User talk:Mzajac/IPA reference
Tried alphabetizing a bit more. I agree we could use /hw/. Always sounded like a consonant sequence to me anyway. We need the rhotic vowels, however, whenever they don't use the normal vowel letters, and I doubt the foreign vowels are any more common in our entries than /hw/. kwami (talk) 20:50, 8 February 2009 (UTC)
- Sorry I reverted, maybe too quickly, but it was hard to sort out the changes because of combined edit and re-order. I still need your input to refine this. —Michael Z. 2009-02-08 23:17 z
Sample words
I tried to choose sample words from the near the top of the Gutenberg frequency list. I like measure over beige because it seems to be more native to English rather than very French, but short is good. —Michael Z. 2009-02-08 23:21 z
Uh-oh vs. button: I guess the former is a better example, but does it convey what the sound is? Button is more clearly indicated, but is it pronounced with the stop very widely? —Michael Z. 2009-02-08 23:24 z
In Canadian English for = [fɔɹ], but I think bought = [bɒt], raw = [rɒː]. Is there a better example without the rhotic r?
- But our IPA key is not intended to be regional. Glottal stop is only widely found in paralinguistic expressions like 'uh-oh' and 'hm-m'. In generic IPA-en, 'button' is /ˈbʌtən/, 'for' is /fɔr/, 'bought' is /ˈbɔːt/, and 'raw' is /ˈrɔː/. If you don't want the 'r', there's 'law' /ˈlɔː/, 'jaw', 'maw', 'paw', 'saw', 'yawn', etc. (I don't have [ɔ] at all in my dialect, apart from ɔɪ, ɔr, and ɔl, so I always have to look this up.) kwami (talk) 23:46, 8 February 2009 (UTC)
- Dictionary.com is quite close to IPA-en (given a few minor diffs like the length mark), though what appear to be alternate pronunciations are very often regional variations (like the regional collapse of ɒ and ɔː), which we ignore, rather than true speaker alternates like 'vase' and 'tomato'. kwami (talk) 23:53, 8 February 2009 (UTC)
Order of examples
I modelled the order approximately after the CanOD's, in the bottom margin of each page. they present the simpler symbols first, then throw in the less familiar symbols, compound symbols, and foreign sounds at the end. —Michael Z. 2009-02-08 23:32 z
- I was trying to guess where the reader would look first, but that's complete OR on my part. I'm sure the CanOD has spent some time on this, so you're probably right to go with them. kwami (talk) 23:48, 8 February 2009 (UTC)
- I wouldn't mind surveying a few North American dictionaries for their order, because the CanOD isn't perdfect, but I don't have any other paper dictionaries at hand. As long as it's systematic, I think it's good, because then the reader can perceive the pattern and predict. But do we alphabetize by the vowel letter in the sample word, by the nominal Latin equivalent of the IPA symbol, or by some relationship of the sound to the English Latin alphabet? —Michael Z. 2009-02-09 00:07 z
- Hardly any US dictionaries use the IPA, but the new OED transcription is listed here. For foreign vowels, however, they pretty much just follow the foreign pronunciation, and don't try to incorporate them into English phonotactics. For instance, they also use pure [a e i o u] vowels. I guess the idea is that if you pronounce a [y] or [ø], you're not pronouncing the words as if they were English, so there's not much point in trying to transcribe them as if they were English. That was my point above about leaving these symbols out, though if Canadians pronounce tu with a [y] while speaking English, then that could be an exception. But really, how many foreign names are pronounced with non-English sounds, without being pronounced as foreign names? kwami (talk) 01:04, 9 February 2009 (UTC)
- Point taken. From memory, the CanOD does say that their pronunciations are “based on thorough surveys of English-speaking Canadians from across the country” and “all pronunciations of foreign place names and people's names are transcribed as they are most commonly spoken by English speakers, as are the pronunciations of words recently borrowed from other languages. Whether the pronunciation given is in the ‘foreign’ or Anglicized form reflects the degree to which the word has been naturalized by English-speaking Canadians.” In the chart, this would only be deux and the three nasals. Schön and tu came from respelling, and I assumed that was because they were used in some English somewhere – I'll remove them until we find evidence this was the case.
- But this chart, and the CanOD's IPA guide, are meant to work exactly the same way as the American dictionaries' guides at the top or bottom of the page, so I'd like to follow their example. (I think our use of phonemic IPA for English is pretty close to their use of letters with diacritics in most ways, or should be.)
ɜː vs. œ/ø
I got /ɜː/ (deux) from the CanOD (yup, length mark), and /œ, ø/ (schön) from Pronunciation respelling for English. (CanOD also has ɜr her.)
I'm not at all certain whether these are different sounds or not in English. I like /ɜ/ because it's already used elsewhere so, if it works, why add another symbol? —Michael Z. 2009-02-09 00:22 z
- Well, it's an unrounded vowel, whereas the French word is rounded. It's also not what people who know the IPA would expect, as it's not used in the French or German textbooks English speakers use. I doubt English speakers make a distinction between /œ, ø/ or /y, ʏ/, but who knows, maybe some do. kwami (talk) 00:52, 9 February 2009 (UTC)
collapsible
Can we make it collapsible, so that it can be added to list articles which have dozens of transcriptions without taking up too much space? I've tried with class="wikitable collapsible collapsed", but it completely screws up the formatting. Maybe you can find the underlying code? kwami (talk) 23:22, 10 February 2009 (UTC)
- You know what would be nice, if we could get it so that this popped up when you hovered the cursor over the "IPA:" or "pronounced". I tried incorporating this using only line breaks and the IPA template into the title parameter in the template, but no dice. kwami (talk) 07:42, 11 February 2009 (UTC)