Re: [tied] Re: Numerals query again

> > In fact, neither I nor the literature on the subject
> > know of any language which has atoms over 10 (at least diachronically).
>
> Well, there are some primitive systems of counting that have no base
> (let alone an exponential base) but employ names of body parts as
> numerals; these go beyond 10. For example, Telefol (from Papua New
> Guinea, where such systems are rather common) is reported to have 27
> atomic numerals, from <maakub> '1' = 'little finger of the left hand' to
> <kakkat> '27' = 'little finger of the right hand' (just fancy having
> different terms for them!). The middle number <mitkal> '14' is _of
> course_ the word for 'nose' (well, on second thoughts it could be
> something else).

That's right - although the ones I've seen that work like this can be
said to have bases in that they can (and will) reiterate the cycles
for higher numbers. But there are of course numeral systems that do
not go beyond the atomary forms and they cannot be said to have
(sufficiently grammaticalised) bases. In any case, you are indeed right
that bodypart-numbering systems do provide atomary forms over 10 (if
the bodyparts themselves do not have composite names). Telefol has -kal
(meaning 'left'?) in 5-14 but I am sure there are other examples from
PNG that do the trick.

> > The 20-count in the IE languages I mentioned in the original query
> > have all innovated it in that they were originally IE-decimal. But I have
> > a difficulties finding information even on when they got it and so on, the
> > origin question aside. It's hard to believe howvere that Piotr for example
> > doesn't know anything about the distribution and age of the score-counts
> > in older stages of English or that Jens doesn't know how long the Danes
> > have counted so peculiarly :-)
>
> Your faith in us is touching :-). All I know, however, is that the
> _attested_ use of OE scoru in counting goes back to the 11th c. The
> etymological meaning of the word (perhaps a calque from ON skor) was
> 'notch' (as e.g. on a wooden tally for keeping records).

Yes that's consistent with what _Indo-European Numerals_ has to say.
I also recall that it mentions that monastic latin in England does
present some 20-counting which monastic or other latin anywhere else
at the time does not. Just for the record, in ON, did skor only mean notch
and was borrowed/calqued/resembled into English with that meaning only or
was the 20-meaning present in ON as well?

> What's interesting about early Germanic numerals is the use of *xunda-
> as 'approximate hundred' (70-120), as in Old English:
>
> 70 hund-seofon-tig
> 80 hund-eahta-tig
> 90 hund-nigon-tig
> 100 hund-te:on-tig (= hund)
> 110 hund-ændlæf-tig
> 120 hund-twelf-tig

That is indeed interesting? How early is this system attested
btw? 500 A.D or 1000 A.D -ish? Also, what theory do you subscribe to
as to the proto-Germanic 70-120?

> This must have something to do with the use of ON hundrað (and
> occasionally of Eng. hundred) for the "long" or "great" hundred (120).
> It's somethimes called the duodecimal hundred, though strictly speaking
> in a duodecimal system a hundred = 144 (in decimal terms).

Yes. To add some more detail, I believe Old (but not modern) Icelandic
has hundrað-of-the-tens and hundrað-of-the-twelves when there was a
need tp be specific. That hundrað could indeed mean 120 rather than
100 is beyond all doubt when there are examples such as the
three-hundred-five days of the year.

thanks for your reply,

Harald