[Spellyans] The quantity system

[Michael Everson]

At 16:01 +0000 2008-06-24, Jon Mills wrote:
>In the 2 pronunciations [Os'tri:@] and ['Ostrja] the word stress is 
>placed on different syllables.

Yes, and spelling them differently enables the learner to predict the 
stress. <ia> is always stressed. <ya> is always unstressed.

>So we have [i:@] in a stressed syllable and [ja] in an unstressed 
>syllable. Given that ryal/rial is monosyllabic, the diphthong would, 
>according to this rule, be [i:@]. However that implies that the 
>diphthongs [i:@] and [ja] are in complementary distribution.

Possibly, though I wouldn't jump to conclusions.

>If that is the case, then we can write <ia> for both and say that 
><ia> is pronounced [i:@] when it is word stressed and [ja] when it 
>is word unstressed.

You're not suggesting doing away with <ya> are you?

We could do what you suggest (if you proved that they really were in 
complementary distribution) but I don't think it would be helpful to 
learners or writers.

The SWF is using this same distinction: it writes bian, lien, lies 
have [i:@] as described in KS16 1.5.8; also in the Conditional 
(Pluperfect) of bos.

>Provia [pr@'vi:@] 'provide' and profya ['prOfj@] 'offer' are another 
>good example of complementary distribution.

Yes, but distinguishing <ia> and <ya> enables the learner to know how 
to pronounce these words. I wouldn't recommend provya/profya or 
provia/profia as these are ambiguous. i/y allows us to make this 
useful distinction -- why not?
-- 
Michael Everson * http://www.evertype.com