From: Miguel Carrasquer
Message: 32436
Date: 2004-05-01
>> >Miguel wrote:Yes.
>> If you can find a Celtiberian demonstrative masc. gen.
>> *sosyo, that would in itself contradicty my theory (I would
>> definitely expect *so). All the forms I've been able to
>> trace are a nom. n. <soz>, a dat. <somui>, what looks like a
>> loc. <somei>, a gen.pl. <soisum>. From *yos- we also have
>> nom. <ios>, acc. <iom>.
>
><iomui> is also attested...
>I don't think that *sosyo is attested in Celtiberian, but it is attested inSince Gaul. <sosio> is NA n. sg. "this"(*), and not a
>Gallic <sosio> (does that affect your theory on -i: < *-osyo?).
>> If *<sosyo wiri:> was analogically altered to *<sosyo wiro>,Stifter's 'Einführung in das Kontinentalkeltische' lists
>> why then didn't *<sosyo Koitunos> "of Koitu here" become
>> *<sosyo Koituno>?
>
>It could be explained by the connection of o- and a:-stems, like
>analogically Ab sg -az in a:-stems from o-stem -uz (I am not sure if this -z
>is attested in other declensional types like in Latin, if it is this is not
>a good example).
>They are also pairs in adjectives (*newos - *neweh2 etc.)Not the closest from a phonological point of view. It's a
>so it could be argued that the analogy affected only the closest paradigm
>(o-stems).