THEORY: ambisyllabicity & gemmination

On Sun, 8 Oct 2000, Raymond Brown wrote:

> But what I'm really wanting to see is how Dirk's notation which he used for
> Japanese _hatten_, works out with English _happy_ and Welsh _hapus_ where,
> although the {p} is written only once, it is certainly geminated and the
> English {pp} is not.  The English is ['h&pi] or ['hapI], depending on
> dialect, whereas the Welsh is ['happ_h1s] or ['happ_h1s] depending on
> dialect.
Your wish is my command :-); here you go.

        'happy'      'hapus'

          s   s        s     s        syllable
         /|\  |       /|\   /|\
        / m m m      / m m / m m      mora
        | | | |      | | |/  | |
        h & p i      h a p   1 s      segment

It should be remembered that I consider 'happy' to be syllabified
after the /p/; the transcription also reflects a General American
pronunciation rather than an Insular one. For the northern dialects
which you mentioned, I don't know enough to talk about the
syllabification properties of medial consonants. My hunch is that it
receives a less marked structure

        'happy'

          s   s      syllable
         /|  /|
        / m / m      mora
        | | | |
        h a p I      segment

But I just don't know for sure.

The representation of the Welsh example should be pretty clear by now.

Dirk

--
Dirk Elzinga
dirk.elzinga@m.cc.utah.edu