I picked up this amazing little book the other day. It explains how linguistics (specifically the frequency of letters) has affected Arabic *calligraphy*. It's in French, so it'll take me some time to read through, but it appears to derive some very interesting conclusions from some seriously quantitative analysis, although it's all tempered with a "big picture" approach.

The author shows wonderful examples of calligraphic art, but couples them to statistical graphs! The type of dual-discipline thing you almost never see - highly fascinating. And it also contains some great data (like a graph showing the relative letter usage of 7 Latinate languages), plus a killer bibliography.

http://mywebpage.netscape.com/atanasiuvlad/frq/

hhp