Print

Print


Wholly and fully have the same length of /l/ for me, FWIW. I suppose if I
were to need “tall+ly” for some reason the morpheme boundary might trigger
germination, as it does in “bookkeeper”..
On Mon, Feb 5, 2018 at 10:23 And Rosta <[log in to unmask]> wrote:

> On 5 Feb 2018 14:42, "Mark J. Reed" <[log in to unmask]> wrote:
>
> On 04/02/2018 16:31, And Rosta wrote:fds
> >
> >> For you? How common is that? The t-less stem occurs in most (but not
> all)
> >> dialects in eigh-ty and eigh-teen, but I'd not known it can happen in
> >> eigh-th.
> >>
> >
> I don't recall ever hearing "eighth" as /eɪθ/, FWIW.
>
>
> Alex's dad says it, apparently. Good datum (eighth rhyming with faith, I
> mean, not the fact that it's Alex's dad).
>
> Alex tells me that 18 in N America is in fact usually eight-teen. I was
> surprised, because to my ears it sounds very Northern (English).
>
> Usually it's /eɪʔθ/.
>
> Nor do I see any deletion in eighty and eighteen, just a merger of the two
> adjacent /t/s (which is automatic since most ModE lects don't have phonemic
> gemination).
>
>
> It isn't uncomplicatedly automatic. You do get, say, fu-lly, but you also
> get whol-ly, and nonce -ly forms from l-final adjectives have double l-l.
> Un- and -ness would always geminate. Irregularly 'degeminated' forms have
> to, in effect, be lexically specified. Possibly the incidence of gemination
> is connected to the presence or absence of phonological word boundaries,
> but they can't be located in a pretheoretical way: in some lects
> degemination might follow from the absence of a phonological word boundary,
> but the incidence of phonological word boundaries is not always
> predictable. At any, do bear in mind that some dialects have eight-teen and
> eight-ty.
>
> --And.
>