Odp: [tied] Re: House and City

--- In cybalist@..., Miguel Carrasquer Vidal <mcv@...> wrote:

> On Sat, 24 Mar 2001 13:56:19 -0000, tgpedersen@... wrote:
>
> >--- In cybalist@..., "Piotr Gasiorowski" <gpiotr@...> wrote:
> >> in time from MP to PA, the striking similariy of the word for

"2"

> >> (most impressive for the form <dua>) to its IE counterpart

begins

> >> to fade somewhat.
> >The obvious retort to that is that these words were borrowed into

IE

> >and AA from MP, not PA. (Ha! Gotcha!)
>
> I'm not familiar with the PMP reconstruction, but judging by the
> attested forms (www.zompist.com/anes.htm), it must have preserved at
> least a trace of the /s/ in PAN *dewsa. I'd say *duha (Bisayan
> <duha>), if not *dewsa (Tominin <deisa>).>

To my knowledge, the current reconstruction is *DuSa, where *D
represents a probable retroflex stop (Formosan languages allow the
reconstruction of at least 2 other d-sounds with very mixed reflexes
in all other MP languages). The *S is one of our so-called
laryngeals, >Formosan s, s^ usually, Philippine h, most other MP
zero. The form *DewSa or older *Dewha was proposed by I. Dyen back
in the 60s and has been abandoned by all but his partisans: it
served to explain only l or 2 aberrant Phil. forms, and posited the
unique pre-consonantal sequence *-ew- which occurred only in this
form.

I personally have never been totally convinced by the (large) number
of "laryngeals" posited on the basis of Formosan evidence only. This
remains a rather controversial area in AN studies.