RP is the accent with the most phonemes?

Jim   Wednesday, September 15, 2004, 04:24 GMT
... aren't bothered with by him but speaketh he for all American ESL teachers? Good question nonetheless.
Mxsmanic   Wednesday, September 15, 2004, 04:37 GMT
Jim asks: "If 'the number of non-phonemic diphthongs varies greatly from one variety of English to the next, but the number of phonemic diphthongs is virtually constant at three (now, toy, and eye)' then what about /ei/ as in 'saint', 'weight' and 'cake' and /Ou/ as in 'tow', 'phone' and 'post'?"

Neither is phonemic. There are no words that distinguish between [e] and [eI] or [Ou] and [o], for example, so these diphthongs are not phonemic.

There are zillions of words that distinguish between [OI] and [O] or [I], though. The same is true for the phonemic diphthongs [AU] and [AI]. But you only need these three; all other diphthongs are allophones of monophthong phonemes.
Mi5 Mick   Wednesday, September 15, 2004, 07:04 GMT
>>... aren't bothered with by him but speaketh he for all American ESL teachers?<<

Nope but he/she normally knows what he/she's on about, especially these things. Maybe he/she's too high and holy to answer this one. hehe!
Jim   Wednesday, September 15, 2004, 08:07 GMT
The problem is calling phonemes "diphthong", "long monophthong" and "short monophthong". There are phonemes which are diphthongs for some and monophthongs for others.

The /ei/ in "saint", "weight" and "cake" and the /Ou/ as in "tow", "phone" and "post" may not be diphthongs in all accents thus it would be fair to say that they are not phonemic diphthongs but to say that they are not phonemes is another matter.

Mxsmanic,

Am I right to guess that you're using SAMPA? If so then

[e] is the close-mid front unrounded vowel equivalent to Tom's /e/ (in my accent and RP).
[o] is the close-mid back rounded vowel equivalent to Tom's /o:/ (in my accent).
[I] is the lax close front unrounded vowel equivalent to Tom's /i/ (in my accent).
[O] is the open-mid back rounded vowel equivalent to Tom's /o/ (in my accent) or Tom's /o:/ in RP.

http://en.wikipedia.org/wiki/X-SAMPA

Or are you just using Tom's alphabet and accidentally capitilised the "I"? Either way I still say that the vowels in "saint" and "no" are very definitely phonemes in my accent.

"There are no words that distinguish between [e] and [eI] or [Ou] and [o] ..." you claim. I distinguish "get" from "gate", "sent" from "saint", "bet" from "bait". I distinguish between "caught", "coat" and "cot".

What do you mean not calling /ei/ and /Ou/ phonemes? They most are in my accent ... and RP amongst many others for that matter. What is are allophones for one person may not be for another.

I'm sorry that I don't share the confidence Mi5 Mick has in your word.
Mi5 Mick   Wednesday, September 15, 2004, 13:05 GMT
What kind of work are you in Jim?
Mxsmanic   Wednesday, September 15, 2004, 18:52 GMT
I'm using Kirshenbaum's IPA transliteration. I don't know how close that is to SAMPA, and frankly, I don't care, as I spend my time using the real IPA, not crude ASCII approximations.

/ei/ and /Ou/ aren't phonemes because they contain more than one symbol, for one thing. There's no difference between [e] and [eI] in English; you can represent the corresponding phoneme as /e/--the /i/ is redundant.

There are only three phonemic diphthongs in English, meaning that there are only three diphthongs that MUST be pronounced as diphthongs (and thus must be transcribed this way). When I transcribe them, I always put a non-syllabic diacritic under the second vowel to make it clear that they are two vowels pronounced as a single diphthong, and not two independent monophthongs (which would be the default assumption in IPA).
Jim   Thursday, September 16, 2004, 05:18 GMT
It's no good trying to use the IPA on the net as you might have noticed. I'm afraid that we're stuck with these "crude ASCII approximations." I find it useful to use the same one as everyone else when writing on Antimoon but I don't mind if you don't agree though if I'm using something else I always make a point of saying so.

This is the Kirshenbaum vowel table from
http://www.hpl.hp.com/personal/Evan_Kirshenbaum/IPA/faq.html

________----- unr -----_________unr_______----- rnd -----
________fnt____cnt___bck______cnt______fnt____cnt__bck
_____________________________rzd__________________
hgh______i______i"_____u-______________y______u"___u
smh______I____________________________I._____U____
umd______e_@<umd>__o-_____R<umd>____Y_________o
mid____________@______________R_____________@.___
lmd________E____V"___V__________________W___O"__O
low_________&____a___A___________________&.__a.__A.

hgh high/close
smh semi-high
umd upper-mid/close-mid
mid mid
lmd lower-mid/open-mid
low low/open
fnt front
cnt centre
bck back
unr unrounded
rnd rounded
rzd rhoticised

So just as in SAMPA:

[e] is the close-mid front unrounded vowel,
[o] is the close-mid back rounded vowel,
[I] is the lax close front unrounded vowel &
[O] is the open-mid back rounded vowel.

"/ei/ and /Ou/ aren't phonemes because they contain more than one symbol," says Mxnmanic.

What relevance does it have how many symbols one uses to transcribe them? I could easily use a transcription system whereby they are transcribed as /A/ and /O/. In Shavian and the Pitman Initial Teaching Alphabet (ITA) they are transcribed with one symbol. One could well choose to transcribe them using Chinese characters. Indeed phonemes existed before writing.

http://www.omniglot.com/writing/shavian.htm
http://www.omniglot.com/writing/ita.htm

Then on the other hand, the vowels in "boy", "bough" and "buy" are transcribed, in Tom's alphabet, as /oi/, /au/ and /ai/. Notice something? They contain more than one symbol but you're not saying that these aren't phonemes. Again, that they contain more than one symbol is another simple matter of choice of transcription system. I could choice to write them /Y/, /W/ and /I/ if I liked. They only need one symbol in Shavian and the ITA

So what is the significance of there being two symbols in Tom's transcriptions of these vowels? In making his alphabet Tom attempted to base things on the IPA as used in various dictionaries. In most dictionaries which use IPA transcriptions /ei/, /Ou/, /oi/, /au/ and /ai/ are written with two symbols. Why? In most dialects these vowels are diphthongs. IPA transcriptions of diphthongs must use more than one symbol.

"There's no difference between [e] and [eI] in English;" sure, as much as there is no difference between "get" and "gate", "wet" and "wait", "pen" and "pain", "men" and "main", "west" and "waste", etc. If you're teaching your students that these are homophones rather than minimal pairs, Heaven help them.

There very much is a difference between [e] and [eI] in Australian, New Zealand, British RP and General American English. In Australian, New Zealand and British RP English (Tom's) /e/ (the vowel in "met") is pronounced as (Kirshenbaum's) [e] in General American it's (Kirshenbaum's) [E]. In RP and GA (Tom's) /ei/ (the vowel in "mate") is pronounced as (Kirshenbaum's) [eI]. In Australian and New Zealand English it's (Kirshenbaum's) [&I].

What strife you'd get into if you couldn't tell the difference between "I met my mother." and "I mate my mother." In the Californian accent things might be a little different. I've read that (Tom's) /ei/ is the monophthong [e:] (Kirshenbaum's) but even here it's dictinct from (Tom's) /e/ which is the monophthong [E] (Kirshenbaum's).

http://www.ling.mq.edu.au/units/ling210-901/phonetics/ausenglish/auseng_vowels.html
http://www.ling.mq.edu.au/units/ling210-901/phonetics/vowelgraphs/USE_Monophthongs.html

Now, in my accent there is no monophthong that could possibly replace the diphthong [&I]. If an Aussie says [get], (s)he means "get" never "gate". If an Aussie says [re:], (s)he means "rare" never "ray". The vowel in "gate" and "ray" MUST be pronounced as a diphthong in Australian English and it must thus be transcribed as two (or possibly more) symbols if you're using the IPA (but not if you're using the ITA).

The same goes for /Ou/. In Aus. Eng. /Ou/ (the vowel in "coat") is [@u"] no monophthong would do. Try [kO:t] and people would probably think you meant "cot" which is actually pronounced [kOt] in Aus. Eng. Try [kot] and people would think that you meant either "caught" or "court" which is actually pronounced [ko:t] in Aus. Eng. It's much the same in other dialects of English except Californian.

The Californian /Ou/ is the monophthong [o:] but it's still distinct from /o:/ (the vowel in "caught") which is pronounced [kAt]. So sure, there are accents in which /ei/ and /Ou/ are not diphthongs but they're always phonemes.
Jim   Thursday, September 16, 2004, 07:12 GMT
Mick,

ESL teaching.
Mi5 Mick   Thursday, September 16, 2004, 07:14 GMT
Ah, good on you.
Ben   Thursday, September 16, 2004, 17:10 GMT
Somebody asked earlier if there were any sounds RP was short of.

There is one that I can think of:

The strong /O/ sound that your hear in Irish, Scottish and Middle American speech is missing (usually the dipthong [Ou] is pronounced [..u] or [^u] in RP.
Mr. Phones   Thursday, September 16, 2004, 22:58 GMT
I asked if there were any phonemes ''not sounds'' that RP was short of.


That strong [O] sound that you mention is the way Scottish and Irish pronounce the /Ou/ phoneme.

Ben, I think you think that If something is not a monothong but a diphthong then it's not a phoneme. That's not true. The British diphthong /Ou/ ''i.e. [..u]'' is just as much of a phoneme as the Scottish and Irish monothong /Ou/ ''i.e [O]''.

''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''

There both the same phoneme /Ou/ but different phones [O] and [..u].
Eng.   Friday, September 17, 2004, 02:02 GMT
the General American and RP accent both lack the [W] phoneme. So, RP is not the accent with the most phonemes (although some people claim it is).
Jim   Friday, September 17, 2004, 06:43 GMT
Not if those dialects which distinguish /w/ and /W/ also merge /u/ and /u:/, for example.