Re: accents

On 30 Mar 2005 12:49:17 -0800, Iain
<iain_inkster@xxxxxxxxxxx> wrote in
<news:1112215757.499082.9150@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>
in sci.lang:

> Peter T. Daniels wrote:
>> Iain wrote:

[...]

>>>>>>>> Wasn't it Iain who showed up recently with the
>>>>>>>> claim that English doesn't have any shwas?

>>>>>>> Shwas feature in speech but they aren't part of the
>>>>>>> code(language) as such -- A Scot says a shwa where
>>>>>>> someone else doesn't, etc.

>>>>>> We'd appreciate seeing your phonemic analysis of
>>>>>> English that manages to do without shwa.

>>>>> Well of course shwa features as a sound in English --
>>>>> that's obvious -- but not uniformly placed like other
>>>>> vowels tend to be, which sets it apart.

>>>> We have no idea what you're saying.

>>>>> I'll clarify. In a pronunciation dictionary, "get"
>>>>> only has one pronunciation. However, the vowel sound
>>>>> in "get" sounds different worldwide but everyone says
>>>>> it the same way they do in "pet", which is why it
>>>>> only deserves one pronunciation in the dictionary.
>>>>> The dictionary is basically saying "just say pet so
>>>>> that it rhymes with get, however you say it".

>>>> Correct.

>>>>> But "girl" and "curl" don't rhyme when I say them.
>>>>> "world" and "twirled" don't rhyme when I say them.

>>>> All four have the same vowel (which isn't a shwa);

[...]

>>>> "girled" and "curled" are perfect rhymes with each
>>>> other and with "world" and "twirled." If they don't
>>>> rhyme for you (and you're Scottish, so presumably
>>>> you're rhotic), they have different syllabic nuclei
>>>> from each other.

>>> If by the syllabic nuclei you mean the vowels as
>>> distinct from the Rs, yes. I say them rhotic as an
>>> American, but with a U and I like in "cut" and "piss",
>>> respectively -- "G-ih-rl", "Kuh-rl".

>> The "Rs" are part of the syllabic nuclei. If I had meant
>> vowels, I would have said vowels. Your phonemicization
>> of those words is presumably /girl/ and /kVrl/; neither
>> of them seems like what the several Scotsmen I have
>> known would say.

> Not all Scots speak the same. My accent is from Troon,
> with a lot of Highland thrown in.

> Highlands -- girl

[gIrl]

> Lowlands -- G@rl

> England -- G@l.

[gV":l], with [V"] = reversed epsilon, I suppose.

Not universally: there are plenty of rhotic varieties of
English in England.

>>>> But what's that got to do with English having or not
>>>> having (a phoneme) shwa?

>>> See below.

>>>>> "Early" has no shwa for me. Even

>>>> Nor for me. Which syllable are you referring to?

>>> Shwa is Ә, right?

>> I see cap-O-acute followed by tilde, so I don't know what
>> you typed there. The ASCII equivalent for the shwa
>> symbol ("turned e") in the scheme usually used here is
>> @.

He typed in a genuine schwa.

[...]

>>> Dictionaries of mine show a shwa at the end of "finger",
>>> and the same sound at the start of "early". My ears
>>> don't challenge that when listening to RP, but neither
>>> are they shwas for me.

The OED has <early> ['V":lI], <finger> ['fINg@(r)];
different vowels.

>> You are consulting dictionaries of RP, which to me are
>> useless.

I don't know what he's consulting, as I'd hardly expect a
dictionary of RP to confuse [V":] with [@].

> That relates to my point. So can we clarify? Dictionaries
> of RP are usless why?

Read it again: useless _to_Peter_.

> Is one correct dictionary possible for most
> pronunciations? If so, can it correspond with bits of
> spoken RP words? Is RP inconsistant with other
> pronunciations?

Of course it is. RP doesn't have the same phonemes as SSE
(Scottish Standard English), and the phonemes distributed
differently; here are some correspondences as given by April
McMahon.

RP SSE
---------
i: i beat
I I bit
eI e bait
E E bet
A: a balm
æ a bat
O: O bought
A. O bomb
u: u food
U u foot
oU o boat
V V but
aI aI bite
aU au bout
OI OI boy

V": Ir bird
V": Vr word
V": Er heard
I@ ir beer
E@ er bear
A: ar car
O: ur poor

> Would you say there is a shwa sound in RP's "early"?
> (unlike other accents)

No. The sound is [V":].

> Would you say there is a shwa sound in RP's
> "entrepreneur"?

If you mean the RP pronunciation that doesn't nasalize the
first vowel or otherwise attempt to approximate the French
more closely, there are two schwas: it's [,A.ntr@pr@'nV":].

> (you said yes) Why is one useless as an example of English
> pronunciation and the other not? Is RP's girl therefore
> straying from Standard English? Why? Because of the shwa?
> Because for that one word at least, shwa isn't part of
> the language, right?

This is grotesque. You might as well say 'For the one
number 2005 the digit '3' isn't part of the language'.

>>>>> "Entrepreneur" has no shwa for me. "Policeman" _does_
>>>>> have a shwa for

>>>> "Entrepreneur" has two shwas, in the middle syllables.
>>>> What do you have?

>>> Mine rhymes with "manure". Ohn-treh-preh-nyoorr,

Given the imprecision of this transcription, it's entirely
possible that your \eh\ actually *does* represent a variety
of schwa. No one has a schwa in the last syllable, so the
rhyme is irrelevant.

>>> with subtly bigger emphasis on the penultimate syllable
>>> than usual. Mine is a *** Highland accent. But I do
>>> say shwa in "entrepreneurship".

>>>>> me, but if I'm asked to say the word slowly it'd come
>>>>> out as an /A/.

>>>> Seems to me if you had to say "policeman" _that_
>>>> slowly, it would be [o].

The two of you are at cross purposes: Iain is talking about
the last syllable, Peter about the first, which is generally
[p@-] in U.S. English. I see that the OED acknowledges both
[p@'li:sm@n] and [p@U'li:sm@n].

[...]

>: -) No no -- I was talking about policem [@] n.
> I just asked my brother to say it slowly and clearly and
> it came out as "Ah".

Asking him to say it slowly and clearly is exactly what you
should *not* have done to elicit a normal pronunciation.

[...]

>>> All I'm really saying is that shwa is different from
>>> other sounds in English because it tends to disappear
>>> with emphasis, unlike "wet", which can bear massive
>>> emphasis without phonemic alteration.

>> Shwas can't "disappear" with emphasis, since they are by
>> definition in unstressed syllables.

> Right, so when the syllable is stressed, the shwa is no
> more! I said "disappear with emphasis" -- i.e. It
> disappears because of emphasis.

When a normally unstressed syllable is stressed, you no
longer have the same word. The noun <object> is commonly
['AbdZ@kt] in U.S. English; if you stress the second
syllable, the schwa is replaced by [E] and you get the verb
<object> [Ab'dZEkt], a different word.

> Emphasis only has denotation in words such as "excuse",

There's no stress shift between the noun <excuse> and the
verb <excuse>: the difference is that the noun has [s] and
the verb [z].

> etc but shwa has even less -- Shwa seems to have little
> or no denotation in language that cannot be replaced by
> the un-neutralised vowel.

This is a misuse of 'denotation'. I think that you're
trying to say that schwa can always, or at least almost
always, be replaced by some other vowel, one that can occur
in stressed syllables, without changing the meaning of the
word. I doubt it, but suppose for the sake of argument that
this it's true; you might get understandable speech, but it
wouldn't sound like any native variety of English.

> In other languages, the same sound is not always a
> neutralisation.

(Reduction; neutralization is something altogether
different.) Nor is it generally a reduction in English,
synchronically. It's true that diachronically many schwas
have arisen from reduction, but the first vowel of <about>
and the last vowel of <sofa> are schwas in Present-Day
English (PDE); use some other vowel, and you'll sound
foreign.

[...]

>>>>> Shwa isn't a propor stead like the middle of "get".

>>>> Usually your typos are self-correcting, but I don't
>>>> know what <propor stead> was meant to be.

>>> There is a parallel associated with E along which "get"
>>> and "wet" rhyme in various locations worldwide, but a
>>> parallel involving shwa doesn't stretch as far as where
>>> I am sitting.

>>> That's what I mean by "stead". Shwa is not steady.

Neither the OED (s.v. <stead>) nor the Concise Scots
Dictionary (s.v. <steid>) notes such a usage.

>>> It can crop up anywhere, and when it goes, the original
>>> vowel returns, hence the /a/ in policeman.

No, it can't crop up anywhere. There is no possibility in
PDE of pronouncing the middle syllable of <policeman> with
schwa. (Oh, you can *say* it, but it isn't English.) And
in the varieties of U.S. English with which I'm familiar,
using the /a/ -- typically [æ] -- of <man> in <policeman>
sounds extremely odd and unnatural.

[...]

>> If, I say again, you produce a phonemic analysis of
>> English that manages to do without shwa, I will be able
>> to see what you are saying.

> Shwa's status as a phoneme: Isn't varying amounts of
> neutralisation a facet of accent? If shwa is a
> neutralised vowel,

It isn't.

[...]

Brian
.