[Spellyans] Problems with SWF

[Michael Everson]

At 12:53 +0300 2008-06-26, Owen Cook wrote:

>A) <u> for MC /y(:)/, which becomes LC /i(:)/. This is the most common
>value; we can take it as default.

Yes.

>B) <u> for early MC /y(:)/, which becomes /Iw/ rather early on. We
>find this in stressed open syllables (as in 'du') and before /h/ (as
>in 'uhel', 'a-ugh').

I don't think we have to mark these, as they can 
be listed exhaustively. How do you transcribe 
'uhel' and 'a-ugh'?

>C) <u> for /u:/ as in frût, gûn. KS has used <û> here, with which I
>heartily concur.

Yep.

>D) <u> for /U/ or /7/ or something similar, as in 'pùb'. KS has used
><ù> here; again I heartily concur. (By-the-bye, I think I've seen KS
><ù> in unstressed syllables as well as stressed ones -- is this
>right?)

Yep; it is quality that is being marked.

>E) <u> in the word 'usya', which may be /ju/ or 
>/Iw/ or who knows what. I keep suggesting that 
>we press <uw> into service here, but nobody's 
>commented so far. Gendall evidentally feels that 
>'universita' belongs in this category too. KS 
><û> seems quite unsatisfactory to me here.

I hate to disagree with you... but I think it is 
easiest to use the standard mark for long [u:] 
and simply say that this word (I am not sure 
about <ûnyversyta>) has a special pronunciation.

>If <uw> seems too letter-heavy, perhaps a diacritic is necessary?

<uwnyversyta>? <uwsya>? This would lead to 
another divergence from the SWF (and from UC, 
UCR, and KK) for little yield. Using the 
circumflex treads more lightly.

>I have always wondered why nobody has seemed to 
>be in a mood to consider an umbrella graph for 
>these items.

Because none suits.

>The situation seems similar to our dydh/dedh 
>words -- there's an awful lot of words with s~j 
>alternation. <z> wouldn't pass the authenticity 
>test, I suspect.

Yes, and I guess this is just going to have to be Something To Learn.

>I've always liked <zh> for this situation; it 
>evokes Lhuyd, but would not make anybody else 
>happy. However, I'm sure any solution would be 
>too radical for our present undertaking.

I agree.
-- 
Michael Everson * http://www.evertype.com