Re: THEORY: Ray on ambisyllabicity

Nik Taylor:
> And Rosta wrote:
> > Yeah. We need to ask Dirk how he accounts for:
> >
> >   Sal   [saw]
> >   Sally [sali]
> >
> > in demotic SE Insular English, if /l/ isn't in an onset in "Sally".
> > Also:
>
> I've noticed that all these different arguments are using different
> dialects.  Is it not conceivable that in some dialects it *is*
> ambisyllabic, while in others it isn't? Or that some dialects have
> /h&p.i/ while others have /h&.pi/, and perhaps some with ambisyllabic
> consonants?
Yes. But the nonambisyllabics would be unremarkable, since there is such
a superabundance of data that do not call for an ambisyllabic analysis.

> In my dialect, for instance, I can find no evidence to
> support the idea of the "ambisyllabic" consonants as even being onsets,
Do you have affrication of /t/ /d/ before /r/?

> yet it seems that in this dialect it is an onset.  I wonder, does that
> dialect have evidence supporting /l/ as being a coda at all in Sally?
First, there's the evidence (from glotalling) that /t/ in the same
position is a coda. Second, there's the putative generalization that
all stressed syllables must be 'bimoraic'. Third, there's the holy/wholly
contrast where /@U/ has its [OU] allophone only in the latter, which
suggests that /l/ is a coda only in the latter.

I'd like to add that the evidence for "ambisyllabicity" is evidence for
a segment being sensitive to its structural position with respect to both
preceding and following material. The analysis doesn't have to involve
literal ambisyllabicity. For example, Government Phonology rejects literal
ambisyllabicity (a segment belonging to both a coda and an onset), but it
allows other nonconstituency relations between onsets and codas.

--And.