Tuesday, June 4, 2013

Using Technology in Manuscript Studies, Part I

One of the techniques I have used to try to identify a problematic owner's mark was a query in Google Image Search. GIS, for those who many not have used it yet, is a search engine that looks for images rather than (or in addition to) key words. I was not expecting this to offer a solution, necessarily, to the problem that the paleography of the signature poses, but at least to see if it exists in any manuscript other than the one I'm working on. So far, it's not proved to be a hugely useful tool.

This is the image, a Hebrew signature about which I'll write in the next post.

And this is a selection from the first page of results:

Just to summarize the results: Not only did Google Image Search not find another exemplar of this insignia, but it seems to think it is a cross between the strokes of Chinese calligraphy and the curvature of a toilet seat. Partly this is a problem of the image just not having been uploaded to the internet for Google Images to find. But it's also clear to me that the visual search engine needs a good bit of tweaking before it can be useful for paleographic or codicological searches.

To be continued.

No comments:

Post a Comment