Talk:International Phonetic Alphabet chart for English dialects/Archive 2

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1 Archive 2 Archive 3 Archive 4 Archive 5

Length marks

The use of [ iː] and [ uː] or [ ʉː]for British and Australian variants of / i / and / u / seems to me unwarranted. Phonetician Peter Ladefoged, himself British, uses [ i ] and [ u ]. I do not believe that, in RP, the length marked with ː is a phonetic reality (both vowels are noticeably shorter than diphthongs, and not appreciably longer than the lax vowels); it is certainly not a phonological reality, as it is never distinctive. I can guess why the OED has chosen this representation, but regardless of the cause, if the actual time-length of the vowel is not significantly (say 1.5x) longer than the length lax vowels in minimal pairs produced by native speakers of RP, it doesn't really belong in representations of RP. I strongly recommend eliminating the mark ː from this representation. (I would also recommend a little more attention to the proper use of phonemic vs. phonetic brackets.) RandomCritic 07:47, 1 February 2007 (UTC)

It's not just OED; Jones, Gimson, and Wells all also use the length mark for RP. Ladefoged, although personally British, spent the majority of his working life in America and usually transcribes American English in his books, because his target audience was primarily American phonetics students. —Angr 08:42, 1 February 2007 (UTC)
I'm referring to Ladefoged's transcription of RP specifically in contrast with American English. So your explanation is mistaken, Angr.RandomCritic 19:44, 2 February 2007 (UTC)
In my British variety, I feel (without measurements, so this may be nonsense) that bit, bid and beat are all about the same length, but bead is longer. So I'm not that bothered either way about the length marks on /iː/; however, in the OED transcription of the SQUARE vowel as /ɛː/ (which makes more sense to me than the traditional diphthongal transcription), the length marks are clearly required as it contrasts with short /ɛ/: merry/Mary, and in non-rhotic speech led/laird.
Regardless of the above, there are already enough competing transcriptions of RPish varieties of British English, most of which use length marks, that Wikipedia shouldn't really be inventing a new one
--JHJ 08:59, 1 February 2007 (UTC)
It's not inventing a new transcription, it's simply choosing which one to use. Ladefoged lists (in A Course in Phonetics, 1993, p. 76), side by side, six different transcription systems (not counting the Webster's dictionary transcription). Only two of these (Gimson/Jones and Wells) use length marks, but Gimson/Jones use [ iː] contrastively with [ i ] (sc. [ ɪ ]), which at least makes phonemic sense even if it conflicts with the phonetic data; only Wells uses [ iː] to contrast with [ ɪ ]. This usage is inconsistent with the purpose of IPA: [ː], which is to express length where it is contrastive -- I hardly need to mention that length is a relative, not an absolute phenomenon. If [ i ] consistently contrasts with [ ɪ ] in quality, then there is no good reason for including the length mark. At the very least, it should be shown that the length mark is not a mandatory concomitant of [ i ] and [ u ].
By the way, are British universities now teaching their students to use slanted brackets with phones rather than phonemes? If this is an international oddity, I'd like to be aware of it. RandomCritic 20:04, 2 February 2007 (UTC)
I no longer have the third (1993) edition of A Course in Phonetics, but by the fourth (2001) edition, Ladefoged was using the doubly contrastive /iː/ and /ɪ/ (even when transcribing American English!), as do Gimson's An Introduction to the Pronunciation of English (third edition, 1980, ISBN 0-7131-6288-0), and Collins English Dictionary (fourth edition, 1998, ISBN 0-00-470453-3). Everyman's English Pronouncing Dictionary ("originally compiled by Daniel Jones, extensively revised and edited by A. C. Gimson", fourteenth edition, 1977 ISBN 0-460-03029-9) also uses the doubly contrastive /i:/ and /ɪ/. —Angr 21:02, 2 February 2007 (UTC)
You might like to read Wells's own explanation of his preferred transcription. One point he makes is that some people apparently believe that the contrast between KIT and FLEECE depends on both length and quality, and so prefer the transcription to reflect both.--JHJ 21:23, 2 February 2007 (UTC)
I don't see him making that point at all. To me he seems to attribute the usage of the length mark solely to historical considerations, i.e., as a merger between two types of transcription. At any rate, no case is advanced that speakers actually pronounce [ i ] with more duration than [ ɪ ]; a factual question, which could be easily answered by recording voices and measuring the length of the sounds in minimal pairs. Being familiar with languages that do have phonemically functional length contrasts, I can assure you that the contrast ought to be easily distinguishable. The transcription has considerable relevance to foreign-language speakers who attempt to reproduce the sounds of English based on IPA. If we are talking about a transcription intended solely for English-speakers who already know what the sounds of their language are, that's one thing, and even purely arbitrary symbols could be used (as of course they are in some systems). But IPA is supposed to be international.RandomCritic 01:57, 8 February 2007 (UTC)
Sorry, I pointed you to one Wells article, but the explicit bit I was remembering is actually in this one (PDF), on page 11.--JHJ 09:35, 8 February 2007 (UTC)
From Wells is also his MA thesis from 1961, which has a page documenting vowel duration and describes /i/ as being consistently longer than /ɩ/ (i.e. /ɪ/): his conclusion for RP is “British English has a much sharper long/short contrast, since even the longest British short vowel averages shorter in duration than the shortest long vowel [before /d/], while this is not the case in American English”. In fact, Wells gives the ratio of long:short as 1.9:1, which is well above the 1.5:1 RandomCritic wants. Can RandomCritic back up his statement that /iː/ and /uː/ are “not appreciably longer than the lax vowels”. —Felix the Cassowary 11:20, 8 February 2007 (UTC)
I can’t speak for British English, but as for Australian English, if length marks are used, they are used consistently for all monophthongal long vowels. /o:/, /ʉ:/ and /i:/ do not contrast precisely with any short vowel quality, but they are all substantially longer than the nearest short vowels and behave as long vowels. It may be that the situation is different in British English. —Felix the Cassowary 22:07, 2 February 2007 (UTC)

ditch "i or ɪ," "eɪ or e," "oʊ or o"

"i or ɪ" etc. unnecessarily confuses the structure of the table. The structure is: single IPA element on the left, multiple usage examples on the right. The point is to find usage examples for each IPA letter, not to specify all possible variant pronunciations of each word.

RP and GA are, to some extent, fictions, in that they're well defined accents where each question has a right answer. But they're still useful models. If one person is pronouncing "city" or "goat" differently from another person, then the two people are not speaking with the same accent. At least one of them is deviating, perhaps only slightly, from the standard, whatever it is.

More examples of words could surely be found where the IPA elements could not be conclusively determined. They would be bad examples and have thus been left off the chart, but "city" does not seem to be such an example, given my memory of OED pronunciations and corroborating testimony above, i.e. "happy" = RP /ˈhæpɪ/. And /e/ and /o/ by themselves as "diphthongs" in America is in fact an unusual and distinctive Northern pronunciation, not universal even in Wisconsin or Minnesota. Rent Fargo for a crash course on that accent.

So, "i or ɪ," "eɪ or e," and "oʊ or o" are structurally bad for the tables, argumentative toward the concept of standard accents, and not warranted by the facts. Three strikes.

That's the reasoning, anyway. Angr asked for some discussion. If there is a General American accent, then I think I'm a native speaker, or pretty close. 67.168.216.176 23:37, 2 February 2007 (UTC)

They're slightly different situations. The "happy" vowel was discussed above; this is a genuine change within RP. Older sources gave it as [hæpɪ], while newer sources give it as [hæpi]; this seems to reflect a change in pronunciation within RP, although both varieties are still accurately called RP. The /e(ɪ)/ and /o(ʊ)/ issue in General American is not really a difference in pronunciation but a difference in style. Some sources prefer to transcribe these vowels as diphthongs at the phonemic level, while others prefer to transcribe them as monophthongs at the phonemic level, while conceding that they have a diphthongal surface pronunciation, especially in open syllables. Since the main function of this article is to provide a quick reference guide for Wikipedia users who encounter IPA transcriptions in other articles, both transcriptions should be listed here, because both transcriptions are likely to be found in other Wikipedia articles. —Angr 00:25, 3 February 2007 (UTC)
Traditional IPA transcription for RP happy is indeed /ˈhæpɪ/, but the current OED pronunciation guide can be found here, and happy is shown as /ˈhapi/. See also Wells's article, as mentioned in the "Length marks" section. I think the evidence is that this /i/ is standard for transcriptions of present-day RP, even if its phonemic status is dubious.--JHJ 10:40, 3 February 2007 (UTC)


A Standardized Transcription for English, a hover guide

I've been working on developing a standardized transcription for English, and a hover guide that would pop up over IPA text to make reading it easier for people not familiar with the system. This scheme would complement or replace the use of this page as a reference guide for people wishing to decode simple English IPA, or to write it themselves. A justification of the principles underlying the transcription, a long with some discussion of it, can be found at Wikipedia:Pronunciation, and the transcription itself at User:Gheuf/Sandbox2. The hover guide is being worked on independently at Template:IPA hover. Let me know your thoughts.--Gheuf 06:44, 13 February 2007 (UTC)

I like the idea. The large majority of readers can't understand IPA and it's a nuisance to have to come to a page every time to check it. What confuses me more is that there are a variety of different symbols used for the same vowel sounds, and some articles even mix and mingle between them. The word I was looking for just now seems to use symbols that aren't even on here. Richard001 05:18, 16 February 2007 (UTC)
What word was that? —Angr 05:30, 16 February 2007 (UTC)

Pronunciation box

Hi all

I don't know if this has ever been tried or whatever, but I had the idea to suggest the inclusion of a pronunciation box, or some other link, to Wikipedia articles. I've seen a lot of articles with an IPA key in the first line after the subject, and it just looks ugly in an encyclopaedic article, we're not Wiktionary. There are also instances of alternative codes being given, e.g. both GenAm and RP (such as in the Mars article), which could be more neatly addressed in a separate box somewhere. I've mulled it over and it does seem like a bit of a complicated "mission" to undertake. Does anyone know if this has already been tried? Or perhaps more to the point, where a more appropriate place to make the suggestion is on Wikipedia? Kris 23:45, 20 February 2007 (UTC)

I'm not getting how this is supposed to work.

If I look at the pronunciation in an article, I assume I should be able to read it and figure out how to pronounce the word (kind of like in a dictionary). But in most cases, it is just a jumble of funny characters. Do I need some sort of plugin for it to work? -- Mufka (user) (talk) (contribs) 02:01, 27 February 2007 (UTC)

If it's a jumble of boxes, then your browser doesn't support the IPA characters, so yes. But if it isn't, then that's how it's supposed to be; IPA is a jumble of funny characters. -Amarkov moo! 02:06, 27 February 2007 (UTC)
Amendment to the above: "If it's a jumbled of boxes or question marks, then..." —Angr 05:32, 27 February 2007 (UTC)

Tin-Ring?

From this chart it appears that tin and ring would be represented with the same final consanant in IPA, when they don't rhyme to me. I know nothing about this but I'd be very interested if someone would explain?137.138.46.155 17:48, 9 March 2007 (UTC)

No, tin ends with [n] and ring ends with [ŋ]. Those are different characters; if they look the same to you, there's something wrong with how your browser is displaying them. —Angr 20:51, 9 March 2007 (UTC)
Ok thanks, yeah it must be a browser thing, I thought something must be up cos they're obviously different sounds. 137.138.46.155 08:32, 12 March 2007 (UTC)

[kʰ], [k] and [g] as well as [pʰ],[p] and [b]

Since we are a multi-lingual dictionary, I must raise this question. In the chart, k is described as the sound for cat, kick etc. Techincally, this is not quite accurate. Since English does not make a distinction between [k] and [g], this is not generally a problem, and [k] may be assumed to actually represent [kʰ]. However, if we want to be consistent, cat and kick should be spelled with [kʰ]. I point this out because in Min Nan , [kʰ], [k] and [g] are all different sounds. Similarly, English does not distinguish between [pʰ], [p] and [b] whereas Min Nan does make such a distinction. So when the table says that [p] represents the first letter in pen, pin etc, it should actually saying that [pʰ] represents the first letter in those words. I make an issue of this because if a native English speaker were to come across one of my Min Nan entries with IPA, and wanted to know how to pronounce it, the inconsistency between IPA chart for English and what is described at Voiceless velar plosive ([k]) and Voiceless bilabial plosive ([p]) may lead to confusion. On the other hand, I know that there are already a number of English words that use the "wrong" IPA, which would make fixing the problem a pain. Does anybody else think that it would be worth the effort to correct, or am I just being nitpicky? -- A-cai 12:53, 2 April 2007 (UTC)

You're being nitpicky. This article is specifically about English, so English phonemics can be assumed. If we were to transcribe English in such a way as to reflect all possible phonemic distinctions in all other languages, we would be doing an extremely narrow phonetic transcription, and miss the forest not just for the trees but for the individual leaves on the trees. —Angr 13:11, 2 April 2007 (UTC)
I'm not sure what you mean by my missing the forest for the trees. English is one language (one tree). IPA stands for International (forest, i.e. many languages) phonetic alphabet. I believe your sentiment is that IPA should be simplified in the case of English so that it is easier to use for English speakers. But as a native English speaker, I don't need to rely on IPA to tell me how to pronounce the word (hopefully). Presumably, the IPA is primarily included for the benefit of students of the language. If this is the case, having consistent rules is important. IPA is supposed to be a phonetic transcription system, which means that one symbol matches exactly one and only one sound. Otherwise, IPA becomes just another unreliable and inconsistent phonemic transcription system. I agree that I'm being nitpicky, but wasn't IPA invented by nitpicky people :-) Finally, how exactly is adding an ʰ in front of k and p being extremely narrow? I'm proposing one minor modification to an otherwise acceptable table. -- A-cai 14:14, 2 April 2007 (UTC)
The forest I mean is the phonological system of English. As I see it, the main audience of this page is native or very fluent speakers of English (since that's the vast majority of English Wikipedia's readership) who don't necessarily know IPA very well, but encounter it in a Wikipedia article. Granted, for the vast majority of English lexical words, they don't need it, but as an encyclopedia, Wikipedia includes a lot of names with non-intuitive pronunciations as well as obscure words like chough and slough whose pronunciation will not be known even to most native speakers. Foreigners learning to speak English have other and better resources than Wikipedia for learning correct pronunciation. —Angr 15:01, 2 April 2007 (UTC)
A-cai, in case you're not familiar with transcription practices, broad and narrow refer to the detail of the transcription; phonemic transcription can be broad, using only as much detail as is needed. Phonetic trascription can be narrow, using as much detail as possible. This article (currently, at least) is mainly phonemic. --Kjoonlee 16:03, 2 April 2007 (UTC)
It is perhaps worthwhile to point out that the IPA chart for English is indeed only meant to apply to English words. When words from Min Nan are described in IPA, certainly the indication of distinctive aspiration would be required. Apart from this, I wonder about some details mentioned in the first paragraph in this section. It states that English does not distinguish [k]/[g] and [pʰ]/[p]/[b]. This is incorrect. There is no phonemic distiction between only the pairs [k]/[kʰ], [t]/[tʰ] and [p]/[pʰ]. The phonemes [g], [d] and [b] are separate from all these. −Woodstone 20:30, 2 April 2007 (UTC)
Let me use an example Min Nan sentence to illustrate my meaning in the first paragraph ([kʰ], [k] and [g] as well as [pʰ],[p] and [b] are all different):
  • Góa kā ìn kóng, "Pháiⁿ-sè, ēng Tiong-bûn sī bô khó-lêng thang piáu-ta̍t." (English: I told them, "Sorry, it is not something that can be expressed in Chinese.").
    [gua ka in kɔŋ, pʰaĩ se, iŋ tiɔŋ bun ɕi bo kʰo liŋ tʰaŋ piau tat]
You can try to hear the differences by listening to the sound files above. However, if you're a typical native English speaker, you may find it difficult to hear the distinctions, especially between the unaspirated voiceless ([k], [p]) and unaspirated voiced ([g], [b]) sounds. As far as I know, English initial consonants only include the aspirated voiceless ([kʰ] and [pʰ]) and unaspirated voiced ([g], [b]) sounds. However, if you're saying that IPA chart for English is meant to be a phonemic chart, I'll accept that. -- A-cai 23:28, 2 April 2007 (UTC)
Yes, we're saying that IPA chart for English is meant to be a phonemic chart. —Angr 05:24, 3 April 2007 (UTC)

Scwhi as well as schwa?

What do you all think about including an entry for ɪ/ɨ, right below the one for ə? --Atemperman 03:30, 11 April 2007 (UTC)

Please don't move

Hi, please don't move this article to IPA pronunciation guide (English). That makes it sound like an enunciation guide. --Kjoonlee 02:25, 19 May 2007 (UTC)

I don't understand your concern. The current title is misleading, since the article has several charts and notes several tables and a notes section.
Further, "pronunciation guide" is the term used in dictionaries.
BTW, there are redirects from IPA pronunciation guide (English) and IPA pronunciation guide to this article.
--Jtir 09:46, 19 May 2007 (UTC)
I'd say the use in dictionaries is more confusing than sg/pl concerns. --Kjoonlee 11:27, 19 May 2007 (UTC)
And if there are redirects, all the less reason to move. --Kjoonlee 11:28, 19 May 2007 (UTC)
There are hundreds of articles linked here:
Obviously, I didn't look at all of them, but it is clear that WP editors regard this article as a pronunciation guide.
I can also see that the IPA templates should display the word "pronounced".
--Jtir 13:02, 19 May 2007 (UTC)

I find nothing wrong with using "pronunciation key" or "pronounced" as an anchor, but to say "pronunciation guide" seems awkward. --Kjoonlee 13:42, 19 May 2007 (UTC)

For an anchor, I would favor "pronounced", because it can be used to form a sentence, if so desired.
--Jtir 14:02, 19 May 2007 (UTC)

This page is great; thank you to the builders; native speakers DO often need pronunciation transcriptions; the phonemic, lay-friendly approach is exactly appropriate

Having read the topics on this talk page, I just want to say all of the things that I just said in the heading above. Wikipedia definitely should continue to have this resource, regardless of how it may be titled or which namespace it's put in. Native speakers DO often need pronunciation transcriptions; for example, when I discovered the article on the Levant, I had never heard this name before. The first thing I wanted to know, after learning the gist of what the Levant is, was this: What's the standard EN-US pronunciation of the name? So I looked it up in AHD4, then I added it to the Wikipedia article in IPA format. I strongly disagree with users who say "I think those little IPA squigglies make our articles look ugly. We don't need them because we're not a dictionary." Why should I have to bother to go look in some other reference work to find out how to pronounce, for example, Levant? Why can't I be efficiently informed of that info at the same instant, in the same place, that I'm reading to learn what the Levant is?

Re the phonemic, layperson-friendly approach—it's exactly what's appropriate given this page's main raison d'être.

In summary, thank you so much, and keep up the great work.

— Lumbercutter 21:08, 7 June 2007 (UTC)

I see from your edits of Levant that you had to work around the {{IPA}} template linking to IPA instead of this article. How did you find this article?
I like to put the IPA pronunciation in a Note, in part, to simplify the lead sentence. (We (novel), Sergey Korolyov) --Jtir 17:49, 8 June 2007 (UTC)
I can't remember now how I stumbled onto this article—clearly it was right at the time that I made those edits to Levant. A lot of people just link "IPA" to International Phonetic Alphabet without thinking too hard about it, but my problem with that is that I think most readers (i.e., linguistic laypersons) take a 20-second look at that article and conclude, "woah, sorry I asked—TMI." They really just wanted to know how to pronounce a certain word. This page is a much better link for that purpose, in my opinion. — Lumbercutter 02:38, 9 June 2007 (UTC)
D'oh! I'm embarrassed. I see now that the way that I stumbled upon this article is that there's a link to it right at the top of International Phonetic Alphabet! Nevertheless, it seems to me that I spent a lot of time reading that article before ever clicking on that link. Guess I should have been paying more attention! Probably what happened is that, being in a big hurry as usual, the first thing I did after landing on that article was to scroll down into the lead and start reading. Then eventually I paid enough attention in subsequent visits to the links at top. Oh well. This page is still cool, regardless! — Lumbercutter 03:19, 9 June 2007 (UTC)

// vs []

Should // vs [] be added on this page? Also doesn't using [] usually mean that what type of accent is used should be described as well? -Ravedave 02:55, 24 June 2007 (UTC)


Not user-friendly enough

As I see it, this article is a tool or guide for laypeople. But it doesn't work very well because on many screens there are rectangles instead of symbols. DON'T say that it is the reader's responsibility to down-load the appropriate software to make it readable. That is 1)often not possible on public machines and 2) something we cannot assume the average reader is both willing and able to do. It is the editors' responsibility to make articles that can be read almost as easily as a book is read. Basta. (I'm talking about the main article, notthis discussion page, where things can be laxer.) Kdammers 17:57, 24 July 2007 (UTC)

There is virtually no article on Wikipedia that is 100% usable for all users on all computers and all browsers. Should we delete all images because they can't be seen by blind users using screen readers? Should we delete the Chinese Wikipedia, just because not everyone can correctly view the Chinese characters (after all, downloading the appropriate fonts is often not possible on public machines)? —Angr 19:39, 24 July 2007 (UTC)
Should we TRY and make everything accessible? yes. The symbols can be replaced by images on this page since this is the page used to decipher hundreds of other pages. The IPA page itself can remain the way it is. I will try and do so soon. -Ravedave 20:36, 24 July 2007 (UTC)
I agree, images would make them much more accessible, but mysteriously, they've (the characters) found themselves a place in my computer. Only my browser, though. It's the same with all of the languages listed as the largest other-language Wikipedias on our Main page. No downloading of language packs, no availability of, say Japanese, or most of the IPA-specific characters in my word processor or IMs. This hasn't happened on other computers I've used (perhaps because I'm a regular viewer(?), on my own machine), so the problem still stands. The Chairman 15:17, 25 July 2007 (UTC)
There was some talk at the Korean Wikipedia about using images instead of text for IPA; if you're going to use images, please add them instead of replacing text. --Kjoonlee 17:58, 25 July 2007 (UTC)
On another note, you can find a lot of IPA images at X-SAMPA, but almost all of them use an opaque white background instead of using a transparent one. Almost all are in the Gentium typeface, with approx. 3-pixel margins. The letters are about the same size as "22" in Inkscape. --Kjoonlee 18:05, 25 July 2007 (UTC)
I have posted an image of the chart (which is also handy for other reasons as well) -Ravedave 00:54, 11 August 2007 (UTC)
Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:International_Phonetic_Alphabet_chart_for_English_dialects/Archive_2&oldid=1091452995"