U of Michigan digitizing Islamic manuscripts
By Paul Biba
This is from Ann Arbor.com and there is a video that goes along with it.
The University of Michigan Special Collections Library needs help cataloguing its vast Islamic Manuscripts Collection. …
The manuscripts are mostly in Arabic, but also include works in Turkish and Persian, with one in Chaghatai. Works date from about 750 to 1906.
Subjects covered are varied, including Quran texts, commentaries and criticism, Islamic traditions, philosophy, and poetry, history and mathematics, among others - and all of it is hand-written.
“It will be presented to the public in Wiki or blog-type interface, so people can comment on what they see. In that way, we hope we can get help from scholars all over the world in identifying the manuscripts and cataloguing them properly,” said Peggy Daub, director of Special Collections.
Daub hand-selected the texts to be scanned, leaving out only the most delicate. …
Now, the pages are being scanned with special equipment in the Buhr Building by U-M’s in-house digital library production service. The scans can’t be digitized for searching using the typical Optical Character Recognition process because the manuscripts are handwritten. …
The Web site will eventually work with Mirlyn and HathiTrust, a shared digital library repository that stores digitized library content from a number of organizations, including U-M.
The project is funded with a $225,000 grant administered by the Council on Library and Information Resources, with funding from the Andrew W. Mellow Foundation. The principal investigator is Jonathan Rodgers, head of U-M’s Near East division.
August 10th, 2009 at 8:29 pm
It’s interesting that they can’t OCR them. Handwritten Arabic, unless it is unduly caligraphic, doesn’t differ too much from “printed” Arabic, except for some orthographic changes it has undergone through the centuries in different regions (i.e. in the Moroccan “Maghrebi” script, the dots of some letters are place below the line and not above it). This should be less of an issue with more recent texts, even if they are handwritten. Then again, if they are annotated like some of the traditional religious texts that I possess, then the spiraling maze of commentaries and footnotes would be very difficult to compute.