Wikipedia Help NA‑class

This page is within the scope of the Wikipedia Help Project, a collaborative effort to improve Wikipedia's help documentation for readers and contributors. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks. To browse help related resources see the Help Menu or Help Directory. Or ask for help on your talk page and a volunteer will visit you there.Wikipedia HelpWikipedia:Help ProjectTemplate:Wikipedia Help ProjectHelp articles NA This page does not require a rating on the project's quality scale.

Archives

October 2007 to March 2008 April 2008 April to December 2008 January to February 2009 March to May 2009 June 2009 to July 2009 August to September 2009 October 2009 to April 2010 February to March 2010 April to May 2010 June 2010 to March 2011 /Archive 10 /Archive 11

This page has archives. Sections older than 90 days may be automatically archived by Lowercase sigmabot III.

I am not sure wheher this article is aimed at helping Wikipedia readers to understand IPA symbolization, or helping writers to use the symbols properly. But the Key section seems to me very disorganized. The vowels and consonants are listed in what seems like random order, compared with the tidy layout seen in the article on English phonology, and this makes it very difficult to check what is there (where is the NURSE vowel, for example?). Some sections in the Key just don't work - the syllabification section has a lot of errors. Readers may well find it hard to understand (as I do) why the vowel section of the key has a separate section for vowels combined with /r/, with some vowels corresponding to those in the other vowels column and others (e.g. some triphthongs) not corresponding. Can this be overhauled? Peter Roach. RoachPeter (talk) 17:44, 11 March 2012 (UTC)[reply]

This page is aimed at helping both readers and editors understand how IPA is used for English at Wikipedia. The symbols are in quasi-alphabetical order, but of course for symbols that aren't part of the normal Latin alphabet we have to improvise, e.g. by listing θ as if it were a variety of O. The NURSE vowel is in the r-colored vowel column after ʌr; apparently whoever put it there considers it the heterosyllabic realization of ʌ + r. Personally, I'd put it between the SQUARE vowel and the mirror/Sirius vowel, leaving the non-r-colored side empty. As for the syllabification section, I don't see any flat-out errors, though I do see some implicit claims that may be controversial. (AFAIK there's very little consensus in the literature about where syllable breaks fall in English.) As for why there is a "vowels followed by /r/" section, it's because the key is trying to accommodate as many accents as possible (cf. International Phonetic Alphabet chart for English dialects). So for example, most North American speakers using this key have to plot the symbol "ɒ" onto their /ɑ/ phoneme in words like lot, but they have to plot "ɒr" onto /ɔr/ in words like moral, so we list plain "ɒ" and "ɒr" separately. Which triphthongs are missing? Angr (talk) 07:23, 12 March 2012 (UTC)[reply]

OK, let's start with smaller matters. Regarding syllabification, the IPA dot symbol is used to show where syllables are divided. There is quite a good consensus on where English syllable breaks fall, and the English phonology article says a bit about it under Phonotactics. Both major British pronunciation dictionaries show syllable divisions in all polysyllabic words, and explain their criteria for doing so in their Introductions (see the Longman Pronunciation Dictionary edited by J.C.Wells and the Cambridge English Pronouncing Dictionary which I edit). Now the examples in the present article show sensible use of this convention for 'hire', 'higher', 'moai'. The syllable division is supposed to help readers see that 'Windhoek' does not have two syllables in the part following 'Wind-", if I have understood correctly, so here it the ABSENCE of the dot which is supposed to be helpful. In 'Vancouvieria' it seems again that the absence of a dot is supposed to tell us something, but it makes it look as if everything after the stress mark is just one syllable. As for 'Myki' and 'Mikey', I simply cannot see how they could be syllabified differently. Regarding triphthongs, the usual presentation for British English (RP/BBC) is to take the set of diphthongs in FACE, PRICE, CHOICE, GOAT, MOUTH and add a schwa on the end. It is not normal to regard triphthongs as unit phonemes. Whether you get triphthongs followed by /r/ in the same syllable varies from one accent to another. While on smaller matters, the table of Reduced Vowels contains some things that need fixing, for example the statement that a vowel is frequently dropped in 'nasturtium'.

For the presentation of the vowels and consonant symbols, the usual way is to group them under phonetic criteria. For vowels this usually means one block for short or lax vowels, one block for long or tense vowels and one for diphthongs. Since this presentation is apparently meant to be used for most major varieties of English pronunciation, it does not seem to me to be sensible to present a restricted list of combinations of vowel symbols with /r/. All the vowels and diphthongs may in principle be combined with /r/ , so there is no point in having a separate table, except perhaps to give some helpful examples.

For consonants, again it would be better to use phonetic criteria, starting with plosives /p, t, k, b, d, g/, going on to fricatives, affricates, nasals and so on. There is no need to have a separate entry for the nasal plus /g/ in 'finger'. This is a simple combination of two phonemes. Peter Roach. RoachPeter (talk) 09:58, 12 March 2012 (UTC)[reply]

The ⟨.⟩ is only used where there might be some confusion. Windhoek and Vancouver might be misread as having th and ng sounds in them if we didn't use the dot. "Myki" and "Mikey" might be obvious to you, but the point of a transcription is for when the reader doesn't know the pronunciation.

Wells is hardly universally accepted. I suspect that MW and the OED would disagree in many many cases. And experience has shown that if we do indicate syllable boundaries, people will "correct" them, then correct the corrections, and soon we'll have a fight over who does or does not understand English.

Aren't your triphthongs just the vowels in our rhotic column?

The reason we need a rhotic column is that there is no regular correspondence between /V./ and /Vr./. We do not use a phonemic transcription because RP speakers balked at it being too American. It was one of the compromises we made to get an English IPA convention, rather than listing multiple dialects for each transcription.

Consonants should not be grouped by phonetic principals, IMO. If you know those, you won't need the key. There is a need for /ŋɡ/, due to the confusion from orthography. — kwami (talk) 10:31, 12 March 2012 (UTC)[reply]

(edit conflict)I think the point of the syllable break in Windhoek is not to show that there is only syllable after the "Wind-", it's to prevent nonspecialists from interpreting the "th" of /vɪnthʊk/ as /θ/, while in Vancouveria it's there to show the absence of nasal assimilation of /n/ to /ŋ/. I cannot agree with you that there is quite a good consensus on where English syllable breaks fall. Even in a comparatively simple word like happy, there's considerable disagreement as to whether it's [hæ.pi], [hæp.i] or [hæpi] with an ambisyllabic [p]. (And there's no way to use the '.' to mark ambisyllabicity without creating the illusion of a geminate consonsant, e.g. [hæp.pi].) I'm not familiar with the word "Myki", but as it seems to be derived from "my key", I think the point is that in "Mikey" the /k/ belongs (partly or wholly) to the first syllable while in "Myki" it's entirely in the second syllable, which depending on accent could have an effect on both the quality and the duration of the diphthong in the first syllable. You say "all the vowels and diphthongs may in principle be combined with /r/", but in most varieties of North American English the selection of vowels that may occur before /r/ is quite restricted, regardless of whether the /r/ is syllable-final or followed by another vowel. Thus an American for whom /æ/ can never be followed by /r/ will find out from this page that when he sees /ær/ in a transcription, he is to interpret that as the sound of marry or barrow (i.e. [ɛr] for him). Grouping the symbols by phonetic criteria is useful for linguists who are already familiar with the phonetic criteria, but users who come here looking for the meaning of a symbol will probably have better luck finding it if the symbols are in alphabetical order or a reasonable approximation of it. Angr (talk) 10:55, 12 March 2012 (UTC)[reply]

Now that you're familiar with some of the issues at play with the key, Peter, would you be interested in putting up what you have in mind as a sandbox subpage? It wouldn't be a good idea to mess with such a high-profile page (any time someone has a question about an English word's IPA, they go here), but your experience in writing introductory textbooks might prompt us to see how to modify the presentation of our transcription system in a more helpful way to our readers.

When you've done it, you can let us know and we can discuss it and incorporate any things that people like. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 14:31, 12 March 2012 (UTC)[reply]

Since you're relatively new to Wikipedia and might not know what a "sandbox subpage" is, you can create it by clicking here: Wikipedia:IPA for English/Sandbox. Angr (talk) 14:45, 12 March 2012 (UTC)[reply]

thanks, I would be happy to have a go at doing that in a way that does not overwrite what is there aleady. I will delay doing it until I am back in UK and able to use a proper computer instead of pecking away at an iPad on holiday! Peter Roach RoachPeter (talk) 16:59, 13 March 2012 (UTC)[reply]

In the rightmost column, 15th entry, I think there's a typo. The phonetic symbols spell out 'yure' whereas the example word is 'cure.' Add a k-like symbol before the j-like symbol (or potentially a kx-combo, since the aspirated stop 'k' creates friction as the tongue progresses to 'y'), and it'll match.Tdbostick (talk) 08:18, 29 May 2012 (UTC)[reply]

The C is not bolded, so jʊər is intended to be the pronunciation of the URE only. ― A. di M.​  10:16, 29 May 2012 (UTC)[reply]

Nothing is bolded in CURE. Suggest making URE truly boldface, and the problem will be solved 134.176.158.20 (talk) 10:44, 11 July 2012 (UTC)tdbostick[reply]

Upon closer inspection by zooming in the page, URE is bolded, however this is impossible to see at a normal zoom factor because the font and/or size of the text doesn't match similar nearby entries. 134.176.158.20 (talk) 10:44, 11 July 2012 (UTC)tdbostick[reply]

The capital letters mark lexical sets, so we may want to simply add another example. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 12:32, 11 July 2012 (UTC)[reply]

How about replacing the boldfacing with underlining? A. di M. (talk) 14:38, 11 July 2012 (UTC)[reply]

According to this IPA template discussion underline obscures some IPA characters. -- Q Chris (talk) 07:57, 12 July 2012 (UTC)[reply]

I think he's talking about the bold in the example words and lexical sets, not the IPA. Underlining would be fine with me. We could even do both underling and bolding. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 11:58, 12 July 2012 (UTC)[reply]

Ah, in that case it's fine with me too -- Q Chris (talk) 12:33, 12 July 2012 (UTC)[reply]

It says that ɨ can be pronounced as either ɪ or ə, using the examples roses and emission, with the explanation that it differs between dialect. I can understand that different dialects pronounce a word differently, but aren't IPA characters supposed to be the 'atoms' of linguistics, the things that you can't break down? Or does IPA have homophones after all? Could somebody expand on this? — Preceding unsigned comment added by 81.71.110.7 (talk) 20:50, 24 June 2012 (UTC)[reply]

This page uses the symbol ɨ like an algebraic symbol to stand for "ɪ or ə depending on accent", not in its IPA role of "high central unrounded vowel". Angr (talk) 21:05, 24 June 2012 (UTC)[reply]

Right, and this means that the use of the "barred i" symbol as noted here is not IPA practice. RoachPeter (talk) 17:43, 25 June 2012 (UTC)[reply]

The IPA, a phonetic alphabet, has commonly been used for phonemes. Phonemes themselves are abstract groupings of phones. The diaphonemic approach here furthers the abstraction to cross-dialectal correspondences. It's not a common practice to use the IPA explicitly in this way, but it's useful for our purposes. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 18:00, 25 June 2012 (UTC)[reply]

I don't think it's really that rare to use IPA symbols diaphonemically and cross-dialectally, though it is rare to try to represent both RP and GenAm using a single set of symbols. But using a single set of symbols to represent all the accents of Southern England, say, or even Southern England plus the Southern Hemisphere, is actually fairly common. And in other languages, it's common enough to use a single symbol to represent a range of dialectal realizations; for example Duden uses /r/ to stand for all realizations of the German r-sound, whether [r], [ʀ], [ʁ], or [ɐ̯] (in coda position after a short vowel; it does use /ɐ̯/ in coda position after a long vowel). Angr (talk) 18:11, 25 June 2012 (UTC)[reply]

Yeah, it's pretty common to pick a character to represent a phoneme that has different dialectal variants. It's also common enough to have one transcription system for phonetically distinct varieties that share the same phoneme inventory. It's less common to combine two or more varieties that have divergent phonemic systems or even, in the case above, to use a single character to represent systematic differences in incidence. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 18:52, 25 June 2012 (UTC)[reply]

I just went through a few dozen Englishcommonwealth placenames and put in/restored postvocalic r's in the IPA transcriptions. For some reason, I thought that that was something that would be more closely monitored. Is there a walled-garden thing going on? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 17:37, 4 July 2012 (UTC)[reply]

There has been in the past. I haven't been keeping up recently. But in Melbourne I think a note about the local pronunciation is useful. Yes, it's covered by the generic one, but we may have readers who are interested by don't know the local dialect to be able to predict it. (But s.t. like Bathurst is so obvious that IMO it doesn't need mentioning.) — kwami (talk) 19:50, 4 July 2012 (UTC)[reply]

I made the change to Melbourne before looking in the talk page archives. I've re-added the local pronunciation and made its presentation a little clearer. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 20:28, 4 July 2012 (UTC)[reply]

[i] is not a reduced vowel. The long markings should be removed from the other vowels because there aren’t short long contrasts in English, compare to say Hungarian or Japanese. I’d consider this inappropriate for the purposes of a general overview that gets linked to as a general pronunciation guide on the other pages. I keep seeing in the notes “we’re doing this and leaving it unspecified.” If you’re not trying to specify it on some level, what is the purpose of this article?

Also, one of the first rules to writing a phonetics paper is the define each symbol you are using by place/manner/etc on the IPA chart, this is especially crucial if you’re choosing phonemic symbols that do not quite match the sound you are describing. I also *strongly* advise against using a glaringly different symbol as it creates confusion. You can’t expect someone referring to the IPA guide on wikipedia to have studied linguistics enough to understand that this or that person is known to use an inaccurate symbol phonemically. If you don't want to write out the description, perhaps link to it?

Mordrynne (talk) 22:22, 7 July 2012 (UTC)Mordrynne[reply]

While [i] may not be reduced (and this is arguable), it only appears in unstressed vowels. I'm not sure what a better term would be.

English does indeed have short-long contrasts, though the distinction is enhanced by qualitative differences, which differs from e.g. Japanese, which makes contrasts purely on length (Hungarian is like English in this regard).

I'm not sure what "we're doing this and leaving it unspecified" means. Can you point to an example?

The problem with a number of the entries here is that there is no one specific phonetic value, only a range of them. For example, the PRICE vowel can be [ɑe], [aɪ], [ʌi], [ɔɪ], [əi], [aː], or [ai], depending on dialect. That's one of the reasons we've been emphasizing lexical sets, here and in the usage of {{IPAc-en}}. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 23:09, 7 July 2012 (UTC)[reply]

there aren’t short long contrasts in English -- depends on what variety you are talking about. There are in mine. Grover cleveland (talk)

I appreciate Aeusoes1's change or the example from perform, which confused me, but I don't think that the "er" in "perfume" has the same sound. I personally pronounce the 'er' as "ɜr" fitting the wictionary /ˈpɜːfjuːm/, which is contrasts with the wictionary for perform /pə.ˈfɔːm/. I think ideally we could do with a different example, but failing that perhaps we should revert to "perform". -- Q Chris (talk) 13:00, 10 July 2012 (UTC)[reply]

Oh, gosh. I'd assumed everybody pronounced it like I did, with stress on the second syllable. Perhaps perhaps would do. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 13:17, 10 July 2012 (UTC)[reply]

I think I have /ɜr/ in "perhaps". Either syllable could be stressed. — kwami (talk) 13:29, 10 July 2012 (UTC)[reply]

This is more difficult than I thought! Perhaps would certainly do for my pronunciation, and the Wiktionary transcription as /pəˈhæps/ for RP and /pɚˈhæps/ implies it is also correct for both RP and General American, but from Kwamikagami's comments this might not be universal. -- Q Chris (talk) 13:33, 10 July 2012 (UTC)[reply]

It works in connected speech, but as an interjection it's more variable. — kwami (talk) 13:40, 10 July 2012 (UTC)[reply]

Perceive? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 13:45, 10 July 2012 (UTC)[reply]

Again fine the way I pronounce it, and Wiktionary confirms it as OK for RP and GA. -- Q Chris (talk) 16:35, 10 July 2012 (UTC)[reply]

Works for me. — kwami (talk) 00:49, 11 July 2012 (UTC)[reply]