[tied] Re: Numerals query again

>
> I think it's significant that *(d)wi-(d)k^m.t-ih1, the PIE word for
> '20', involved a dual and that in many branches it came to differ
> structurally from the other decadic terms. As a result, it became
> unanalysable, like the basic number terms (the cardinals from 'one'

to

> 'ten'); the same happened to the word for 'one hundred', *(d)

k^m.tom,

> reanalysed as a simplex primitive term, no longer synchronically
> connected with *dek^m.(t). When the other terms for decads were
> restructured (compounds into phrases), '20' resisted change in

several

> branches and became independent, just like '100'. It was therefore
> available as a counting unit.
>

When people still used cheques and mail orders in Denmark, in the
field that says 'write sum in letters' you were supposed, to preclude
ambiguity, to use a _decimal_ system (at least for decades over
twenty). Since it was used nowhere else, it was made up from scratch,
with no phonological modifications:

treti, firti, femti, seksti, syvti, otti, niti

(but cf. 'fire' "4").

I have no idea where this originated. A wild guess: The Scandinavian
currency union (approx 1870 - WWI)?


Torsten