Help talk:Pronunciation respelling key/Archive 2

Is this page still here?
If there's no consensus to delete it, it should at least be moved to the Wikipedia: namespace, because it's not an encyclopedia article. —Keenan Pepper 14:09, 11 April 2006 (UTC)

Pronunciation respelling key → Pronunciation respelling key … Rationale: It's not an encyclopedia article and it's full of self-references, so it should not be in the main namespace. … Please discuss/vote at Talk:Pronunciation respelling key —Keenan Pepper 14:12, 11 April 2006 (UTC)
 * True, it's not an article, but no, it's not full of self references (there were two, but they've been removed). kwami 00:53, 12 April 2006 (UTC)

Survey

 * Add *Support or *Oppose followed by an optional one-sentence explanation, then sign your opinion with  ~


 * Support: this is original research by the editors and therefor cannot be in the main namespace. Indefatigable 17:49, 11 April 2006 (UTC)
 * No, it's not original research. That term means a "novel interpretation" of material, and there is no interpretation involved here. It's simply a convention, the way the layout of an article is a convention. kwami 00:53, 12 April 2006 (UTC)
 * Our conventions on the layout of articles are not documented in the main namespace; they are in the Wikipedia: namespace.  This article also fails the verifiability criteria, so this is another reason it ought to be moved out of the main namespace. Indefatigable 03:17, 13 April 2006 (UTC)
 * Oh for crying out loud, there's nothing to verify! It's a convention. The same as in any print encyclopedia. kwami 09:05, 13 April 2006 (UTC)


 * Print encyclopedias also have a division between the actual articles and things like the index and pronunciation key. This is not an article, so it doesn't belong in the article namespace. —Keenan Pepper 10:42, 13 April 2006 (UTC)


 * Yes, I agree with you there. kwami 10:50, 13 April 2006 (UTC)

We're not getting any votes, but I'll move it myself if I can get answers to the questions below. kwami 07:43, 14 April 2006 (UTC)

Support, if reluctantly, re-spelling. While IPA notation may be more precise, and may have other advantages as well, I fear that most people (at least in the United States) simply will not make the effort to learn it sufficiently well to be able to use it in any practical manner. This must be the conclusion reached by American dictionary publishers as all (unlike British publishers) use some form of (slightly different) re-spelling schemes in their pronunciation keys. I suppose the underlying philosophy could be expressed by the old notion that the familiar is preferable to the ideal. dshep/07aug2006 —The preceding unsigned comment was added by 4.154.111.221 (talk • contribs) 05:34, 7 August 2006 (UTC).

Would it be possible to devise some re-spelling scheme that does not require bold letters, nor for that matter (not that you have them), italics, capitals, underlines or any other modification of standard letter-forms? -- dshep/21aug/2006 —The preceding unsigned comment was added by 4.154.121.107 (talk • contribs) 04:08, 22 August 2006 (UTC).


 * Support - reluctant and, yes, too late since it's already been done. No, it does not belong in the main namespace.  The problem is that it doesn't belong in the helpspace either.

Discussion

 * Add any additional comments


 * If it's in namespace, can we still link to it from main Wikipedia articles? kwami 18:33, 11 April 2006 (UTC)


 * If anything, shouldn't it go in Wikipedia Help, the way WikiHiero syntax is? kwami 00:34, 12 April 2006 (UTC)
 * Support moving to Help namespace Ashibaka tock 02:48, 24 April 2006 (UTC)
 * OK, I'll do that. However, I'll move it back if people start objecting to links within articles. kwami 07:32, 28 April 2006 (UTC)
 * I commend your attitude. But I'm not sure whether or not it's a good idea to move it. Once it's moved, getting it back may will be difficult. On the other hand, it does seem a sensible way forward. The problem is, people are trying to use Wikipedia to promote IPA, and this isn't likely to stop, however much other Wikipedians show that they are not ready to adopt it as a policy.
 * So my advice is, let's be crystal-clear about the policy on wikilinks from articles to help pages first. Andrewa 16:43, 28 April 2006 (UTC) Andrewa 16:43, 28 April 2006 (UTC)


 * A little late now! :)
 * I also promote the IPA; I just don't see why we can't have both for English names, just as we have both IPA and sound files for foreign names. kwami 17:24, 28 April 2006 (UTC)


 * I can even make a case for using IPA exclusively. But, IMO such a policy won't have any chance of adoption. So I must go with the community decision. Andrewa 18:31, 28 April 2006 (UTC)


 * If we could use only one system, it would be no contest. But since the whole point is to help as many people as possible ... kwami 00:19, 29 April 2006 (UTC)

Of course you're not getting any votes. It's impossible to tell what proposal we're voting for or against. —Michael Z. 2006-08-07 06:09 Z 

....

Readers of this page will probably be interested in the respelling scheme used by the BBC to assist its broadcast speakers. To quote from J. C. Wells' blogsite, "The BBC uses a special respelling system rather than IPA, since it is intended for native speakers of English (not phoneticians) addressing other native speakers of English." You may see it at http://www.bbc.co.uk/commissioning/delivery/phonetic.pdf...--dshep/22aug2006 —The preceding unsigned comment was added by 4.154.103.187 (talk • contribs) 03:06, 23 August 2006 (UTC).

....


 * I was too late to vote on the move to the help namespace but I voted to support it anyway because, as I mentioned, the page doesn't belong in the main namespace. The problem, as I also mentioned, is that it doesn't belong in the help namespace either.


 * Kwami is absolutely right: this is not original research but a convention and as such doesn't need varification. However, the problem is that it is not the convention agreed upon by the editors at Wikipedia.  The agreed convention is to use the IPA.  This page goes against accepted practice.


 * Now, going against accepted practice is fine, however, typing up a rival convention is not a productive way of going about it. The way it should be done is to discuss it at the Manual of Style (pronunciation) talk page.


 * It doesn't belong in the main namespace but it doesn't belong in the help namespace either. It just doesn't belong in Wikipedia.  I appreciate that a lot of work has gone into it and I'm not suggesting that it's a particularly bad system, however, until it is accepted as an alternative transcription system for Wikipedia this page should not exist.  I missed the vote for deletion too.


 * Jimp 08:34, 29 August 2006 (UTC)


 * Happy to oblige, if you can come up with an alternative. (It should be obvious from the debate at Manual of Style (pronunciation) that the IPA is not a functional alternative, or at least wasn't the last time I checked.) kwami 16:04, 29 August 2006 (UTC)


 * Functional or not, Kwami, what I'm saying is that it's the IPA which (at least currently) is the agreed-upon standard. It's your key which is an alternative.  Yeah, I think that I could cover the discinctions you make here using an IPA-based system but it would be my version and would have no place here at Wikipedia without concensus.  Anyhow I'm off to sleep. Jimp 18:22, 29 August 2006 (UTC)


 * See also

Overcomplicated key
If this is a key, why not just have "a as in cat" and leave it at that? What's the point of showing how the vowel in cat is pronounced in several dialects of English here? That information may belong in an encyclopedia article, but isn't it just superfluous clutter in this guide. —Michael Z. 2006-08-30 01:49 Z 

Comments on the system itself
Questions of where the page should go and whether is should exist at all aside, I feel like making a few comments on the system itself.

1) The word "after" is transcribed as whilst the word "fern" is transcribed as .  These two s are quite different vowels to me.  Now, I assume that stress marking is intended to take take care of this.  However, this is not mentioned anywhere on the page and, indeed, there is no mention of the "er" of "after" in the table. - Jimp
 * True. Remedied easily enough; there's already a proposal to use er, ir, ur for the sounds in words like fern. However, I'm not competant to make the change, since I don't control these distinctions. - Kwami
 * Using er, ir & ur would be good for those who still make the distinction and would be fine for those of us who don't ... but how do you transcribe "word" and "work"? Jimp 01:03, 31 August 2006 (UTC)

2) Talking of stress marking, Kwami, why indicate it by an apostrophe after the stressed syllable? Why adopt an ad hoc convention contrary to the IPA system when neither would appear more intuitive?  Indeed what might be most intuitive would perhaps be to use bold for stress but this would not be possible unless another way of distinguishing  were found. - Jimp
 * I was trying to keep the system as close as possible to the US dictionary systems that many people are already familiar with. (And Usonians, of course, are the main demographic that has trouble with the IPA.) Nearly all of these (at least that I've seen) have the opposite convention of the IPA. Bold for stress might be a good idea. (Easier on the eyes than ALL CAPS, at least.) - Kwami
 * I didn't know that US dictionaries used the opposite convention. I wonder why they do this.  You could, of course, use SMALL CAPS but I'd prefer bold.  S MALL CAPS could, however, be a nice way of avoiding the use of digraphs (e.g. "I might meet my mate." ==>> ( I m I t m E t m I m A t)) but now we're moving away from a spelling-based system. Jimp 01:03, 31 August 2006 (UTC)

3) Which brings me to my third point. Why use bold this way? I'd prefer to use uu verses oo than oo verses oo.  I would make for a more straight-forward system.  Either way the reader would have to refer to the key to know the difference. Jimp 05:40, 30 August 2006 (UTC)
 * I had a hard time with this. Most keys use diacritics for this. I even used ew instead of oo in places where there would be no confusion, just to avoid bolding the text. Uu would be okay, I guess, but would be unfamiliar to most people. (With oo at least they know it's going to be one "oo" sound or the other.) But then we've gotten rid of conventional ah and oh because a couple people made a fuss, so we might as well add one more. kwami 18:13, 30 August 2006 (UTC)
 * Why forego the common expressions 'oh' and 'ah' simply because some silly person protested. After all, both are actual words known to all, and their sounds known.  —Preceding unsigned comment added by 74.190.180.191 (talk) 05:53, 18 January 2010 (UTC)
 * If this article is to be believed the number of places where using ew instead of oo would cause no confusion will be exactly zero ... at least for the Welsh. A number of spelling reform proposals I've seen use uu.  I've even seen an Oxford dictionary which used uu in its pronunciation key.  The question would remain, though, as to which digraph to use for which phoneme.  The dictionary I mention used uu for  and oo for .  This has the advantage of keeping the key for  similar to the symbol for  which will be nice for those of us lacking the foot-strut split.  However here's a good counter argument (though this is in the context of spelling reform). Jimp 01:03, 31 August 2006 (UTC)

Archive?
This page is getting a bit too long and I feel that it is time to archive this page. If there are no objections over this, I shall be doing this within the next 24 hours. -- S iva1979 Talk to me 04:26, 18 September 2006 (UTC)
 * It is done now. If other editors have any issues regarding this, please inform and discuss it with me. Thank you for your patience! -- S iva1979 Talk to me 17:29, 23 September 2006 (UTC)

Tidy archive
I've moved back from the archive as  was an orphaned subsection left here. jnestorius(talk) 21:20, 28 September 2006 (UTC)

Classical names
This section has several problems. For one thing, it makes generalizations about pronunciation that are not so, for instance that "r always closes a syllable" (and the incorrect stressing of Elara), or the speculation on the change in vowel quality before semivocalic i and e (which, if followed, would make people pronounce "Sirius" like "Cyrus"). For another, it omits a good deal of necessary information on the English pronunciation of classical names. I am reluctant to make corrections, however, since the information really doesn't belong in a help page about "pronunciation respelling", since the pronunciation is really a completely separate question from questions of spelling. It deserves its own page. RandomCritic 17:26, 17 December 2006 (UTC)
 * Let's write that page. I've been longing to see it. As to the idea that "r always closes a syllable", you are right that this is incorrect: but the author obviously meant that "r colors the preceding vowel regardless of what syllable it is in." (Don't know about "Elara": what syllable is that properly stressed on?)


 * As to vowel tensing before "semivocalic" i: the rule as originally stated is generally correct, though you have identified the one big exception: the vowel "i" is not so tensed. (Compare "tibia" with "Boethius".)

--Gheuf 04:34, 30 January 2007 (UTC)

Further comments
There was and is no great urgency with regard to this key -- which is now largely supplemented by IPA -- but I feel that a few comments are in order.

First, the concept of having an agreed-on phonemic spelling system to explain unfamiliar names is in principle sound. However, there are good and bad executions of such a system: the worst are those which require just about as much study of the system as other strictly phonetic transcription systems, like the IPA. This one is moderately badly executed: the spellings frequently suggest a pronunciation quite at odds with the one they are intended to represent. For instance:


 * 1) The distinction between ar and arr, ir and irr, or and orr, er and err -- in these contexts -- is far from intuitive.
 * 2) The spelling oe may to some readers suggest the sound of oe in Goethe
 * 3) The spelling ye may suggest the sound of the word "ye", or else
 * 4) ər is used to represent two sounds, while both ər and er can represent the same sound
 * 5) oo represents two distinct sounds
 * (This could be easily remedied by using 'uu' instead, as some dictionaries do.)  —Preceding unsigned comment added by 74.190.180.191 (talk) 05:57, 18 January 2010 (UTC)


 * 1) or represents two distinct sounds (and the examples given, war and wore are distinct even in dialects that don't have the north/force distinction).

The system frequently doesn't reflect the phonemic unity of certain sounds. Moreover, while originally intended to represent the sounds of classical (especially astronomical) names, it includes a host of sounds that never appear in classical names: as follows

Vowel symbols
Now, to break this down: leaving aside the sounds that never appear in classical names, we have:
 * Simple short vowels a, e, i, o, u
 * r-colored short vowels:
 * tautosyllabic: ar, er, (ir), or, (ur) -- though tautosyllabic er and ir have not been distinguished in English for close to 400 years, and ur is no longer distinct from er/ir in major dialects of English
 * heterosyllabic: arr, err, irr, orr, urr -- given that we have symbols for the short vowels, and these occur across a syllable boundary, it's hard to see how these are necessary. (There's a case to be made for the dialectal American variation of arr, but this is of course a completely automatic substitution)
 * Simple "long" vowels and diphthongs ay, ee, ye, oe, ew, aw
 * r-colored long vowels: air, eer -- other long vowels+r, all of which exist and have some notable dialectal variations (particularly insertion of epenthetic [ə] before [ɹ]) -- omitted for some inexplicable reason, though of course if they're not necessary, then neither are air and eer.
 * Reduced vowels: a or ə and r-colored ər (with variants er, ir, ur). There are two British variants of ər, but these are completely predictable by position.  The point of such a system should be to get the user of the system to employ to the fullest his intuitive knowledge of his own language, and not to try to impose a particular pronunciation.

Also missing from this list is one of the most common sounds in Classical names, which varies by dialect between ee and i, as in Kore, Prometheus, Harmonia. Substituting ee or i is again, an imposition of a particular sound-reading.

For classical names, especially, it seems appropriate to use the old grade-school system of "short and long vowels" distinguished by macrons -- which, after all, is derived from the old English pronunciation of Latin. Adding the symbols ə (oversimplified as it is) and some other symbol (say y) to represent the variable i/e sound, would cover all of the possibilities in Classical names, granted the allophonic distribution before r. Also desirable is a syllable separator -- for which I think a period looks better than the hyphen.

Thus the names of the planets could be respelled as follows:
 * mer.kū.ry, vē.nus, erth, marz, jū.pi.ter, sa.tern, ū.rə.nus, nep.tūn (Mercury, Saturn, Neptune and of course Earth are not strictly classical in form)
 * Moons: ī.ō, ū.rō.pə, ga.ni.mēd, kə.lis.tō, tī.tən etc.
 * Asteroids: sē.rēz, pa.ləs, jū.nō, ves.tə, as.trē.ə etc.
 * Constellations: sen.tau.rus, skor.py.us, pī.sēz, vul.pe.kū.lə etc.

Stress could of course be marked in a number of ways. And of course this system, like the previous one, requires some mastery; but I think it is not quite as arbitrary (and, at times, bizarre) in structure as the current one. RandomCritic 00:35, 24 April 2007 (UTC)


 * RC, I didn't see this till just now. There is an EnPR system similar to this, and discussion on making it more phonemic. (Currently it's just AHD and some of the rhotic vowels are idiosyncratic: air for example is not written with the same vowel as made.) kwami (talk) 21:03, 8 February 2009 (UTC)

I consider my self to be an educated (college degree), well-read, detail-oriented person and most Americans (United States), I agree, would not know what IPA is. I even have trouble with it and I READ the dictionary. Also, the IPA pronunciations do not display on the computers I have used. What is wrong with spelling phonetically (in parentheses) in an entry? The IPA symbols look like gobbledygook to us. Secondly, on the topic of syllable stress, I routinely see the stressed syllable printed in all caps, such that "America" would be rendered

(e-MARE-i-ke) [that is, the vowels being: schwa - long "a" - short "i" - schwa]

75.179.5.126 19:52, 25 April 2007 (UTC)WordPedia

OR?
If this is based on a system published somewhere, then there are bits that are new here? That would be OR and against Wikipedia policy, wouldn't it? --Kjoonlee 16:41, 20 September 2007 (UTC)
 * No, because it is a formatting key, that is, a convention, not an article. kwami 20:29, 20 September 2007 (UTC)
 * But it's used in at least some articles, so that would be OR in articles, wouldn't it? --Kjoonlee 21:06, 20 September 2007 (UTC)
 * It's still just a convention. We have certain arbitrary formatting conventions we use, but none of them are considered OR. Adding an unsourced pronunciation would be OR, how you indicate it is a convention. It doesn't make any claims about substance. kwami 20:39, 22 September 2007 (UTC)
 * That's evading the point. Those conventions can be sourced, in one way or another. --Kjoonlee 20:56, 22 September 2007 (UTC)
 * No, it is the point. It doesn't matter whether or not we can source the layout of Wikipedia. It's just a convention. What of coloring maps? If we assign countries or languages different colors, should that be flagged as OR unless we can provide a source that maps them with exactly the same colors? It doesn't matter, because the colors we assign don't affect the content of the articles. kwami 21:10, 22 September 2007 (UTC)
 * Bad analogy. We use IPA and SAMPA (to some extent) because they're sourced. --Kjoonlee 21:42, 22 September 2007 (UTC)
 * To make another analogy, layout styles and editing conventions are like suprasegmentals, but transcription conventions are like segmentals: you can't use OR for those. --Kjoonlee 04:52, 25 September 2007 (UTC)
 * I don't understand why OR would be okay for suprasegmentals, nor why transcription conventions should be equated with one, but layout conventions with the other.
 * It's standard for dictionaries and encyclopedias to come up with their own transcription system, so there's nothing out of the ordinary in Wikipedia doing the same, though of course most are moving to the IPA. In any case, a convention isn't research, any more than chosing your user name, so by definition it can't be original research. kwami 05:08, 25 September 2007 (UTC)
 * Transcription conventions lead to transcriptions. Hence the OR connection is valid. --Kjoonlee 17:02, 28 September 2007 (UTC)
 * And formatting conventions lead to formatting. Should that be disallowed as well? kwami 00:34, 29 September 2007 (UTC)
 * Bad analogy. Formatting conventions don't affect content. --Kjoonlee 01:06, 29 September 2007 (UTC)
 * Neither do transcription conventions. I really don't understand the distinction you keep trying to make. kwami 01:10, 29 September 2007 (UTC)

That is demonstratably false. Just look at Talk:Ötzi the Iceman to see evidence of conventions affecting content. --Kjoonlee 01:27, 29 September 2007 (UTC)
 * And your example is a straw man. This transcription was designed exclusively for English. It obviously won't work for non-English phonetics, any more than the IPA will work for languages with sounds that fall outside the IPA. kwami 04:45, 29 September 2007 (UTC)
 * You're taking my example out of context. I'm not saying that it's bad, I'm saying it affects content. You can't deny that. --Kjoonlee 12:43, 29 September 2007 (UTC)

Examples
Some examples would be good. --Abdull 21:04, 1 November 2007 (UTC)
 * I was just about to say that. If anyone is interested gave me this example: prə-NUN-see-AY-shən. +Hexagon1 (t) 04:53, 8 January 2008 (UTC)


 * Just for fun I did some exercises to see how this turns out:

Not really all that intuitive! &minus;Woodstone (talk) 12:43, 30 September 2008 (UTC)


 * You're right, it is certainly not intuitive for people who aren't used to this format. It's meant for the people who can't get their heads around the IPA, who are mostly Usonians and are familiar with this kind of thing. kwami (talk) 18:53, 30 September 2008 (UTC)


 * This convention has been used on the moon and asteroid articles for a couple years now, and the only real complaint has been that we should restrict ourselves to the IPA, not that we should modify this to something more intuitive. BTW, it becomes more intuitive with longer words. I would normally use a rhyme of homonym for short words like your examples: worse, rhymes with purse, etc. There's nothing in the MOS that we should use this convention, only that we can, which is as it should be. kwami (talk) 19:05, 30 September 2008 (UTC)


 * Yikes.


 * Shouldn't that be wyen and hyed, following the “after a consonant” direction? —Michael Z. 2008-09-30 20:42 z 
 * Indeed, now modified in list. Makes it even worse. &minus;Woodstone (talk) 07:49, 2 October 2008 (UTC)
 * Yeah. Sometimes I think we should have ss at the ends of words. A little leeway is a good thing, to avoid homographs with different pronunciations. (You don't want to respell as THIS, for example.) kwami (talk) 06:50, 2 October 2008 (UTC)

Isn't there a balance between IPA, which can only be read by people who are trained in linguistics and therefore probably know how to pronounce most words already, and this, which gets words completely wrong? Normal people would read "wyen" as "wai-en" or "wee-en". --86.168.108.98 (talk) 10:27, 18 December 2008 (UTC)


 * As an aside, that's not really true of IPA. The subset of IPA for English is used in dictionaries just like the respelling systems many of us used in grade school.  And I have no linguistics training at all, but I've learned to read IPA for a number of other languages. —Michael Z. 2009-01-17 16:42 z 

I'd favour deleting this page and abandoning all instances of it in use. I just saw the following"
 * Makemake ( or    and formally designated (136472) Makemake), is the third-largest known dwarf planet in the Solar System and one of the two largest Kuiper belt objects (KBO) in the classical KBO population.

and I have to say I think it is both intrusive and unnecessary for us to have proper IPA alongside Berlitz-style respellings. I also think that people here "inventing" the convention are engaged in Original Research. -- Evertype·✆ 09:09, 4 February 2009 (UTC)


 * It's not research. But I agree that it can be rather intrusive to have lots of pronunciation hints. Prob'ly best to put them in a footnote. A lot of Americans are simply hopeless with the IPA. It's like using imperial units. We can't really expect someone to learn the IPA just to look up stuff on Wikipedia. Another possibility would be an American-dictionary style transcription, like the Compact Oxford before it went IPA, or American Heritage (though hopefully more phonemic and browser friendly). kwami (talk) 11:36, 4 February 2009 (UTC)
 * Devising new conventions is research. In any case I don't feel that catering to the lowest common denominator is the right thing to do. We should use IPA throughout, and give tools to these Americans you feel disadvantaged to work out the IPA. -- Evertype·✆ 17:21, 4 February 2009 (UTC)


 * The problem with Makemake can be solved by moving the pronunciation down to its own short paragraph. Many biographies and articles about places have similar problems because of transliteration and pronunciation info, sometimes for three or four languages. But creating a novel system is not something we're qualified or mandated to do.  It is against the spirit of WP:OR, if not the word.  Does anyone have access to a 1913 edition of Webster's?  Whatever pronunciation system is used there is totally public domain, and it certainly has the credentials as an authoritative American dictionary. (I've seen the 1859, and that early system is not suitable) —Michael Z. 2009-02-04 18:20 z 
 * Begs the question: Why should we need to have more than one phonetic transcription system for the Wikipedia? The encyclopaedia is internationalist; IPA is unambiguous; IPA isn't that difficult. -- Evertype·✆ 18:44, 4 February 2009 (UTC)
 * Hear, hear! &minus;Woodstone (talk) 20:35, 4 February 2009 (UTC)


 * Why do we use imperial units in Wikipedia? Metric isn't that hard. Why do we use AD vs BC? We accept several provincial conventions to make this place more accommodating. kwami (talk) 22:07, 4 February 2009 (UTC)


 * Well, I prefer IPA and believe it is better, too, but we know this view isn't universal. I also think that the readers who may not benefit from IPA might be underrepresented in these talk pages.  Appropriate responses include:


 * Encourage IPA, help people use it, make it as easy as possible.
 * Accommodate additional transcription methods, without detracting from articles.


 * For no. 1, perhaps we need a handier or simpler reference than Help:IPA for English. For no. 2, perhaps we need a transcription preference, or some sort of tighter integration with Wiktionary. But energy is better spent on improving the usefulness and accessibility of IPA. —Michael Z. 2009-02-04 22:15 z 
 * Well, Kwami, it's hard to see what you're advocating. Or why you're advocating it. Are you proposing that in every instance where we have IPA we shove in one of these Berlitz transcriptions? For my part over the years I have been here I can recall numerous occasions where transcriptions were replaced by IPA, indeed where Talk pages requested it. The Berliz transcriptions are imprecise, and they are unsightly with their capital letters SHOUTING stress. We should set the bar higher on the Wikipedia. We should as Mzajac says help users to understand IPA. After all people come to the Wikipedia to learn things. Frankly I think the Berlitz transcriptions condescend to these Americans you say they help. -- Evertype·✆ 09:12, 5 February 2009 (UTC)


 * No, it's not condescending when Americans tend to think the rest of the world should do things their way, not the other way around. And no, I wouldn't want to see the respellings everywhere. I don't want to see imperial units everywhere either, and I wish we'd ditch AD/BC too. But a lot of our readers understand neither metric nor the IPA, and telling them not to use Wikipedia until they figure them out is basically telling them not to use Wikipedia. kwami (talk) 10:48, 5 February 2009 (UTC)
 * You speak for Americans? I was born and raised in America before I emigrated to Ireland. I learnt IPA. (And was familiar with the other systems starting with Thorndike-Barnhardt as a kid.) -- Evertype·✆ 00:38, 8 February 2009 (UTC)
 * Yes, I speak for Americans. Michael, you obviously aren't the typical American when it comes to knowledge of scripts! Are we to use you as the standard? :) I doubt 1% of Americans are familiar with the IPA. I'm not sure 1% have even heard of it. You'd be hard pressed to find a native-born American who hasn't lived abroad or taken a college-level linguistics class and is familiar with it. Nearly every American, on the other hand, is familiar with respellings and AHD-type transcriptions. Asking people to learn a system they often find confusing in order to make full use of Wikipedia is basically shutting that part of the encyclopedia off to most of them. Most are casual users who simply aren't going to bother. kwami (talk) 00:54, 8 February 2009 (UTC)
 * Do you, indeed? "They" asked you to devise a redundant, parallel yet inaccurate system to festoon the English Wikipedia with? You do a lot of good work. I don't think this is some of it. -- Evertype·✆ 00:49, 9 February 2009 (UTC)


 * Yes. Some have, anyway. Quite a few Americans here complain that the IPA is unintelligible. I've also taught intro (and not so intro) linguistics courses in the US, and AFAIK I haven't had a single native-born student who knew the IPA who wasn't already into languages or linguistics. Those taking the course to get a breadth requirement out of the way are almost universally clueless, and sometimes really struggle. US exposure is nearly zero. A university education in the US does not mean it's even likely that a person knows the IPA; without a university education or foreign language study, the chances are very low. kwami (talk) 01:30, 9 February 2009 (UTC)


 * I'm no American, but a Canadian. My paper dictionary uses IPA (the Canadian dictionary, the Canadian Oxford Dictionary).


 * But no person is a standard. I challenge the assertion that even 1% of anybody has “learned” one of the dozen or more “AHD-type transcriptions”, much less all of them.  (Is this a “type” of transcription?  They're functionally identical to the phonemic IPA used in my dictionary.  The only thing which characterizes them is that no two of these proprietary systems use the same set of symbols, while the IPA is an open standard.)


 * Discussion on this topic is rife with mythology. —Michael Z. 2009-02-08 15:37 z 


 * Yes, Canada is different. It's specifically Usonians who are this insular, as we are with imperial units instead of metric. (I mean, there are Usonians who don't even know Canada has its own currency!) True, there is no single AHD-type standard, but then there is no IPA standard either. "/i/" can mean either or, for example. (Which is the reason we don't use "/i/", except at the end of city, where it doesn't matter.) But just as someone familiar with the IPA can quickly adjust to the version we use here, someone familiar with AHD, Random House, or Webster's (the vast majority of the literate US population) or even who just went to a US elementary school will be able to quickly adjust to the version of AHD/RH that we use here, since the differences are minor. kwami (talk) 01:30, 9 February 2009 (UTC)


 * Yeah, but IPA had the same status in Canada as in the US before 1998. That's when it showed up  in the Canadian Oxford Dictionary, 1st edition.  I haven't heard anyone complain.


 * I'd wager that a significant part of our readers' resistance to IPA is being confronted with foreign-looking characters, and that once they got past this, most who actually use IPA wouldn't work significantly harder at it than using the dictionary systems (I would also assert that practically everyone would use it with the reference, and not have actually memorized our EnPR system. With less confidence, I would cheekily speculate that many complainers object to the appearance of IPA, without actually trying to use it).  Of course we can't test this, and I believe it serves the readers to respect their needs anyway. This is also why I think lowering the threshold to IPA adoption is important. —Michael Z. 2009-02-09 02:35 z 


 * "practically everyone would use it with the reference": That's exactly one of the complaints I've heard, that they need to look up every single character individually, which they wouldn't need to do with the AHD, since they already have that or a similar system memorized. ("Can't you guys use normal dictionary pronunciations?") That, and that they can't figure out how the key is organized. If you don't know whether some squiggle is supposed to be a consonant or a vowel, it can take a long time to find it. (I mean, "c" is a consonant, but turn it over and it's a vowel. And how is anyone supposed to know which the numeral "3" (ezh) is?) But with a Latin-based system, you at least know that a-whichever diacritic is a vowel that is often spelled "a". It really is much more intuitive if the only thing you know about phonology is what you've been able to extract subconsciously from English orthography. kwami (talk) 02:54, 9 February 2009 (UTC)


 * Point taken, although the broken o and script zed weren't hard for me. I actually found much of IPA rather intuitive, but it is definitely quite different. —Michael Z. 2009-02-09 05:00 z 


 * And also has very little connection to English orthography. I also find it rather intuitive, but I suppose that's what happens when you know something well. Actually, though, isn't a script zed, but historically a script gee. Or maybe a script gee (yogh) modified by the IPA by analogy with zed. kwami (talk) 07:14, 9 February 2009 (UTC)


 * I think Michael [Michael Z?] got it spot on. Pretty much anyone who's participating in discussion on WP talk pages relating to pronunciation transcription will be someone who has an understanding of IPA.  If having non-IPA transcriptions clutters articles and offends your eye, then perhaps we could have some sort of a button or expansion text after the initial entry that, when pressed, expands to give pronunciation details, and then can be collapsed.  I'm no HTML or wiki markup expert, but surely there is a way to hide the information but then expose it if the reader wishes to see it.  The OED has something like this, no, where you press a "pronunciation" button and then pronunciation information appears?Atemperman (talk) 18:08, 7 February 2009 (UTC)


 * Yeah, Dict.com lets you toggle between IPA and respelling. We could do s.t. like that, or (maybe better) between IPA and mod-AHD. We could prob'ly automate it, so that we could just enter the IPA and the conversion would happen in the coding, the way we do now for Polish, so the two transcriptions don't get out of whack. (We'd probably need to re-enter each IPA phoneme in a separate slot, though, like we do for Polish IPA.)


 * As far as coordinating with Wictionary, they still can't decide which system they're using. They continually modify their transcription keys without updating the entries, so that the keys don't always correspond to what the reader sees. There's evidently a lot on people's private pages that is just assumed to be Wiktionary standard without it ever being made available to the reader. We need something more stable than Wiktionary. kwami (talk) 23:04, 7 February 2009 (UTC)
 * May I suggest that we use the IPA? -- Evertype·✆ 00:35, 8 February 2009 (UTC)


 * Of course we use the IPA. This is about making it more accessible by offering alternatives. kwami (talk) 00:54, 8 February 2009 (UTC)

I don't know if this has been made clear earlier on, but obviously the respelling alternative would only be applicable for English words. Trying to use the same respelling for other languages would create absolute havoc. &minus;Woodstone (talk) 09:42, 8 February 2009 (UTC)


 * I would hope that's clear from the intro to the key. kwami (talk) 10:03, 8 February 2009 (UTC)
 * I'm minded to go on a nice crusade deleting "respellings" and replacing them with accurate IPA. -- Evertype·✆ 00:49, 9 February 2009 (UTC)


 * Please do, if any are there without IPA. That is our MOS. But deleting them just to delete them would be like deleting imperial units where they're already secondary to metric. kwami (talk) 01:34, 9 February 2009 (UTC)

Phonemic IPA chart
This may be off-topic here, but IPA might be more accessible if there were a chart, like the one shown in the margin of most dictionaries. It could be placed in the margin of an article which uses a lot of IPA. It can also be linked directly, offering just a simple chart, and not a screen-full of explanation like WP:IPA for English.

Needs refinement. I linked the diphthongs to Diphthong and the nasals to Nasal vowel – is this optimal? —Michael Z. 2009-02-08 16:45 z 


 * Sounds like a good idea to have this available at least. Some minor improvements might be:
 * use the same more or less alphabetical ordering as in WP:IPA for English
 * remove the foreign sounds from this concise key
 * use the shortest possible words to save space (beige for [ʒ], sing for [ŋ])
 * for vowels try to find words containing a letter similar to the IPA symbol (e.g. piece for [i])
 * for diphthongs use open syllables only (go for [oʊ])
 * use only non-rhotic words as example (raw for [ɔ])
 * &minus;Woodstone (talk) 17:15, 8 February 2009 (UTC)

It looks good. I like it, though I'm not sure how many articles it would be appropriate for. As IPA for English, I think the foreign sounds should go. They're rare, and most people I know are only going to use English phonemes for them anyway. (Do we even have entries which use them?) And since we don't use a phonemic transcription for our rhotic vowels, they need to be listed separately.

Hey, another possibility would be to have a collapsible chart (with a shorter title) that is automatically introduced with the IPA-en or pron-en template, which opens up like this when clicked. kwami (talk) 20:37, 8 February 2009 (UTC)


 * The sounds from foreign languages are all used in English. A major reason for giving transcriptions is to show unfamiliar English words and English pronunciation of foreign proper names.  Perhaps a better example can be found for schön, but the remainder are used in English, at least regionally. —Michael Z. 2009-02-08 23:05 z 

Since we're giving alts here, I'll throw in a US dictionary system, like the one e.g. Random Critic proposed above.

(asterisks) We don't generally need to bother with "primary" stress or syllable breaks. The IPA in parentheses is the Wikipedia IPA-en system where it isn't phonemic. kwami (talk) 21:23, 8 February 2009 (UTC)

Random Critic proposed  for (I believe) the final vowel of city. This is useful in words which vary between a syllabic and non-syllabic i, but I haven't (yet) added it to my proposal. kwami (talk) 21:34, 8 February 2009 (UTC)

Hey guys, I wonder if you'd mind editing a copy of the chart in my user space. I don't agree with every change that's been made to it. For example, I believe that neither bought nor raw is pronounced with an /ɔ/ in Canadian English. It's kind of hard to compare or tell what was changed, because content changes were made at the same time it was re-ordered.

I'm going to revert for now. —Michael Z. 2009-02-08 23:01 z 


 * But this isn't just for Canadian English, is it? Caught is the standard word. (It doesn't have that vowel in my dialect either, but that's not relevant.) Rhotic vowels need to be listed separately. Also, you forgot the length signs on  and, which do not occur before r. There should probably also be a length sign on  deux. And is that really phonemically distinct from ? ( might be better there, because if people ever italicize the IPA,  will become conflated with .) Also, glottal stop should go under 'regional', unless you choose a word like uh-oh that has it elsewhere. kwami (talk) 23:21, 8 February 2009 (UTC)


 * Ideally, we would choose a word with that sound everywhere, or as widely as possible. Otherwise, for example, Canadians will misinterpret the key.  I'm not sure what being the standard word means, but caught is not a universal word, or close to it.


 * Which rhotic vowels? Can they be represented with something like /ər/ or /ə(r)/?  (I know, I should stick to the chart and discuss representation elsewhere.) —Michael Z. 2009-02-09 00:00 z 


 * Further talk about refining details of the chart at User talk:Mzajac/IPA reference. —Michael Z. 2009-02-09 00:27 z 

Missing sounds
How should I represent a word-final /i/ as in happy?


 * Generally systems like this don't distinguish it from ee.

And what about /ʊər/ as in pure, for some people? Gailtb (talk) 10:02, 16 April 2009 (UTC)


 * oor would work. The distinction between oo and oo is lost before r. kwami (talk) 23:18, 16 April 2009 (UTC)

What's wrong with IPA?
This method is a joke. Pronunciation of different words varies by person and location. 'or' or 'aw' or almost any other combination of letters used here to make a sound has many possible pronunciations. IPA has only one pronunciation, it can also be used for foreign words and can indicate differences between regional dialects. Plus IPA can help the speaker with vowel length (long or short vowels), uses an apostrophe as standard to signify word stress and is accessable for English speakers and speakers of any language (IPA is IPA for whatever language is used). This page seems like an American attempt to simplify something which doesn't need simplifying - get off your fat asses and learn IPA.--217.203.154.2 (talk) 01:26, 24 May 2009 (UTC)


 * A couple reasons to keep it:
 * Tens of millions of people do not know the IPA, and they aren't going to learn it just to use Wikipedia. Accessibility is a good thing. And the American sitting on his fat ass who pushed for this actually knows the IPA quite well, and has pushed for primary universal coverage in the IPA as well.
 * Surprising numbers of people from countries that use the IPA are also pretty clueless. Often when I clean up people's garbled IPA transcription, the only way I can make heads or tails of it is the respellings they added along with it. Having two competing transcriptions acts as an error check on our editors: when the two don't agree and the editor doesn't realize it, we know something is wrong.
 * And no, the IPA doesn't have "only one pronunciation". /r/ in English, Spanish, and French are entirely different sounds, for example, as more subtly are English and Spanish /t/ and /d/. And no, the IPA doesn't use apostrophes. It's a real pain to clean up after people who think it does, and don't bother to use the real IPA in their transcriptions. kwami (talk) 04:10, 24 May 2009 (UTC)
 * That's because lazy people don't know how to use IPA. Spanish has /r/ and /ɾ/, English has /ɹ/ and /ɻ/, and French has /ʀ/ or /ʁ/ depending on the dialect. No apostrophes to mark stress? What do you call this then? Also, the English t and d are /tʰ/ and /d/ while the Spanish t and d are /t̪ / and /d̪  /. The reason for writing /t/ and /d/ instead of the more complex forms is because IPA transcriptions in Wikipedia are in most cases meant to be phonemic, not phonetic. By the way, for those who say that "I need to go back and forth to decipher the pronunciation", well, how about you try to learn at least one other foreign language besides your native English, so that you'll familiarize yourself with writing phonetically like real languages do (Latin, Finnish, Serbian, Turkish, Spanish, Italian, etc.)? -- nlitement  [talk] 19:17, 7 June 2009 (UTC)
 * I thought you were serious, but apparently you're just trolling. What, English, Japanese, Gaelic, French, Arabic, Chinese, Thai, and Hebrew aren't "real" languages? (Also, of the languages you cite, none of them actually write phonetically. Most don't even write phonemically.) You also apparently don't realize that your explanation for why English and Spanish /t/ are written the same is also the explanation for why English and Spanish /r/ are written the same. And this is not an apostrophe. —kwami (talk) 20:59, 7 June 2009 (UTC)


 * Because 99% of people reading wikipedia don't know and don't care about IPA. I certainly won't go back and forth between the IPA article to decipher the pronunciation letter by letter. Forcing IPA on people when better solutions (i.e. this) exist is idiotic at best. —Preceding unsigned comment added by 68.40.177.113 (talk) 01:37, 29 May 2009 (UTC)


 * Don't project yourself onto other people. People who know how to use a dictionary generally know the IPA. Yours is an argument for abandoning the metric system because "99% of people" don't know what a meter is. kwami (talk) 02:31, 29 May 2009 (UTC)
 * It's also worth pointing out that IPA is supported by a number of people who don't understand it. For example, I don't personally, I've never bothered to learn but do understand why it is superior whatever flaws it may have. Nil Einne (talk) 17:11, 27 December 2009 (UTC)

I could possibly understand adopting another well-known standard in parallel, but coming up with a new proprietary one is simply ridiculous. Let's get rid of this please. 87.23.5.174 (talk) 09:48, 20 December 2009 (UTC)
 * What well-known standard would you suggest? — Æµ§œš¹  [aɪm ˈfɻɛ̃ⁿdˡi]  18:24, 21 December 2009 (UTC)

I add this edit simply to support the view that a simple system using no, or few, special characters and requiring neither an effort to learn nor reference elsewhere to use will allow more people to more easily read and benefit from wiki articles. Ceratinly this is so in my own case and I believe it will be so for a great many others. After having found only IPA pronunciation guides on other wiki pages on many occassions and found it a trouble to extract the information I desired I was delighted to find the non-IPA pronunciation guide on a page I was looking at just now. It was the degree of that delight that led me to this discussion page to record that fact. I do think there are people who believe that 'we ought' to be using the IPA [only] and who see that as a higher imperative than usability. I commend the alternative view. —Preceding unsigned comment added by 217.44.127.79 (talk) 22:28, 15 February 2010 (UTC)

Vowel of "put" and "foot" seems to be missing
Or have I overlooked it? Grover cleveland (talk) 15:40, 9 June 2009 (UTC)


 * You have :-). It's "oo". — Cheers, Jack Lee  –talk– 16:28, 9 June 2009 (UTC)

Name
In lexicography, respelling is the conventional name for the kind of system still used in US dictionaries. —Michael Z. 2009-08-09 16:23 z 

Weiss
For some reason I looked at at the Rachel Weiss article, which correctly gives her surname in IPA as /vaɪs/. Applying the respelling principles here, we end up with "vyes" or "vyess". Both are in my opinion, misleading. "vyes" suggests to me a voiced final consonant /vaɪz/, while "vyess" (or indeed "veyess") suggests a bisyllabic word /vaɪ.ɛs/. The obvious, unambiguous phonetic respelling is "vice". Since the whole point of the pronunciation respelling is to be intuitive and unambiguous, something is very wrong here. Grover cleveland (talk) 00:56, 26 August 2009 (UTC)


 * If it were two syllables, it would be written as two syllables. That's why we have a key: any system is ambiguous if you don't understand what the symbols mean. (This system works fairly well for polysyllabic words, not so well for monosyllabic words.) Anyway, if it doesn't work in this case, you're free to delete it. Adding "vice" would also be acceptable, just don't link it here. kwami (talk) 05:02, 26 August 2009 (UTC)


 * The page should be amended to allow a respelling as "vice" in order to fit typical notions of phonetic spellings. The page should not attempt to subset the English-language conventions generally followed for the past 3 centuries. -Wikid77 (talk) 14:53, 26 August 2009 (UTC)


 * There are a couple cases where I've left spellings such as double-s for /s/, plus recently eye-o-dyne rather than eye-o-dyen for iodine. Perhaps that is a respelling we should officially include? Another counter-intuitive result is koh-bolt for cobalt, which I don't see a way around: I'm afraid any time a resulting syllable is an orthographic word in English, we're likely to have such problems, given the irregular sound-letter correspondence of English. kwami (talk) 19:33, 30 September 2009 (UTC)

Added: Rationale for respelled pronunciations
To help explain why many people can better understand the respelled, phonetic pronunciations, I have added the text below:


 * "The use of spelled-out, phonetic pronunciations has continued for over 3 centuries in English-language culture, and hence, many people can be expected to more easily grasp the sounds of words as denoted by respelling as simple syllables. Although the IPA phonetic alphabet allows for a greater range of sounds, it comes across as arcane, peculiar "nerd-speak" (unintelligible technical text) to many millions of readers. Consequently, the respelled format should be used to reach the vast majority of general readers, as typical during the past 3 centuries. However, for some words originating in other languages, the IPA format might be the only viable method to denote specific sounds of the preferred pronunciations. To limit confusion, the IPA format or multiple, regional pronunciations could be listed in a footnote, to avoid cluttering the upper text of a page.

As noted above, many people, already familiar with basic English words (such as "cat" or "ginger" or "talk"), can easily interpret the respelled format, with hyphenated "syl-la-bles" as noted with the BBC usage. Instantly, they can tell that "vyess" will be one syllable (similar to "vice"), so that the name "Hugh" could be respelled as "hyu" or "hue" while Beethoven becomes "Bay-tow-ven" or such. Confusion about syllables is a major, confusing problem introduced only in the IPA format, and thus the implicit warping. A leading "h" sound could be indicated as "hwaht" for a regional accent dwelling on the word in "what a dump". -Wikid77 (talk) 14:53, 26 August 2009 (UTC)
 * I do not see how IPA is confusing wrt. syllabification (IME most complaints focus on the non-standard symbols)? Anyway, if desired, syllable boundaries CAN be explicitly mark'd in IPA, using a period. -- Trɔpʏliʊm • blah 17:12, 26 August 2009 (UTC)


 * Also, I've come across editors who insist on syllabifying English as if it were Latin, so it's probably best to leave syllable boundaries out entirely in most cases. kwami (talk) 20:33, 26 August 2009 (UTC)


 * I think the idea of stating the rationale for respelled pronunciation is fine, but some of the statements in the text you've inserted may be too prescriptive. You say that "the respelled format should be used to reach the vast majority of general readers" and that "[t]o limit confusion, the IPA format ... could be listed in a footnote...". My impression is that there are editors who believe that IPA is superior to pronunciation respellings, and who will not agree to any text that promotes respelled pronunciation over IPA or that tries to relegate IPA to footnotes. It may be better to rephrase the text so that it simply suggests that respelled pronunciation and IPA are equally acceptable except when the scheme at Pronunciation respelling key does not accurately reflect the pronunciation of foreign terms. — Cheers, Jack Lee  –talk– 10:42, 27 August 2009 (UTC)

I deleted that section. Although I created this key to give extra flexibility to editors, it is provincial; it's unintelligible to a great number of native English speakers and almost all non-native speakers. It also fails to make some distinctions which are phonemic in RP and other major dialects. Therefore it is this system which should be secondary to the IPA and perhaps relegated to a footnote. Americans may have trouble with the IPA, but then Americans don't know what a kilometer is either—that doesn't mean WP should use Imperial units with metric added in a footnote. kwami (talk) 11:13, 27 August 2009 (UTC)


 * Fair enough. :-) — Cheers, Jack Lee  –talk– 14:54, 27 August 2009 (UTC)

How to use this key in articles
What is the best way to enter text? I have been manually applying the "small" tag to stressed syllables, but there seems to be a better way that I haven't quite figured out. It's something like this: {respell|unstressed|STRESSED} but that doesn't seem to work properly - I recently found the following at Citizendium: {Respell|SI|tə|ZEN|DEE|əm}, yielding, which has the wrong font sizes. (Also "dee" shouldn't be in caps but we can ignore that.)

I'm sure I'm missing something simple here! Lfh (talk) 19:08, 6 December 2009 (UTC)


 * The instructions for the Respell template are on that page. kwami (talk) 04:13, 7 December 2009 (UTC)


 * Got it. Thank you. Lfh (talk) 09:58, 7 December 2009 (UTC)

Displaying the schwa
I noticed some PCs with IE can display the schwa [ə] in IPA, but not in these respellings. I assume that's because the IPA templates are designed to ensure correct display. Could that be included in the template? Lfh (talk) 17:12, 21 December 2009 (UTC)

Hang on, I'll copy this to Template talk:Respell, that's a better place. Lfh (talk) 17:14, 21 December 2009 (UTC)

f
A suggested tweak - ff to be allowed for /f/, similarly to ss for /s/ - it emphasises that the consonant is not /v/ (as in "of"), and looks more natural after short vowels ("buff" vs. "buf"). (And some articles are using it already.) Lfh (talk) 21:35, 16 January 2010 (UTC)
 * That works in the syllable coda, though wouldn't an initial double f look weird? ("fford")? Or are you not suggesting that? — Æµ§œš¹  [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  22:38, 16 January 2010 (UTC)
 * I'm suggesting both be allowed, like for /s/ - we can choose between s and ss, right? Lfh (talk) 11:13, 17 January 2010 (UTC)

Concerns
I am concerned that this "respelling guide" constitutes original research, and does not live up to the claim embodied in This is a pronunciation respelling key used in some Wikipedia articles to unambiguously spell out the pronunciations of English words. It worries me to see this over-simplified and misleading pronunciation guide being added to multiple articles. Was a consensus ever established to use it in this way? If so, where? --John (talk) 15:57, 27 January 2010 (UTC)
 * I believe that WP:NOR applies only to articles and images, and not to help pages and policies. At least that's the impression I get when reading the policy. — Cheers, Jack Lee  –talk– 18:59, 27 January 2010 (UTC)
 * Sure. But these respellings are in the articles, and therefore could breach NOR, according to your understanding and mine. Unless there is some sort of process here, a consensus, I don't think we should be using them. --John (talk) 09:48, 30 January 2010 (UTC)
 * I didn't think that consensus was an excuse OR. But using this for the IPA or whatever in a dictionary is no more OR than converting numbers to Roman numerals. It's simply a matter of formatting. kwami (talk) 11:12, 30 January 2010 (UTC)

Misleading in what way? Lfh (talk) 13:47, 30 January 2010 (UTC)
 * Take barium, for example. I have been a chemistry teacher for many years. I come from Scotland originally. The respelling BAIR-ee-əm does not reflect the way I say the word. The second vowel is more an "i" sound than an "ee" sound. So the respelling adopted here seems to be based on an American English pronunciation. This seems to break the spirit of ENGVAR, which is why I wondered just where and how consensus was attained to add it to articles. --John (talk) 22:27, 30 January 2010 (UTC)
 * Well, any pronunciation system is only unambiguous if you read the key; it's impossible to design a system that is both internally consistent and immediately intuitive to everyone in every case. But you have a good point: the respelling "EE" is being used here for two different sounds - the long  of "seem", and the short  of "Austria - while the key only gives an example for the former.  This should be changed.


 * I assume that the short was pegged to "EE" rather than "I" because if you have so-called "happy tensing" (as I do, in England) you pronounce "y", "ey", and "ee" at the end of words as  rather than  (e.g. "happy", "Aimee").  That's my guess, anyway.


 * I don't think ENGVAR applies here, because the elements are there for everyone and have no nationality. As an Englishman I don't find the system incompatible with my own accent, but I appreciate that it may be different for Scots.


 * As for consensus - I'm afraid I don't know because I wasn't around when this system was designed. Lfh (talk) 23:12, 30 January 2010 (UTC)


 * Yes, this key assumes happy tensing, because it's based on US dictionaries, and all of them presume happy tensing. Other than that, it is compatible with RP. It is not adequate for Scottish English, but then our IPA conventions aren't compatible with Scottish English either: . The reason for that is that very few dictionaries are adequate for that; if it's not in our sources, we have no way of including it. But again, that is a problem for all English transcriptions on WP, not just this one. kwami (talk) 23:53, 30 January 2010 (UTC)

There's no schwa in sofa
The intro paragraph has a nonsense description about the schwa, "ə", which is used for the vowel at the end of sofa.

I don't know where the author learned English, but I've always pronounced sofa as so-fa. Like the so and the fa in the scale of musical notes, do-re-mi-fa-so-la-ti-do. If this page is just about how some author would spell/pronounce words in his local dialect, then this page isn't much use. Gronky (talk) 13:26, 10 March 2010 (UTC)
 * I (born in England, now live in California) have never heard anything but schwa in "sofa". May I ask where you learned English?  Would you have a schwa at the end of "comma"? Grover cleveland (talk) 17:52, 10 March 2010 (UTC)


 * Scotland. If I say "comma" quickly, yes, it's schwa'd, but if I was speaking to a foreigner (thus slowly and clearly), I'd pronounce the "a" fully.  Sofa is schwa-less for me regardless of speed.  Clearly, neither word is a reliable explanation of "schwa".  Better examples, from the schwa article, are the e in taken and the i in pencil.  Doing a few interpretations in my head, these seems more reliable/stable across accents.  Gronky (talk) 18:28, 10 March 2010 (UTC)
 * Ah, but those are then schwas in your own dialect but not most others, where they're syllabic sonorants. I don't know if pronouncing sofa as  is common for your dialect, but everyone else pronounces it with a final schwa.  — Æµ§œš¹  [aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]  19:24, 10 March 2010 (UTC)


 * Gronky, we've been unable to properly accommodate Scottish English. In the IPA key we use the same vowel for fur, fir, fern, which I assume are distinct in your speech. This is an unfortunate effect of the dictionaries we use (the OED, Webster's, etc.) which do not take Scottish into account. In this key we've set up separate transcriptions for those vowels, in an attempt to make the system more universal, but you won't often find the distinction actually made in the articles, since generally all we'd have to go on is their spelling. Different dialects have different degrees of vowel reduction too; what we have generally reflects RP, General American, and Australian. kwami (talk) 21:04, 10 March 2010 (UTC)