Reply Phonemic Transcription in dictionaries.
I have some objections to some of the pronunciation guides used in dictionaries.
One of the things I really don't like is when dictionaries transcribe the ''wh'' words with /hw/. The sound the some people use in ''wh'' words to distinguish them from ''w'' words is a voiceless /w/ not /hw/. The dictionaries would be better if they used an upside down ''w'' for the ''wh'' words. Transcribing them as /hw/ is silly, because no one pronounces them as /hw/. It's a voiceless /w/ not /hw/.
/hw/ occurs in ''Huang River'' and is totally different from the voiceless /w/ that some people use in ''wh'' words.
Would you agree with my objection?
Reply I Agree
Instead of /hw/ the IPA's up-side-down "w" would be better. SAMPA's /W/ would be good to. Also a "w" with a diacritical mark would work (the IPA's circle: the voiceless diacritic). If it were up to me, I'd favour a rounded "w" (i.e. symbol which would look like an up-side-down "m"). Failing any of these, I'd rather see /W/ and /w/ not distinguished than see /W/ written as /hw/.
I think the correct IPA symbols should be used. But English dictionaries make lots of mistakes in their phonetic transcriptions, the most glaring of which are /e/ for /E/ and /i/ for /I/.
Dictionaries use phonemic transcriptions and the symbols they use do not have to correspond to IPA symbols. That is why I say dictionaries use "IPA-based" phonemic transcriptions (emphasizing "-based").
I don't know the justification of your claim that /e/ is the wrong phonetic symbol for the sound in "bed" in BrE. Daniel Jones seems to disagree with you, judging from his work on the EPD.
Anyway, I think the use of /r/ instead of the "upside-down r" should be a much more glaring "mistake", according to your standards.
As far as I'm concerned, dictionaries can use a triangle for /e/ and a circle for /r/. The symbols they use represent phonemes, so there's no reason why they should match the symbols the IPA uses to represent phones, is there?
The IPA itself tends to encourage the use of IPA symbols that most closely represent the most common allophones of the phonemes being represented. For English, this means that the upside-down 'r' is most appropriate for the 'r' sound (because this is the symbol for the most common allophone, a postalveolar approximant), and the lowercase epsilon is the most appropriate for the vowel in "bed" (the /e/ represents the sound of "made").
The problem with English dictionaries using the wrong symbols is that this is very misleading to foreign students of English who actually know how the IPA works, and it can lead to a lot of mistakes and wasted effort. Offhand, I can't think of any English pronunciation in which [E] becomes [e], and the use of the latter symbol causes confusion with the real /e/, which is a separate phoneme.
I agree entirely. More dictionaries should use IPA symbols. That why I like the NTC dictionary of American English. It uses a subset of IPA.
Also you need hw for the actual sound of common Spanish name Juan.
Regards, Paul V.
"I think the correct IPA symbols should be used." writes Mxsmanic overlooking (as usual) the fact that there are no "correct" IPA symbols.
"English dictionaries make lots of mistakes in their phonetic transcriptions," he writes and then goes so far as to give the two "most glaring" examples: "/e/ for /E/ and /i/ for /I/."
Stuff and nonsense! I agree with Tom entirely on this point. It's a phonemic not a phonetic transcription that is used in dictionaries and with very good reason. You can base phonemic transcriptions on the IPA or you can base them on ordinary English orthography. You could use runes if you like or even circles, triangles, etc. like Tom says.
"The IPA itself tends to encourage the use of IPA symbols that most closely represent the most common allophones of the phonemes being represented." I doubt that Mxsmanic would speak Lithuanian in a Scottish accent just because the IPA happens to encourage it. I have enough respect for the IPA but there is no obligation to do as they say.
Mxsmanic goes on to write "the lowercase epsilon is the most appropriate for the vowel in 'bed' ..." This is simply not the case at all. Whether you pronounce "bed" as [bEd] or as [bed] (or something in between) really depends on your accent. Having dictionaries prescribe "standard" accents may suit Mxsmanic just fine but in truth it's a load of rubbish.
I happen to have one of those accents in which "bed" is pronounced [bed]. Oh! But my accent isn't "standard" according to Mxsmanic and we should use the IPA symbol which represents the "standard" accent. I don't believe in any "standard" accent but the realisation that it's a phonemic transcription not a phonetic one makes this irrelavant.
I could go on but I've written what I had to say about this only five weeks ago. http://www.antimoon.com/forum/2004/5719.htm
One thing that I will repeat is that there is nothing wrong with using the ordinary spellings that you find in English orthography as your basis (or part thereof). The letters "e" and "r" are fine for a phonemic transcription of /e/ and /r/: this is their usual spelling in English, that's enough.
Mxsmanic goes on to complain "the /e/ represents the sound of 'made'" ... the vowel obviously. Well maybe in Mxsmanic's accent it might but "made" for me sounds closer to [mæId] not [med]. This is quite an interesting thing coming from Mxsmanic though considering the fact that he transcribes "make" as [meIk] in his "Betty Botter" rhyme.
"The problem with English dictionaries using the wrong symbols is that this is very misleading to foreign students of English who actually know how the IPA works," he writes. If they're so cluey as to how the IPA works, they probably realise that it's a IPA-based phonemic transcription not a phonetic IPA transcription that they fine in their dictionaries.
However, what of the vast majority of us (foreigners and native speakers alike) who don't know how it works? The greater problem, as I see it, is in the other direction: people's being mislead about the IPA symbols based on their knowledge of English pronunciation.
This is why I'd advocate a clean break with the IPA and the development/adoption of an International phonemic alphabet for English, e.g. Pitman's ITA, something closer to English orthography.
The Jym that Blacksmith copies and pastes is me, by te way. I'd like to add that the circle diacritic of the IPA is just one example of a diacritic that could be used to distinguish /w/ and /W/.
I've had a listen to the so-called /E/ and /e/... English speakers don't distinguish between them since they use either interchangeably without realising it. A phonemic distinction is made between them in languages like French: where they are represented as è and é, but not in English.
English speakers confuse this closed /e/ with the English diphthong "ay" as I believe Mxsmanic is doing here. So, "made" is pronounced either /meId/ or /mEId/ by the vast majority native speakers, not /med/ or /mEd/ which is spelt "med".
>>So, "made" is pronounced either /meId/ or /mEId/ by the vast majority native speakers<< (especially Americans).
PS: However, as Jim noted, in Australian English "made" is pronounced [mæId].
I'm not confusing anything. In a phonemic transcription, you don't have to represent the diphthong as /eI/; you can just represent it as /e/. And [e] is an allophone of the phoneme, anyway.
What you must do is distinguish between this phoneme, heard in made and raid and play, from the phoneme heard in bed, shed, and men. They are two very different and very distinct sounds.
Using /e/ for /E/ leads to tremendous confusion among students who speak languages in which /e/ is used correctly. They must constantly be corrected on the incorrect pronunciations they've inferred from the incorrect use of the IPA in English dictionaries.
Pronouncing "get" as [get] is perfectly fine. I do it all the time; as do the rest of us Aussies. The Kiwis have an even more closed /e/ (yes /e/) than we have. Now, Mxsmanic, are you going to say that we pronounce /e/ "incorrectly"? Using either [e], [E] or something in between for /e/ should cause no confusion at all.
What is likely to cause confusion for learners would be transcribing /ei/ with an "e". Sure, you're free to use "e" for /ei/ (as long as you're not using it for something else). You're also free to use "A" or "a" with a macron (as in some old dictionaries). You're free to use whatever you wish: runes, Shavian, katakana, shapes, colours, bits of rotten flesh, etc.
One problem is when you tell folk that you're using the IPA and forget to stress that it's not really the International Phonetic Alphabet that you're using but an IPA-based phonemic transcription. Another problem is when you use an IPA symbol that matches a common English spelling for a sound that this common English spelling does not correspond to.
"What you must do is distinguish between this phoneme, heard in made and raid and play, from the phoneme heard in bed, shed, and men." writes Mxsmanic "They are two very different and very distinct sounds." Well, hello, I wasn't born yesterday nor do I expect was Mick or Tom. I don't know whether you've noticed but none of us are suggesting that /e/ and /ei/ be transcribed with the same symbol.
Mxsmanic writes that "[e] is an allophone" of /ei/. This is true for some accents (Californian, I believe, is one). However, as I keep writing, [e] is the standard pronunciation of /e/ in AusE and NZE and never an allophone of /ei/. We still understand Californians of course ... their accent at least.
If you transcribe /ei/ as /e/ then you're likely to have learners think that it should be pronounced as a monophthong. As I've noted, I'm aware that it can be pronounced as a monophthong but most of us will use a diphthong. In order to be better understood the learner would best be advised to pronounce it as a diphthong rather than a monophthong.
If you go telling students that it's okay to pronounce "gate" as [get] then you're doing them a disservice. They'll risk being misunderstood: people might think they'd said "get". The riskes will be hightened in Australia and New Zealand where [get] is standard for "get". Now, I'm sure that Mxsmanic will again be insisting that AusE and NZE are not standard. My question is "Are the accents in which /ei/ is pronounced [e] amongst those that Mxsmanic would choose to label as 'standard'?"
>> Using /e/ for /E/ leads to tremendous confusion <<
Both are monophthongs. The only difference I hear between them is one is an open "e", while the other is closed, and to English-speakers' ears this is inconsequential. In other words, they are treated as allophones. Further, neither is a diphthong.
However in the following recordings they lose their short/pure monophthong quality somewhat as the speaker lengthens them laxly:
Also, his "epsilon" is _excessively_ open to the point where it's virtually an /æ/.
That said, I'd have to listen to what you're referring to. That could change things.
>>Pronouncing "get" as [get] is perfectly fine. I do it all the time; as do the rest of us Aussies. The Kiwis have an even more closed /e/<<
We pronounce it either [get] or [gEt] and anything in between: there is no difference. So do Americans and other English speakers. And you're 100% re: the Kiwis.
Maybe Mxsmanic has an accent where the /e/ monophthong is an allophone of the diphthong /eI/... unusual as it seems.
Most ESL students will never go near Australia, so whether or not they talk like Australians is irrelevant. Throughout most of the rest of the world, they can pronounce gate as [get] and they will be understood without any problem, particularly if they also pronounce get as [gEt].
For most native English speakers, there's a huge different between [e] and [E], just as there's a huge difference between [i] and [I]. Australia is not representative, and it is not important. No amount of repetition will change this.
When my ESL students start telling me that they want to learn to speak with an Australian accent, then I'll worry about Australian pronunciation. But I have yet to hear any student mention Australians at all. About the only pronunciations they are interested in learning are GAE or RP; and the only other pronunciations besides these that they've expressed an interest in understanding are those of Ireland and Scotland.
If I said [wed], would you think I said "wade"?