The short /ö/ is young and has diverse genesis.
In some suffixes [ö] occurs as an allophonic variant of /o/
conditioned by vowel harmony. As an allophonic phenomenon this might
be very old, i guess, insofar there were old suffixes with -o-.
Examples: nime-tön 'nameless, anonymous' and syy-tön 'not guilty,
innocent' versus raha-ton 'without money, bancrupt'
At some point, certinly very late, I guess late in the first
millennium AD or even later a new stemfinal vowel -ö has occured in
derived words as a front vowel version of the relatively older -o.
Examples:
säästö 'saving' from säästää 'save, spare' in analogy of derivations
like paahto 'roast' from paahtaa 'to roast'.
Also very late, in the middle ages I believe, the glide *-eü- has
become -öü- (in Finnish).
Example: löytää < *leütä- 'find' <= Gmc *xleuta- 'get as your share,
get by chance or by heritance etc.'
Short /ö/ in the nucleus of the stem is still today limited mostly to
words, which fennicist call "descriptive" (that is a concept slightly
wider than, but related to, "onomatopoetic").
Example: töllöttää 'stare' and töllötin 'television' (coll.)
There might also be cases where an original back-vowel stem, such
as 'Coc(C)a' has spontaneously mutated to a front vowel one 'CöC(C)
ä'. This sort of mutating is well attested generally in the language,
and it may also have contributed to the stemfinal -ö described above.
I wonder if törmä 'river bank' could be an example of such mutation
(I don't have an etymological dictionary here at lunch break)?
Needless to say that the system with strict vowel harmony, which is
inherited from Uralic, is very difficult to reconcile with any german
style umlaut. The previous paragrapgh shows how the vowels in the
first and second syllables always move in tandem, and therefore there
cannot be the sort of tension which would launch an umlaut.
The only possibilty of something remotely similar is a conditioning
of a vowel by the adjecent consonant. And indeed Pre-Finnic knows of
this phenomenon, in the case of labialisation, producing a long /ö:/.
The phenomenon is for sure a bit older than the genesis of short /ö/
described above. Certainly the existance of a (Middle or Late) Proto-
Finnic long phonemic /ö:/ cleared the ground for the development
above.
The long /ö:/ (today /-yö/ due to a recent sound development) exists
in old Uralic or Pre-Finnic words that have been contracted from two
syllables to one:
Examples: lyö- < *lö:- < *leXi- or *lewi- 'hit, blow'
syö- < *sö:- < *sewi- 'eat' <= Proto-Iranian cognate to 'chew'
Jouppe
--- In
cybalist@yahoogroups.com, "Patrick Ryan" <proto-language@...>
wrote:
>
> Jouppe,
>
> and did *ö come about as a result of *l being followed in the next
syllable
> by a front vowel?
>
> or how else?
>
>
> Patrick
>
>
> ----- Original Message -----
> From: "jouppe" <jouppe@...>
> To: <cybalist@yahoogroups.com>
> Sent: Wednesday, February 20, 2008 1:05 AM
> Subject: [tied] Re: Finnish KASKA
>
>
> Patrick,
>
> *ü was part of the inventory.
>
> **ö was not.
>
> Jouppe
>
> --- In cybalist@yahoogroups.com, "Patrick Ryan" <proto-language@>
> wrote:
> >
> > Jouppe,
> >
> > let me ask it another way:
> >
> > in your view were *ö and *ü part of the earliest PU vowel
inventory?
> >
> >
> > Patrick
>