested before. My main concern is in maintaining the templates, so that ‹aː› (or whatever we would use) would reliably mean /æ/~/ɑː/ rather than, say, a dictionary variant for /ɑː/. Currently if ‹a› isn't followed by ‹ɪ› or ‹ʊ› we know something's wrong. — kwami (talk) 17:49, 4 June 2011 (UTC)

That's one of the reasons that we might want to choose æː. Are there any other alternatives? — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 18:49, 4 June 2011 (UTC)

The spider vowel

Re. this edit, I'm concerned that ‹əɪ› may be misleading, as we otherwise only use ‹ə› for a reduced vowel, and [əɪ] isn't reduced. Personally I prefer the reverted ‹ʌɪ›. Any reason to avoid it? — kwami (talk) 17:34, 2 June 2011 (UTC)

That's a narrow transcription between square brackets, so we should use ə if we mean a mid central vowel and ʌ if we mean a back mid-open one. (To my hear, Canadian spider is more like [ɜɪ], but I'm not a reliable source.) Anyway, I think we are giving way Too Much Information with the footnotes; we should only mention 1) where major dialects have phonemic and near-phonemic splits/mergers not reflected in our transcription and 2) when major dictionaries use transcriptions different than ours. (Hence, I'd omit footnote 2, as AFAIK no dialect has two contrasting phonemes /r/ and /ɹ/ and no major dictionary uses /ɹ/, etc.) ― A. di M._plé^dréachtaí 19:39, 2 June 2011 (UTC)

The main reason for footnote 2 is that every few months a newbie comes along and leaves a note on this or some other talk page saying "Why do you use r instead of ɹ? r is a trill, and almost no one uses a trill in English. We should use ɹ as it's more accurate." And then one of us has to come along and explain why we use r. By keeping the footnote on the page, we have a central place we can point people to when they ask. —Angr (talk) 21:32, 2 June 2011 (UTC)

Looks like the footnote isn't effective ar that! (See section below.) Anyway, if that is the purpose, wouldn't the FAQ at the top of this talk page be a better place? ― A. di M._plé^dréachtaí 12:14, 4 June 2011 (UTC)

[əɪ] is definitely wrong. You might find [ʌ̈ɪ], [ɜ̠ɪ] or even [ɐ̠ɪ], but [əɪ] is pretty uncommon. I'd say we keep [ʌɪ] as the standard. Hear the difference here Canadian - Edinburgh EMGT.--TheAmericanizator (talk) 11:34, 5 June 2011 (UTC)

Archiving robot problems

It seems as though the automated archiving has created some strange oddities that I've fixed:

A conversation from October 2007 was in the beginning of Archive 2 (April 2008) instead of the beginning of Archive 1
A conversation from June 2010 was in Archive 3 (April to December 2008)
Fifteen consecutive conversations (an entire archive's worth, I'd say) from late 2010 were in Archive 4 (January to February 2009)
Sixteen conversations from late 2010 and early 2011 were in Archive 5 (March to May 2009)
Five conversations from early 2011 were in Archive 6 (June 2009 to July 2009)

Is there someone we should be informing about this? — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 15:16, 3 June 2011 (UTC)

Done. Thanks to User:Wwoods. — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 19:04, 10 June 2011 (UTC)

/hw/

I would really rather see the template give /ʍ/. -- Evertype·✆ 06:28, 4 June 2011 (UTC)

I wouldn't. It's an additional unfamiliar symbol for little or no added benefit. —Angr (talk) 07:28, 4 June 2011 (UTC)

This has been suggested before, but it's dubious that it's actually a phoneme in English, and as Angr said, it's an extra letter for people to learn. — kwami (talk) 17:50, 4 June 2011 (UTC)

How would that be any more useful than using, say, /ç/ in huge? ― A. di M._plé^dréachtaí 12:28, 5 June 2011 (UTC)

What everyone else said. /hw/ has the benefit that it's easily decomposable into /h/ and /w/ for the majority who don't distinguish it from /w/. (It's also the historical form, for what that's worth). Grover cleveland (talk) 18:24, 6 June 2011 (UTC)

Requested move

The following discussion is an archived discussion of the proposal. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

The result of the proposal was not moved.--Fuhghettaboutit (talk) 12:15, 14 June 2011 (UTC)

Wikipedia:IPA for English → Help:IPA for English – The key in this page links from the pronunciation help in article pages. My rationale is sort of based on WP:SELF. Help namespace is for help and this is a help page not a project page. Help:IPA for English is currently a redir to this page. -- Alan Liefting (talk) - 05:33, 7 June 2011 (UTC)

Oppose. Help namespace is supposed to be for help using the MediaWiki markup. This page is about how to use the IPA in Wikipedia articles and is thus similar to MOS pages. That's why we moved it back to Wikipedia: space from Help: space a few years ago. —Angr (talk) 10:22, 7 June 2011 (UTC)
I think the main focus should be for readers who want to interpret transcriptions, rather than for editors who want to add them. ― A. di M._plé^dréachtaí 14:58, 7 June 2011 (UTC)
That's not an argument for putting it in Help: space, though. It used to be in Help: space, but we moved it back when we were told Help: space is for help using the software and the markup language, not for general help understanding Wikipedia. That's what project space is for. —Angr (talk) 15:12, 7 June 2011 (UTC)
Weak support per nom. ― A. di M._plé^dréachtaí 14:57, 7 June 2011 (UTC)
Oppose. There is already a redirect from the proposed name, from which it was moved long ago, because it is more of the nature of a guideline, than technical help. There are arguments for both locations, but there is no point in moving it back and forth. −Woodstone (talk) 15:42, 7 June 2011 (UTC)
Oppose, per Angr. It might be beneficial to compare Wikipedia:Project namespace and Wikipedia:Help namespace. — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 19:13, 10 June 2011 (UTC)
Support makes sense with reference to WP:SELF. Donama (talk) 05:52, 14 June 2011 (UTC)
- What does WP:SELF have to do with anything? This page isn't in article space, which is what WP:SELF is about. It's in the Wikipedia: namespace, which pretty much by definition is the place for self-references. —Angr (talk) 05:57, 14 June 2011 (UTC)

The above discussion is preserved as an archive of the proposal. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

Vocal Pronunciation

Has anyone out there come across a website with the option of inputting IPA, with an output of actual sound? IPA is so impenetrable to the average user that I feel linking to something like this would be an invaluable feature. Just a suggestion/idea.

Jrtf83 (talk) 18:26, 30 June 2011 (UTC)

I've seen one but I can't remember where it is. Check out the external links at International Phonetic Alphabet. — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 12:00, 1 July 2011 (UTC)

Simplifications

What do you think about these simplifications inspired by Merriam Webster's Learner's dictionary :

1)Merging /Vɪɚ/ to /Vjɚ/ and /oʊV/ to /owV/ (V-vowel).

/flaɪɚ^r/ --> /flajɚ^r/

/leɪɚ^r/ --> /lejɚ^r/

/lɔɪɚ^r/ --> /lɔjɚ^r/

/ˈʃoʊi/ --> /ˈʃowi/

They're easier to write and on the eyes. --TheAmericanizator (talk) 20:06, 10 July 2011 (UTC)

I don't think that transcribing the same phoneme in two different ways depending on whether the following sound is a vowel or a consonant is a good idea. ― A. di M._plé^dréachtaí 20:54, 11 July 2011 (UTC)

ill-formatted IPAc-en

Just to let everyone know, I modified {{H:IPA}} to link to Category:Ill-formatted IPAc-en transclusions whenever an unsupported letter or sequence is used with that template, so we may be getting requests to fix. (This won't catch mistranscriptions using valid strings, of course.) I'm going through them myself, but it looks like there are a couple hundred. I tried to get the template to just display "??" for unsupported strings, but with the template added, it makes a mess – maybe s.o. here knows a more elegant way? (Though once we get this batch cleaned up, it won't necessarily be a bad idea to make a mess of the article: that'll discourage people from misusing the template, or maybe even encourage them to ask for help.) — kwami (talk) 22:05, 26 August 2011 (UTC)

baht

is baht a word? if so not particularly common. is there a better choice for table? Feldercarb (talk) 05:57, 19 August 2011 (UTC)

I don't think there's any particularly common word in the PALM lexical set ending with a voiceless consonant. Indeed, Wells (1982) only lists baht, Bach, and a few Italian loanwords which in BrE can also have /æ/. And I'd rather not use a foreign proper name such as Bach. ― A. di M._plé^dréachtaí 21:05, 24 August 2011 (UTC)

The original idea behind including "baht" was to have a large minimal set: bit, bet, bat, bot, butt, beat, bait, baht, bought, boat, boot, bite, bout. However this seems to have gone by the wayside. Grover cleveland (talk) 17:50, 30 August 2011 (UTC)

And yes, baht is a word, though not one in most English-speakers' everyday active vocabulary. Angr (talk) 05:53, 7 September 2011 (UTC)

Political bias

There should not be political bias in the words used to represent particular sounds. A generic name like John is fine. But to use Barack Obama, is a clear case of high-jacking the page for political advertising. 92.22.7.121 (talk) 08:57, 30 August 2011 (UTC)

In addition, "Barack" is widely pronounced with the TRAP vowel in British English. So it's not even a useful keyword. Grover cleveland (talk) 17:45, 30 August 2011 (UTC)

Who cares about politics? It's simply a matter of finding a widely recognized word. "John" has the wrong vowel. — kwami (talk) 02:53, 7 September 2011 (UTC)

{{IPA-en}} → {{IPAc-en}}

An editor has been spending the past two days converting almost 200 {{IPA-en}} templates in articles to {{IPAc-en}}.[1] I'm not necessarily criticizing the editor's actions, but I just want to know if the new template is being added to phase the other one out. Also, according to the template page, {{IPAc-en}} is supposed to replace {{IPA-en}} and {{pron-en}}, but the editor is leaving the latter template in place. I have not yet contacted the editor, but I am just questioning the reason for this massive change, and hoping it is not a matter of WP:POINT. –Dream out loud (talk) 00:31, 7 September 2011 (UTC)

IPAc started being phased in last year. 6,700 articles have been converted, almost half the total. It's supposed to eventually replace both IPA and pron, but doesn't have a parameter to automate the latter. It also isn't always appropriate, so there will probably remain a smallish number of IPA-en transclusion after conversion. I don't have much preference either way; it's probably a good idea to have individual popups for phonemes (I once tried generic popups with IPA-en, but WP coding isn't up to the task), but it's more difficult to check and maintain IPAc-en, so as an editor I prefer the former. — kwami (talk) 02:49, 7 September 2011 (UTC)

How do you determine when it is and isn't appropriate? –Dream out loud (talk) 07:27, 7 September 2011 (UTC)

If the template doesn't support the transcription, so that it triggers the error category we have set up for unsupported transcriptions, then IMO it isn't appropriate. But that may just be me. — kwami (talk) 07:35, 7 September 2011 (UTC)

So if we fix {{IPAc-en}} to support all transcriptions, does that render {{IPA-en}} to be obsolete? Also, should {{pron-en}} be removed from the articles with the new template in place? –Dream out loud (talk) 06:39, 10 September 2011 (UTC)

Well, we've got a way to go, but there isn't much point in having 3 templates for the same thing, so I expect that they will eventually be eliminated. IPAc-en also gives us greater consistency. Say, if we decide that putting optional elements in parentheses makes it too hard to read, we simply don't support parentheses. Or if we decide we want a phonemic transcription for vowels, we can just change the output. Etc. — kwami (talk) 08:59, 17 September 2011 (UTC)

transliteration of Merriam-Webster

Could someone please help with the IPA for Nigerien? Should naɪdʒərɪˈen be naɪdʒɪriˈen? --Espoo (talk) 06:49, 17 September 2011 (UTC)

Neither. That sounds to much like "Nigerian". It's /niˈʒɛriɛn/. At least, that's the only pronunciation I've heard in English. — kwami (talk) 08:56, 17 September 2011 (UTC)

But when Merriam Webster says ˌnī-ˌjir-ē-ˈen what they mean in IPA is [naɪdʒɪəriˈɛn]. Angr (talk) 09:25, 17 September 2011 (UTC)

r and oo

is there any good reason that wiki uses <r> and not <ɹ> for the rhotic approximate [ɹ~ɻ]? [r] doesn't exist in any dialect of English, excepting perhaps dialects which have been influenced by languages like Afrikaans (SAE) and Scots or the Gaelic languages. Also, why is the oo sound written <ju:> and the put vowel wirtten <ʉ> when [u:] again only exists in radical dialects and most English dialects use a diphthong [ʉʊ] or ([yʊ] in some aussie dialects). While we're on the tpoic, why is [ʍ] represented as <hw>? The benevolent dictator (talk) 12:58, 3 June 2011 (UTC)

See here, here, here, here, here, here, here, and here for other times this has been brought up. — Æµ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 15:16, 3 June 2011 (UTC)

A vast majority of english speakers recognise [ɹ] and [ʉː] as the phonemes /r/ and /uː/. Plus they're easier to type on a standard keyboard, widely used (even in dictionaries), and better recognised. Plus tradition or something. As a native speaker, I'm sure that AuE /uː/ is a monophong, and the many mistakes involved in transcribing accents is yet another reason to use the "standard" /r/ and /uː/. P ø r k (talk) 16:03, 7 August 2011 (UTC)

Whenever I write in the IPA, I always use the real symbol for r, ɹʷˤ. It's important that we be precise when we write in the IPA; since r in fact represents a rolled r, an intelligent reader of the IPA from another language who read that in a dictionary may say things like, "RRRRoberrrrt, wherrrre's yourrrr trrrruck?" Tradition is not always best, and the IPA really should reform this. Inter change able|talk to me 17:39, 17 September 2011 (UTC)

What, complete with pharyngealization mark?? No, in fact it's not important that we be precise when we write in IPA. (It's important that we be accurate, but accuracy and precision are two very different things.) The IPA is useful for both broad and narrow transcriptions, and broad transcriptions are most useful in contexts like encyclopedias and dictionaries. In broad transcription, using the symbol /r/ for a sound that's phonetically realized as [ɹʷ] (or [ɻʷ] or [ɻ̠ʷ] - the high degree of variability in this sound is another reason to avoid excessive precision) is no more inaccurate or problematic than using the symbol /iː/ for a sound that's phonetically realized as [ɪi̠˕], or failing to mark nasalization, glottalization, aspiration, etc. etc. Angr (talk) 17:55, 17 September 2011 (UTC)

I don't care at all about how Wîkipedia treats it - they can transcribe it as r for the sake of "convenience" and "tradition" all they like; I strive for precision because many English speakers tend to stereotype foreign accents, so it's important that they know the exact correct sound when speaking (regardless of the dialect they learn). In my mind it's just plain silly to make people who use the IPA to study English think that they have to roll their r's, simply for the sake of our "traditions". But what the heck do you mean by [ɪi̠˕]? Does that sound exist in any dialect of English? Inter change able|talk to me 19:45, 17 September 2011

Yes, [ɪi̠˕] is the usual pronunciation of the vowel of see in most educated British and North American accents. If you wanted to be precise for the sake of foreigners learning English, you would have to use that symbol instead of [iː], lest anyone think the English vowel sounds like French [i] of si and German [iː] of Sie, which are much higher, much fronter, and much more monophthongal than the English vowel. If you want to transcribe ɡreen as [ɡ̊ɹʷɪ̃ĩ̠˕n] in your own work, of course that's up to you, but I think both native and non-native speakers alike would be better served by a more comprehensible transcription like /ɡriːn/ combined with various conventions about the details of English phonetics. Angr (talk) 20:42, 17 September 2011 (UTC)

Truly, I don't believe that I say "unvoiced g, r, nasalized ɪ, nasalized/retracted/voiced (are vowels not already voiced?) i, n", to use a crude transcription which I assume you mean by that green. I assume that means I'm uneducated. But let's keep this to r. What I am trying to say is that it is silly to use a symbol for a completely different sound that sounds nothing like the one that we intend to transcribe a sound which already has its own symbol and has had one for some time, simply for the sake of "convenience" and "tradition". I don't want to reform Wikipedia here, but it is a pet peeve of mine that our transcriptions may be misdirecting those who are trying to learn English. Your examples can be easily usurped by a simple /i/, since the pronunciation is so undetectably similar, but the difference between a rolled and true labialized and pharyngealized r will get laughs from any native speaker. Inter change able|talk to me 22:51, 18 September 2011 (UTC)

superscript consonants

Thought I'd try bringing this up again. There will never be an end to the edit wars from people who insist that there is no /r/ in "here", and to a lesser extent from those who insist there is no /j/ in "new". Why not just superscript these, as we do with vowels which are optionally dropped? The previous argument against was that would not be phonemic, but we're not exactly phonemic as it is. And there's plenty of precedence in English dictionaries. It would make things so much easier than having to constantly police the articles, and having many of them inconsistent with the key. — kwami (talk) 15:22, 5 October 2011 (UTC)

One argument against the proposal is that the superscript j already means something else in IPA and people are already complaining that we use too many IPA symbols like algebraic variables instead of to stand for their intended phonetic values. (I'm thinking specifically of ɨ, ɵ, and ʉ.) Also, I don't think it will work: people who object to transcribing new as /njuː/ will not actually be mollified by transcribing it /nʲuː/. Angr (talk) 18:32, 5 October 2011 (UTC)

I fully agree with Angr; English has no palatalized consonants (and let us keep it that way!). People who insist that their own dialects of English are the best should simply be ignored. Inter change able|talk to me 19:37, 5 October 2011 (UTC)

boor, moor and tourist

The article says that boor, moor, and tourist have the same (first) vowel. While it lists boar and more as having a different vowel. To me boar and boor are alternate spellings for the same word, and moor and more are homophones. In case this has been discussed before, I'm commenting on it here. Imc (talk) 18:50, 6 October 2011 (UTC)

Did you read footnote 15? Angr (talk) 19:46, 6 October 2011 (UTC)

Different dialects of English differ so much that it frequently amazes me that (for the most part) they are all mutually intelligible. In my dialect, the words that you provided are pronounced [buːɹʷˤ], [muːɹʷˤ], [tʰuːɹʷˤɪst], [ˈboɚ], and [boɹʷˤ] (the distinction between boar and bore is very slight but observed thoroughly). Wikipedia's pronunciation does not revolve around any dialect (nor does its spelling; proposals to enforce one variety of spelling have been repeatedly rejected). Inter change able|talk to me 22:32, 6 October 2011 (UTC)

No aspirated consonants?

Why are the aspirated consonants (e.g. /pʰ/, /tʰ/, and /kʰ/) not listed on the article as separate consonants as they are on the article Wikipedia:IPA for Icelandic? Since aspiration is nearly universal to these phonemes across English dialects, why not list them (and perhaps add a footnote stating the dialects that don't use them, if complaints arise). Dialects of English that do use these aspirated consonants include all the major ones of the world (specifically Australian English, Canadian English, General American, Irish English, Received Pronunciation, New Zealand English, Scottish English, South African English, and Welsh English), so the minority would be those dialects which don't. Since this article even includes as a phoneme "ŋg", why not a much more common phenomenon: aspirated consonants? Wolfdog (talk) 23:27, 10 October 2011 (UTC)

Because it is traditional to transcribe them /p t k/. There is no benefit to adding the diacritic, and it would be counter-intuitive to most of our readers. /ŋɡ/ is a sequence; it's given so that the reader understands the distinction between it and /ŋ/. — kwami (talk) 05:52, 11 October 2011 (UTC)

Must tradition always prevail? Inter change able|talk to me 15:03, 11 October 2011 (UTC)

No, but in this case the tradition is not without good cause. The phonemes /p/, /t/, /k/, /tʃ/ are aspirated only word-initially and before a stressed syllable (not preceded by /s/). Elsewhere they are unaspirated, possibly glottalized, or in the case of /t/, even flapped. Arguing for this nondistinctive feature (as aspiration is not itself distinctive in English, as it is in Icelandic) is tantamount to arguing for a much more narrow transcription generally.

An unrelated question: does anyone know why Wikipedia's IPA templates suddenly underline the transcription? This never happened until a few days ago, and I find it quite off-putting, but I can't track down the relevant edit in the nested templates. — ˈzɪzɨvə (talk) 22:40, 11 October 2011 (UTC)

I don't see any underlining. — kwami (talk) 05:29, 12 October 2011 (UTC)

Perhaps ˈzɪzɨvə witnessed recent IPA-en to IPAc-en replacements. (IPAc-en should display a dotted underline which provides a hint to the phoneme help available on mouseover.) – RVJ (talk) 10:25, 13 October 2011 (UTC)

No, it wasn't that. I suspect it was some setting with my browser in combination with a minor change Wikipedia-wide. When I checked my WP settings, the "underline links" option had gotten set to 'always'. Switched it to 'browser default', which doesn't underline any links, except on mouseover, and never the IPA ones. Oh well, good enough for me. — ˈzɪzɨvə (talk) 10:50, 13 October 2011 (UTC)

Citations for pronunciations

If various editors keep sticking in incorrect pronunciations (at Vermont, for example) how can I provide a citation to a reliable source, when the dictionaries I have access to do not use IPA? Jc3s5h (talk) 03:03, 31 December 2011 (UTC)

dictionary.com uses IPA. I believe there is also a poorly-touted guide somewhere around here that compares the conventions of various dictionaries with the IPA system used at Wikipedia. — Æµ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 04:16, 31 December 2011 (UTC)

The OED uses IPA and has an entry for Vermont. I corrected the article. — kwami (talk) 04:34, 4 January 2012 (UTC)

/ɒr/ and /ʌr/ examples

Can someone explain why Moral and Forage are used as examples for /ɒr/? Do I misunderstand the pronunciation of /ɒr/ or does it not sound like "awr?" While I certainly often hear people say "Mawral" and "Fawrage," I also hear people say "Moral" and "Forage," as in /moʊral/. I think a more universal example word (or words) would be appropriate, since the split between "moral" and "mawral" seems about even. I would suggest an example for /ɒr/ be "swarm" or something else. Or am I just misunderstanding the key for /ɒr/?

For /ʌr/, the "borough" example is also a fairly even split between "bʌroʊ" and "boʊroʊ", and thus unsuitable as an example. I would suggest removing "borough" outright or replacing it with something such as "urn". — Preceding unsigned comment added by Nitrowolf (talk • contribs) 15:56, 5 January 2012 (UTC)

The words you gave do not have those sounds, though it may not make any difference in your dialect (as in mine). Neither occur before an /r/ at the end of a syllable. Hurry, for example, has /ʌr/, but it's impossible in a word like urn. — kwami (talk) 17:10, 5 January 2012 (UTC)

I'm ok with Hurry being an example... but I don't think borough is a good example. Perhaps it should just be removed?

What sound is /ɒr/ then, if not the ar in "swarm?" For me, Moral and Forage are both /oʊr/ — Preceding unsigned comment added by Nitrowolf (talk • contribs) 20:06, 5 January 2012 (UTC)

/ɒr/ is the sound of, well, moral and forage. It may be that it's the same sound as the /ɔr/ of "swarm" in your accent, but it's different in other people's accents. For example, for some people foreign rhymes with snorin’, but for other people they're different, and this key shows different symbols for the sake of people who pronounce them differently. People who pronounce them the same just have to remember that /ɒr/ and /ɔr/ (and, for that matter /ɔər/) stand for the same sound in their accent. Angr (talk) 20:48, 5 January 2012 (UTC)

So how to account for or in "oral" being /ɔər/ or /oʊr/ and not /ɒr/ or /ɒʊr/, since they are the same sound in the key? Websters seems to agree that moral, forage (and foreign) are /ɔər/ as primary and /ɒr/ being a secondary pronunciation. The examples are fine so long as the key is consistent, but it's not, which is where the problem is IMO. I don't think moral and forage are good examples of /ɒr/, since pronunciation is fairly evenly split between the two. Nitrowolf (talk) 14:28, 6 January 2012 (UTC)

Webster's is not reporting 1ary and 2ary pronunciations, but a dialectical difference. We don't do that. — kwami (talk) 14:42, 6 January 2012 (UTC)

Ok, if their first pronunciation is not considered the primary, in either case, they are split and thus either a) unsuitable or b) should be made consistent with the rest of the key. Honestly, I am just after consistency in the key - the specifics of the consistency is not my primary goal here. Can we come to an agreement on consistency? How do you reconcile "oral" being /ɔər/ or /oʊr/ and "moral" being /ɒr/?

We need to either change "oral" and all related words to match Moral/Forage, change /ɒr/ to be consistent with words that have that sound or remove Moral/Forage from examples of /ɒr/ and replace them with something else. They can't all be correct as they are inconsistent with each other. — Preceding unsigned comment added by Nitrowolf (talk • contribs) 15:58, 6 January 2012 (UTC)

The thing is, in RP and other non-North American varieties of English, oral and moral don't rhyme, so we transcribe them differently (/ˈɔərəl/, /ˈmɒrəl/) to accommodate those accents. Angr (talk) 16:19, 6 January 2012 (UTC)

Be vigilant when using North American sources for /ɔr, ɒr, ɔər/ or /ɛr, ɛər, ær/, even /ɒ, ɔː, ɑː/, as considerable confusion can set in, even when the dictionary uses IPA. I know I for one cannot transcribe them accurately without guessing from the orthography (which is, as you can see, an unreliable method) or using a dictionary. — ˈzɪzɨvə (talk) 18:31, 6 January 2012 (UTC)

References Section

Is it not a glaring mistake that this article has no References, or am I missing something? The only reference it needs is the most current Journal of the International Phonetic Association, which can be found here: http://journals.cambridge.org/action/displayJournal?jid=IPA. Lee (talk) 02:39, 19 November 2011 (UTC)

This isn't an article (it's in Wikipedia: namespace), so it doesn't need references. Angr (talk) 09:58, 19 November 2011 (UTC)

The idea that the Journal of the IPA should be the only reference needed is quite wrong in any case. The Journal publishes highly specialized articles on phonetic topics, plus "specimen" descriptions, with audio recordings, of particular languages or accents (I wrote the one on British English ("RP")). I think the writer might have intended to refer to the Handbook of the International Phonetic Association (Cambridge University Press, 1999), which does indeed explain all the principles of phonetic and phonemic symbolization, and would be an excellent basic reference. Peter Roach RoachPeter (talk) 10:16, 11 March 2012 (UTC)

"Mayor"

Note 7 says: "Note that many speakers distinguish monosyllabic triphthongs with R and disyllabic realizations: ... hire /ˈhaɪər/ from higher /ˈhaɪ.ər/, ... mare /ˈmɛər/ from mayor /ˈmeɪ.ər/."

I myself am one of these speakers, but in spite of the spelling, mayor is pronounced in one syllable, the same as mare. I think layer would be a better example. Or is this a case (similar to, for example, waistcoat) of a word whose traditional pronunciation has given way to one more in accordance with the spelling and I'm the odd one out for continuing to pronounce it the older way? Does anyone know? 2.25.121.54 (talk) 19:25, 2 January 2012 (UTC)

I've heard layer pronounced with one syllable too; let's try player. ― A. di M. 20:26, 2 January 2012 (UTC)

It's not clear what the note means by "distinguish". It is certainly true that many speakers of British English "Received Pronunciation" or "BBC Pronunciation" have different pronunciations of what traditionally have been represented as triphthongs, i.e. the vowel sequences /eɪə, aɪə, ɔɪə, əʊə, aʊə/. The presence or absence of the letter 'r' or the phoneme /r/ (or whatever the writer of the note meant by R) is not relevant as far as I can see. The difference is explained in the Introduction to the 18th Edition of the Cambridge English Pronouncing Dictionary (p vii) in this way: "Some triphthongs are pronounced like single syllables ... while others are more likely to be divided into two syllables. We usually find the two-syllable pronunciation (i) when the schwa is a separate morpheme (e.g. '-er' in 'buyer' /baɪ.ə/, (ii) when the word is thought to be foreign (this includes many biblical names originating from Hebrew, e.g. 'Messiah' ... and (iii) when a word is not used very frequently (e.g. 'cyanosis')". This section also points out that the single-syllable pronunciation found in e.g. 'fire' is often reduced almost to a pure long vowel as a result of smoothing. All the above relates to RP/BBC, and I feel more thought needs to be given to other accents of English. Peter Roach RoachPeter (talk) 10:42, 11 March 2012 (UTC)

"New"

The IPA template translates /nj/ 'n' in 'new' . In many (most?) NA regions, /nj/ is not 'n' in 'new' and leads to confusion, especially with proper names. Most NA regions pronounce "new" as /nuː/ (North East is an exception, but even then few people say "Ny-oo York" as opposed to "N-oo York"). Can we please change the template to a more agreed upon word to represent /nj/, such as "onion"? — Preceding unsigned comment added by Nitrowolf (talk • contribs) 16:47, 3 January 2012 (UTC)

Yes, as the explanatory guide says, we're trying to encode for multiple dialects. If your dialect is one that has dropped this j after alveolar consonants, you can ignore it when you read our transcriptions. — Æµ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 18:26, 3 January 2012 (UTC)

I'm not sure I understand your answer. Are you saying "Yes, go ahead and change it to "onion" as opposed to "new?" I realize people can drop the j the dialect applies, but my point is the example is not accurate for all dialects and there are alternatives that ARE accurate for all (or at least most) English dialects, and thus that/those should be used instead. — Preceding unsigned comment added by Nitrowolf (talk • contribs) 22:14, 3 January 2012 (UTC)

I'm not sure I understand where you want to make this change. On this page, the only place /nj/ is mentioned is specifically in the discussion of the fact that some dialects have /nuː/ where others have /njuː/. It would make no sense to change it to onion there. Angr (talk) 23:21, 3 January 2012 (UTC)

Perhaps it's a little confusing to Nitrowolf because there are still instances of /nj/ in yod-dropping dialects (onion being an example). Our explanatory guide covers this, but the template he's talking about just uses new as an example. Is there a way we can make this clearer? — Æµ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 02:32, 4 January 2012 (UTC)

Yes, that's what I was trying to get at. The template (Template:H:IPA) was the only thing I was talking about - just changing the example to something that is more universal - not changing the guide. Which, incidentally, I'm not sure specifically what is being referenced when you say the explanatory guide. Do you mean the heading "Understanding the key" on the IPA page? Nitrowolf (talk) 14:33, 4 January 2012 (UTC)

The fact is, in this transcription scheme /nj/ is used diaphonemically to mean ‘whatever you pronounce new like’ (e.g. /njuːˈtriːnoʊ/ for neutrino); if the tooltip said “ni in onion” it would be inaccurate for speakers from North America, as they don't pronounce the beginning of neutrino the same as the middle of onion. (BTW, I've just noticed that IPAc-en doesn't support /u/ for sinew or influenza; would it be useful?) ― A. di M. 17:14, 4 January 2012 (UTC)

Who distinguishes that from /u:/, and how would we know which was which? — kwami (talk) 17:30, 4 January 2012 (UTC)

Some old-fashioned accents pronounce it the same as /ʊ/ rather than /uː/ (it's the analogous of the lack of happy tensing). The LPD uses /u/ for it[2] and some American dictionaries use the convention of marking (primary or secondary) stress on all syllables with unreduced vowels regardless of lexical stress, hence in those dictionaries /uː/ without any stress marker means inTO rather than GOOSE. ― A. di M. 17:39, 4 January 2012 (UTC)

Distinguishes what from /u:/? For me, sinew works for me as well, as the "new" in sinew is /nj/ to me. Is there another pronunciation? I've not heard one, other than /sin'nju/. I'm not sure what the tool tip of ni would have to do with it, as the /ni/ in onion is RP /nj/, and thus as you say unsuitable... whereas, to my knowledge, the /ni/ in onion RP as /nj/ is nearly universal across all dialects? I'm not saying this is true, but it is to the best of my knowledge. I am absolutely open to another example word for /nj/ than 'onion', but 'new' is definitely not a good example under any circumstance. Nitrowolf (talk) 21:47, 4 January 2012 (UTC)

A. di M., I think this has come up before, but it was decided that the functional load of /u/ was so low, and the chance of spurious /u/ so high, that it wasn't worth it. It might be time for a rethink, though, if you find it useful.

Nitrowolf, I suspect that any word will meet with the same objection that new does. Can you think of a word that has [nj] in RP, [n] in GA, and would not cause a problem? — kwami (talk) 17:14, 5 January 2012 (UTC)

Does "onion" not qualify? Nitrowolf (talk) 19:54, 5 January 2012 (UTC)

No, because it has /nj/ in both RP and GenAm. Angr (talk) 20:02, 5 January 2012 (UTC)

Wait, I don't understand - why do we want a word that is /nj/ in RP and /n/ in GA if we are giving an example of /nj/? Wouldn't we want a word that is /nj/ in both RP and GA? — Preceding unsigned comment added by Nitrowolf (talk • contribs) 14:16, 6 January 2012 (UTC)

The sequence /nj/ has its own entry in that template just to ward off yod-droppers from removing the /j/. It's not its own phoneme, and when it occurs after a stressed vowel—in a word like onion—you should just transcribe the /n/ and /j/ separately. — ˈzɪzɨvə (talk) 18:31, 6 January 2012 (UTC)

I'm still not sure why you'd want an example of /n/ in GA used to explain the sound of /nj/, do Brits not pronounce "Onion" with /nj/ ? ... but ok.. How about "Neural", "Neuron", "Neurotic" or "Neurosis" or a derivative there of? — Preceding unsigned comment added by Nitrowolf (talk • contribs) 19:39, 6 January 2012 (UTC)

Look, the entry /nj/ in that template should only be used in those cases where North American has /n/ but others have /nj/. It's a workaround to make clear that yod-droppers should just ignore the /j/ in this diaphonemic transcription, similar to the /r/ we encode with the vowel for the sake of nonrhotic speakers. In words where nobody drops the /j/, like onion (though personally I have /ŋ/ in onion, but that's another story...), just transcribe them separately, because they're two phonemes that just happen to be next to each other. Your examples would work just fine for that (at least I think—I'm a yod-dropper myself, so I'm relying on the orthography), as does new, but what's wrong with using new? — ˈzɪzɨvə (talk) 05:52, 7 January 2012 (UTC)

I can't see what this discussion is about. The idea of using IPA symbols is to have one symbol per sound. If you say /nu:/ for 'new', then you have two phonemes and you should use /n/ and /u:/ to represent the word. If you are not a "yod-dropper" and say /nju:/, then you have three phonemes. There is simply no need to have a separate item /nj/ in Wikipedia's IPA inventory. Peter Roach RoachPeter (talk) 10:51, 11 March 2012 (UTC)

Our transcription scheme encodes for the /j/ in new, even though part of our readership consists of yod-droppers. The /nj/ element is indeed biphonemic, but is necessary to help yod-droppers understand not to remove the /j/. — Æµ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 16:43, 11 March 2012 (UTC)

Examples used for Tooltips

In the table provided in this article, the examples for i are fleece, seed, feel, mean, and sea, but the example given in the IPA tooltip is bead. This may be clear to anyone who is already comfortable with goofy English pronunciation, but given the ambiguous pronunciation of "ea" in read and lead, I suggest that fleece of the standard lexical set or at least feel or seed would be more helpful to the readers actually using the tooltip to learn how to pronounce a word. -AndrewDressel (talk) 14:54, 3 March 2012 (UTC)

Actually, I don't link the phoneme-by-phoneme tooltips at all, given the large amount of allophony in English. I'd pretty much prefer tooltips for entire onsets and rhymes specified by hand, such as /kwɔrk/. Anyway, this discussion belongs on Template talk:IPAc-en. ― A. di M. 17:43, 3 March 2012 (UTC)

Thanks. I've cross posted this discussion there. -AndrewDressel (talk) 21:20, 3 March 2012 (UTC)

To any future readers: please respond there not here. (I've found out that without such an explicit request, cross-posted discussions will eventually branch, which is no good.) ― A. di M. 23:49, 3 March 2012 (UTC)