Re: Why is walk spelt that way ?

On Jun 9, 2:02 pm, mb <azyth...@xxxxxxxxx> wrote:
On Jun 9, 11:20 am, "ranjit_math...@xxxxxxxxx"
...

wouldn't solve the problem of how to pronounce an /a/ that's not
silent since /a/ has realizations of [E],[V],[@] and [i"].

But not of [a] (non-English [a])?

[V] represents the "non-English" [a]; [a"] might be an alternative
representation. Using [a] seemed confusing because the British
realization of /&/ is sometimes described as [a]. Non-English [a] is a
confusing term too; French [a] is non-English but isn't used in Hindi.

If that is true, why call it a
phoneme /a/, of all things?

Good question. Perhaps a phonemic reanalysis is called for. Malayalam /
a/ and /i/ both have allophones of /E/. Is it possible for two
phonemes to have an allophone in common?

Or of zero?

What is transcribed as zero is not always zero. "apna" is pronounced
[api"nA].

By the way, is this really one Hindi phoneme /a/,

Good question. I'm not sure whether a phonemic reanalysis would leave
it as one vowel. /mahal/ can pronounced [mE:h@l] where [E:h@] means a
long breathy diphthong [E@]. What would that turn /a/ or /h/ into in a
phonemic reanalysis?

or just the
different sounds of the grapheme equivalent to <a>, and that one
equivalent to different phonemes, or part of?

It's an unmarked vowel. It is marked only initally, where it's always
[V]/[a"].

Also, did the idea of representing all these sounds with <a> in
Western transcription originate with Sanskrit?

Yes. It also has to do with a shortage of Latin vowel graphemes. For
example, <e> is already in use for [e:], so it's not available for
[E].

Asking for details because I have not the foggiest idea.

.

Quantcast