🚀 go-pugleaf

In article <memo.857761@cix.compulink.co.uk>, petex@cix.compulink.co.uk (Peter Christian) writes:
> In-Reply-To: <1993Jan11.090344.4952@memstvx1.memst.edu> connolly@memstvx1.memst.edu

[marvelous stuff omitted]

>> A final mystery is why *all* German feminines lack genitive -s.  In Indo-
>> European, the morphology of masculine and feminine i-stems was identical,
>> and had -s,
>> as may readily be seen by looking at Greek and Latin.  But in Old High 
>> German they're different for some reason.  To wit:
>> 
>>                         Masc.           Fem.
>>         Nom. Sing.      gast 'guest'    anst 'favor'
>>         Gen. Sing.      gastes          ensti
>>         Nom. Plural     gesti           ensti
>> 
>> Somehow -s survived in the masculine, but not in the feminine.  Apparently
>> the patterning of *other* masc. resp. fem. categories is to blame.
> 
> Could the answer lie in the coalescence of the masc. a-stems and
> i-stems in the singular? Perhaps the fem has the `correct' reflex of
> IE *-eis and the masc has taken over the reflex of *-os from the
> a-stems. Just a thought, I haven't checked the phonology.

Apparently so.  The IE genitive ending of o-stems was -e/o-s(j)o, while
i-stems had -e/os added to the stem: PIE *ghostej-es etc.  Final -s
generally (but not always!) dropped in Germanic, hence the i-stem geni-
tive ought to be s-less in both genders.  The o-stems kept their -s-
because it was not originally final. -- Phonologically, this works.
But the question is why the masculine i-stems adapted only partially
to the o-stem paradigm.  Seems an awkward way to proceed, though I know
perfectly well that morphological change in generally tends to be messy
in exactly this way.

--Leo Connolly