Making a TEI-encoded version of Bédier's 1913 edition of Jean Renart's Lai de l'ombre. Gallica.bnf.fr provides a scan and OCR. Hand work in emacs yields a valid TEI document. Half an hour for an ixml grammar for the foot-of-the-page apparatus, several more hours to tweak it for slips and OCR flaws. And voila, a fully parsed negative apparatus.
Next: align with base text and re-encode in parallel segmentation form.
Hurrah for Gallica, and Emacs, and XQuery, and Invisible XML.
Still thinking about Bédier's Paradox. #Bedier #stemmas
In his book Textkritik, Paul Maas attempts a solution: "Zunächst ist zu erinnern, daß von den 22 stemmatischen Typen, die bei drei Zeugen möglich sind, nur einer dreispaltig ist" / "First, remember that of the 22 [labeled] stemmas possible for three witnesses, only one has three branches [at the root]."
Nice try, but ... why would those 22 forms be equiprobable? Textual traditions don't toss dice to choose the form of their stemmas.
Can Bédier's Paradox be stated in 500 characters?
How likely is it that in a manuscript tradition where the most recent common ancestor of all extant mss is lost, mss will fall into two families, not more? In the stemmas found in editions, the odds are 20:1. Bédier thinks that's high. Is it a miracle or a flaw in the stemmatic method? If the method is flawed, then the editions are flawed.
As far as I can tell, a century on, Bédier still awaits a satisfactory answer.
Michael Sperberg-McQueen