a MUCH better photo of the icosahedron

[Gerald McGarvey]

My take on this...

That they are at least read left to right is indicated by 1 being to the left
of the zero in the representation of 10.

Web searches show that nowadays for languages that use Arabic script,
including Arabic, Urdu, Farsi, etc., numbers are written left to right, in 
fact
software for these languages needs to have bidirectional script support.

As to why they are written left to right, this document
http://www.dsv.su.se/~hercules/papers/FarsiSum.pdf
states 'Persian numbers have the same origin as the Latin numbers and
are written left to right.' According to this article they originated in India:
http://en.wikipedia.org/wiki/Arabic_numerals
(there's probably more to it)

I think manuscripts might show signs of the numbers being written left to 
right.
The writer would need to guess the amount of space needed, if they don't guess
correctly,  there could be a gap to the right of the number, or crunched up 
numbers.
That would pretty much be physical proof that numbers are written left to 
right.

More examples...
Persian (Farsi) numbers 0-10:
http://www.omniglot.com/images/writing/persian_num.gif

Kurdish (Kurdí) number 0-10:
http://www.omniglot.com/images/writing/kurdish_num.gif

Gerald

At 02:56 AM 12/17/2006, Ralf Stephan wrote:
>Brendan McKay:
> > Now here is an interesting trivium: Even though Persian and Arabic
> > are written right to left, numbers are written left to right.
> > You are all about to jump up and tell me that the order of writing a
> > decimal positional system is arbitrary, so what I call left to right
> > might as well be called right to left starting at the high-order
> > digit. However, that is not correct and I reassert that numbers are
> > written left to right.  Can anyone guess what my proof is?
>
>The fact that our number system is derived from Arabic,
>and we use it left-to-right?
>
>(Not a hard proof but a good reason.)
>
>
>ralf