2010-09-19 18:08, Alex Fink skrev:
> On Sat, 18 Sep 2010 13:27:48 -0700, David Peterson<[log in to unmask]>  wrote:
>
>> In Finnish, there's a sound change which produced a number of
>> Finnish's characteristic diphthongs. Specifically:
>>
>> SAMPA
>> e:>  ie
>> 2:>  y2
>> o:>  uo
>>
>> Modern Finnish, however, has a full set of long and short vowels,
>> including long mid vowels.
>>
>> My question: Where did the modern long mid vowels come from?
>
> Since John V hasn't stepped in yet I can at least point to his Finnish
> diachrony page on Frath
>    http://wiki.frath.net/Finnish#Proto-Finnish_to_Standard_Finnish
> section "Subsequent vowel changes in unstressed syllables" which says that
> at least some of them are reductions of diphthongs in that position.
> Wikipedia says long mid V are frequentest in unstressed position, so that
> might be the main sourse.

Actually the biggest source of new /e:/ /2:/ /o:/ was probably
the loss of intervocalic /h/, /?/ and */G/ in unstressed
syllables. (*/G/ was lost or changed to /v/j/ *everywhere*, but
it only occurred intervocalically/unstressedly, being the
morphophonemically conditioned weak grade of /k/, and /?/
was lost or at least never written, depending on dialect) e.g.:

_tehdä_   'to do'             < */tekt&/
_tekee_   's/he does'
_teen_    'I do'              < */teGen/

_meri_    'sea'               < */mere/
_mereen_  'into (the) sea'    < */mereh@n/
_talo_    'house'
_taloon_  'into (the) house'  < */talohon/

but /h/ is preserved after a long vowel/diphthong:

_maa ~ maahan_  'land' nom. ~ illative
_tie ~ tiehen_  'road'

There is a third illative allomorph _-seen_ used with
polysyllables ending in long vowels which AFAIK must
be the result of analogy, since intervocalic /h/ was
itself the result of */S/ > /h/, but the (pre) history of
sibilants is very convoluted and I'm sure the old
cannons had some different more convoluted explanation!
There may be an *S/*tS alternation lurking somewhere.

Then there are some freak occurrences in words which used
to end in */k/ or */h/:

perhe < *pereh         'family'
perheet < *perehet     'families'
perheeseen < *perehesehen (I believe < *perehes@n)

kirje < *kirje? < *kirjek     'letter'
kirjeet < *kirje?et/*kirjeGet < *kirjeket

Some dialects/colloquial styles also change /oa ea e&/
into /oo ee ee/, which frequently affects the partitive
singular ending, but that is not reflected in the
standard written language.  Social realist novelist
may be another matter though!

NB I don't have the books to hand, and as usual it's exactly the
relevant pages which are blocked on google books so details may
be wrong, but the patterns are right, modulo my doubts wrt the
traditional account of the allophony of the illative ending.
Something *very* hairy was going on there for sure, and the
details may be irrecoverable.

I'm also unsure to what extent unstressed diphthongs
monophthongized, and to what extent new unstressed
diphthongs arose through loss of intervocalic /h ? G/.

Further there seems to be some t > 0 / V_V, the details
of which I'm not filled in on, as evidenced by the
allophony of infinitives and partitives. IIRC there
occurred both *t > *D > d and *D > *d > t at different
times and in different contexts, possibly also *T > *h > 0,
though the norm is *T > ts (which may be a spelling pronun-
ciation, since /T/ used to be spelled _z_ os _tz_. There
may have been a /T/~/TT/ distinction; my recollections are
hazy.

/bpj