*gwh in Gmc.

[Miguel Carrasquer Vidal]

On Sun, 25 Feb 2001 09:03:49 -0000, "petegray"
<petegray at btinternet.com> wrote:

>> This looks suspiciously like the Caland pattern.

>I'm both intrigued and ignorant - as normal.   Could you outline the Caland
>pattern, please?

Caland (in 1892/93) noticed that certain Indo-Iranian adjectives in
-ra- and -ma- change that ending to -i- when thay are the first part
of a compound:

Av. d at r@z-ra- "strong"  => d at r@ez-i-ratha "having a strong chariot"

The phenomenon also applies to adjectives in -u, and examples can also
be given from Greek:

Grk. ku:d-r-os "famous", ku:di-aneira "bringing fame to men"

Another thing to note is that the pattern usually also includes an
s-stem neuter (e.g. Grk. <ku:dos>).

The pattern has usually been considered a "Suffix-verband", i.e.
merely a group of suffixes that "go together", but the possibility of
a set of sound-laws behind the pattern has to my knowledge never been
proposed.

What I would suggest is that these forms are in origin athematic
adjectives in *-n (secondarily thematicized).  Besides the normal
primarily thematic pattern in *-no-, we would have:

[old abs./acc.?]  *-n(a) > *-r (> *-r-o-)
[old erg./nom.?]  *-n(u) > *-u or *-m (> *-m-o-),

with a "status constructus" [old gen.?] in *-n(i) > *-i.

=======================
Miguel Carrasquer Vidal
mcv at wxs.nl