From: Miguel Carrasquer
Message: 26238
Date: 2003-10-04
>03-10-03 22:05, Rob wrote:Greek eíkosi/Fíkati (*ewi:km.ti) points to *h1wi-h1k^m.tíh1 (presumably
>
>> I hope I'm not bringing up something already established, but I was
>> wondering what your opinions on the etymology of Latin vi:ginti
>> ("twenty") are. Personally, it seems to me that vi:ginti is the
>> result of PIE *dwi(s) kmtih2 'twice ten.'
>
>You're on the right track. The actual reconstruction is *wi-(d)k^m.tih1
>a compound in which the second element is the _dual_ of *dek^m.t- '10'=======================
>(hence the ending *-ih1). Whether *wi- is really a simplified
>(dissimilated?) variant of *dwi-, opinions vary. It may be. If it is,
>the assumption of the adverb 'twice' is not necessary: *dwi- is simply
>the normal form of 'two' in compounds.