USAGE: /t/ (was Re: USAGE: Schwa and syllabification

Ray:
> Of course in the London area the 'dark l' has become [w] so the word
> sounds more like ['lI?U] - I'm probably not spot on with that last sound
> and no doubt And or Joe who are more familiar with 'Esturine English' will
> be more accurate  :)
_skittle_ is [skI?o], _little_ [lI?o] or [lIo].

> > People are probably sick of my pointing out that this or that bit
> > of linguistic exotica or impossibilia occurs in one dialect or
> > another of English... But here goes anyway:
> [snip]
>
> Nor do you just have to make do with a Brit English dialect for an example
> either. The change from [t] to [h] did happen in Gaelic. In Scots Gaelic
> it's spelled |th| and also in Irish when written in standard modern Roman
> letters; but in the Irish alphabet it's just |t| wih a dot on it.
>
> It's true not all instances of /t/ went this way, but those subject to
> 'soft mutation' did. The steps are:
> [t] --> [t_h] --> [T] --> [h]
Scouse would have followed a similar path, except that the coronal
segments are apicoalveolar rather than laminodental. (So _pith_,
_pit_ and _piss_ all end in fricatives and all contrast.) Because
the step to [h] was lexically selective, you end up with contrasts
like _not_ [nQh] vs _knot_ [nQ$] (where $ is a fricative one
might symbolize in IPA by [t] with an openness diacritic added).
(All things considered, I'd say that _not_ has allomorphs /nQt/
and /nQ/.)

--And.