This is quite amusing, for this is plainly the less smooth of the two
easiest possibilities. We find that some morphemes of non-sg. number are
consistently plural in non-Anatolian IE, while other are consistently dual
in non-Anatolian IE (caring only about non-Anatolian languages that
possess a dual category). Now we want to explain why that is so. Many
believe that the old functions of the number markers have been retained
outside Anatolian and have merged in Anatolian (and some of the
non-Anatolian languages). That scenario offers no problems I can see. The
other way of looking at it is to imagine that the number markers existed
in pairs of synonyms in the parent language and then leave it up to the
non-Anatolian languages to differentiate them in a later period, each time
reserving one morpheme for dual function and another for the plural. Are
there parallels for anything like that? And why can't the actual process
of change be recovered in any detail? Is that a sign of its being very
young? And why can we establish element functions between IE animate dual
markers and non-IE dual markers if none of them ever were dual markers in
a shared past?
So we are to believe that the dual forms existed in the PIE ancestral to
both Anatolian and the rest, only in that language they functioned as
plural forms? And it is a complete mystery how some of them developed dual
function outside of Anatolian? Why is such silliness preferred?
Jens
>
> Jens:
>>Also the Hittite ntr.pl. forms in -i must come from somewhere.
>
> Yes, the Hittite ntr.pl. forms come from somewhere... but why
> the "dual" per se? We know that *-x (*-h2) is used in a
> collective, non-singular sense, unspecific to dual or plural.
>
>
> Miguel:
>>There's also the compund noun huhha-hanna- "grandparents",
>>where the first member (huhha-) appears to be in the dual.
>
> That's not very apparent at all.
>
>
> Jens:
>>It fits the dual of the other languages, so does the ntr.pl. -e of the
>> enclitic stem -a-, and so do the -w- forms of the 1pl in the verb.
>
> Obviously, Hittite endings are going to compare with dual endings
> of other IE languages if the rest of IE innovates existing endings for
> the purposes of marking a new dual number that originally
> didn't exist! There is nothing inheirantly "dual" about 1pp *-wes
> and is used in Anatolian languages to convey the NON-SINGULAR,
> both dual AND plural.
>
>
> - gLeN
>
> _________________________________________________________________
> The new MSN 8: smart spam protection and 2 months FREE*
> http://join.msn.com/?page=features/junkmail
>
>
>
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/