"whose"

[David L. White]

>> Yes, the OE form is "hwaes", but I was (over-)generalizing to
>> the rest of Germanic, where /e/ seems to have been the rule.  The usual view
>> is that OE /ae/ got there simply by a difference of opinion about whether
>> to use /o/ or /e/ in IE.

LC > Certainly possible, since other genitives show both vowels.

        Yes, but it is annoying, since we (or at least I) would like to
reconstruct one genitive of the interrogative for Common Germanic, not two.
But such dreams are perhaps naive.

>> Another possibility that occurs to me is that there
>> might have been a change of unstressed /e/ after /w/ (voiced or
>> voiceless) to /ae/, but I do not know if this checks.

LC > I don't think so.  Why would it have been unstressed in an interrogative
> like that?

        Because it's really from unstressed compounds with the interrogative
as second element, not from the true interrogative itself.
The Conventional Wisdom, at least as represented by Wright, derives the
vowel of the nominative from an unstressed form, so I do not see that an
unstressed genitive would be notably problematic.
        A greater problem, in my view, is that the sound change that the
Conventional Wisdom depends on applies, as far as I can see,  to only to
words:  /hwa/ -> /hwaa/, and /swa/ -> /swaa/.  This is seen as lengthening
of short vowels to long, some sort of de facto mora minimum.  But there are
two suspicious problems.  First, the fact that both words have /w/ is not
mentioned as relevant, and is therefore implicitly left as a concidence.
Maybe, but maybe not.  Second, it is not mentioned that /hwaa/ and /swaa/
quite commonly co-occur, as in the usual OE word for 'whoever', "swa hwa
swa".  Some sort of (for lack of a better term) sympathetic association
seems likely.  The sequence is vaguely reduplicative, a fact that speakers
must have been sensitive to.
        So (just when you thought there was no point) the view I would (very
tentatively) take is that 1) /hwaa/ is a plural intruding into the singular,
originally from indefinites.  Or perhaps it is from a sound change that
applied only to unstressed indefinites, but either way indefinites cannot be
avoided, and 2) that /hwaes/ is from /hwes/, with a change or reanalysis of
unstressed /hwes/ to /hwas/, due to influence from the preceding labial,
from which point /hwae/ is regular.  But perhaps there is something fatally
wrong with all this.

Dr. David L. White