--- In
cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...> wrote:
> There's a famous article the name and authos of which I forgot
which
> investigates the distribution of 'huis'/'muis' "house"/"mouse" in
> Dutch; the vowel developped this way: *u: > *ü: > *ui > *äÜ (vel
> sim.!).
>
> The interesting part was that in some places it
was 'huis' "house",
> but 'mu:s' "mouse". In other words, since the "mouse" word belonged
to
> a 'loer' sphere, it stayed behind in the prestige-driven
development
> of the vowel.
It's famous because Leonard Bloomfield discussed it in _Language_ in
the chapter on "Dialect Geography". (In my edition the map is on p.
328), on the basis of work bij G. Kloeke. As Torsten relates, it is
about two innovations: an automatic fronting of long *u: to ü: and a
subsequent diphthongization of ü: to [öü] or [äü] or something along
those lines paralleled by a diphthongization of *i: to [ei] and
comparable with similar diphthongizations found in English and
German.
The eastern periphery retains the original *u:, all the rest has
fronting. Most of the fronting area has undergone diphthongization
too, but there are large peripheral areas on both the east and the
west that still have monophthongs. Between the eastern monophthongal
area and the archaic area with retention of *u: there is a large area
with inconsistent reflexes showing innovating items with fronting
(hü:s) on an archaic non-fronting background (mu:s).
(Bloomfield - and no doubt Kloeke also - doesn't say that those
systems can do that because they have a phoneme /ü:/ from umlauted
*u:, so there was a point of entry. That is the reason the diphthongs
can't spread piece-meal in similar fashion: systems that have not
undergone diphthongizaiton don't have the diphthong, so there is no
point of entry.)
W.