On 3.9.2007 Paul Bennett wrote:
> It is, in short, the kind of environment in which I
> suspect my nicely-laid-out Germanic vowel grid just would
> not stay stable. Worlds are colliding, and things are
> going to collapse. The only problem is: which things?
> It seems that the front-rounded vowels would need to go,
> and one or more of the back rounded ones, too. Where from
> there, though? Pages and pages of core morphology would be
> flushed down the same tube.

I say you'd get mergers galore when Alg(onkian) speakers
tried to speak ON. Surely that's where you should start out.
IMHO you should also look at what merger tendencies are/were
already there in Scandinavian.

As for effects on morphology you should be aware that Old
Scandinavian only distinguished three vowels [i/e], [u/o]
and [A] in fully unstressed syllables, and thus in
practically all suffixes and in all inflexional endings.
That's why the Scandinavian Younger Fuþark gets away with
having only three vowel graphemes {a i u} and a
practically superfluous {a~}. Inflexional endings could
always be correctly identified, and roots could be
identified from context.

So starting from the classical ON vowel system

:    i i:    y y:        u   u:
:      e:
:    E E:    9 9:        o   o:
:                    A   Q   Q:
:    Ei  9y  Qu

[2:] and [9:] most likely were in free variation.  Icelandic
has 9: > E: > &: > ai however.

In addition there were nasalized vowels, which were
preserved in 12th century Icelandic but later lost

The Younger Fuþark reduced them in writing like this:

- {i} = i(:), e(:)
- {a} = A, Q(:), E:
- {u} = u(:), o(:), y(:), 2, 9:
- {ai} = Ei
- {au} = Qu, 9y

However I'd suggest that Finlæsk does *not* get exactly
these mergers, since the mergers that actually happened in
Scandinavian languages were different.

The first thing you'd definitely get is that front rounded
vowels lose their rounding (itacism). Since it happened in
Icelandic it's a pretty sure one.

As for reducing the number of vowel heights among the back
vowels I'd say o(:)/u(:) merger is a pretty sure thing too,
since in Scandinavian there was already dialectal variation
between o(:) and u(:) in many items. Moreover you have a
chain shift Q: > o: > u: > u\: in much of Scandinavian,
which may have been a pull
chain, though a push chain is IMHO more likely.  Anyway o(:)
and u(:) are clear merger candidates.

Gutnic (of Gotland in the Baltic) had e: 2: o: > i: y: u:.
In that language as in most peripheral Scandinavian dialects
long vowels later diphthongize in various ways, primary
diphthongs also being preserved.

Q: would most likely become A: to an Alg. speaker.  NB that
short Q mostly reverted to A in Peninsular Scandinavian,
though there are some instances of Q > o and Q > 9.

Interestingly many Peninsular dialects have o > 3\, so you
may get o/Q > @.  Icelandic of course has Q > 9.

Alg. languages allow Vj and Vw, so Ei/9y > ej and Qu > aw
are likely. Q: may go to aw as in Icelandic. Likewise E: may
become any of e:, ej, Aj.

o: may be pushed to ow too, if u: > o:.  Icelandic has o: > ow.

So you definitely can merge it down to i(:), e(:), o(:),
A(:) without violating existing Scandinavian tendencies.

/BP 8^)>
Benct Philip Jonsson -- melroch at melroch dot se

    a shprakh iz a dialekt mit an armey un flot

                                 (Max Weinreich)