[Spellyans] i ~ y

[Michael Everson]

On 3 Feb 2009, at 14:01, Daniel Prohaska wrote:

> The SWF doesn’t, unfortunately, go by the attested spellings, but  
> rather assumes a context conditioned variation of [ð] and [θ] in  
> unstressed syllables where etymologically /ð/ is expected, i.e.  
> British /d/ or intervocalic /j/. As far as I can recall this was an  
> AHG decision and it is mentioned in the final specification.

More of a waffle than a decision.

Michael Everson * http://www.evertype.com